Icru Report 74

doi:10.1093/jicru/ndi016

PATIENT DOSIMETRY FOR X RAYS USED IN MEDICAL IMAGING

Report Committee

J. Zoetelief (chairman), Faculty of Applied Sciences, Delft University of Technology,Delft, Netherlands

D. R. Dance, The Royal Marsden NHS Foundation Trust, London, United KingdomG. Drexler, GSF-National Research Center for Environment and Health, Neuherberg, Germany and

State University of Rio de Janeiro, Rio de Janeiro, BrazilH. Jarvinen, Radiation and Nuclear Safety Authority (STUK), Helsinki, Finland

M. Rosenstein, Clarksburg, Maryland, USA

Commission Sponsors

H. G. Paretzke, GSF-National Research Center for Environment and Health, Neuherberg, GermanyK. Doi, The University of Chicago, Chicago, Illinois, USA

A. Wambersie, Universite Catholique de Louvain, Brussels, Belgium

Consultants to the Report Committee

P. Allisy-Roberts, Bureau International des Poids et Mesures, Sevres, FranceH. Bosmans, University Hospital Gasthuisberg, Leuven, BelgiumC. J. Moretti, National Physical Laboratory, Teddington, UK

J. Van Dam, University Hospital Gasthuisberg, Leuven, BelgiumE. Va~nnoo, Complutense University, Madrid, Spain

B. F. Wall, Radiation Protection Division, Health Protection Agency, Chilton, UK

The Commission wishes to express its appreciation to the individuals involved in the preparation of this report,for the time and efforts which they devoted to this task and to express its appreciation to the organizations withwhich they are affiliated.

All rights reserved. No part of this book may be reproduced, stored in retrieval systems or transmitted in anyform by any means, electronic, electrostatic, magnetic, mechanical photocopying, recording or otherwise,without the permission in writing from the publishers.

British Library Cataloguing in Publication Data. A Catalogue record of this book is available at the BritishLibrary.

ISBN 0199203208

Journal of the ICRU Vol 5 No 2 (2005) Report 74 doi:10.1093/jicru/ndi017Oxford University Press

THE INTERNATIONAL COMMISSION ON RADIATION UNITSAND MEASUREMENTS

INTRODUCTION

The International Commission on Radiation Unitsand Measurements (ICRU), since its inception in1925, has had as its principal objective the develop-ment of internationally acceptable recommenda-tions regarding:(1) quantities and units of radiation and radio-

activity,(2) procedures suitable for the measurement and

application of these quantities in clinical radio-logy and radiobiology, and

(3) physical data needed in the application of theseprocedures, the use of which tends to assureuniformity in reporting.

The Commission also considers and makes similartypes of recommendations for the radiation protec-tion field. In this connection, its work is carried outin close cooperation with the International Commis-sion on Radiological Protection (ICRP).

POLICY

The ICRU endeavors to collect and evaluate thelatest data and information pertinent to the pro-blems of radiation measurement and dosimetry andto recommend the most acceptable values and tech-niques for current use.The Commissions recommendations are kept

under continual review in order to keep abreast ofthe rapidly expanding uses of radiation.The ICRU feels that it is the responsibility of

national organizations to introduce their owndetailed technical procedures for the developmentand maintenance of standards. However, it urgesthat all countries adhere as closely as possible tothe internationally recommended basic concepts ofradiation quantities and units.The Commission feels that its responsibility lies in

developing a system of quantities and units havingthe widest possible range of applicability. Situationsmay arise from time to time when an expedientsolution of a current problem may seem advis-able. Generally speaking, however, the Commission

feels that action based on expediency is inadvisablefrom a long-term viewpoint; it endeavors to baseits decisions on the long-range advantages to beexpected.The ICRU invites and welcomes constructive com-

ments and suggestions regarding its recommenda-tions and reports. These may be transmitted to theChairman.

CURRENT PROGRAM

The Commission recognizes its obligation to pro-vide guidance and recommendations in the areas ofradiation therapy, radiation protection, and the com-pilation of data important to these fields, and toscientific research and industrial applications ofradiation. Increasingly, the Commission is focusingon the problems of protection of the patient andevaluation of image quality in diagnostic radiology.These activities do not diminish the ICRUs commit-ment to the provision of a rigorously defined set ofquantities and units useful in a very broad range ofscientific endeavors.The Commission is currently engaged in the

formulation of ICRU reports treating the followingsubjects:

Approaches to the Dosimetry of Low-Dose Exposures toIonizing Radiation

Assessment of Image Quality in Nuclear MedicineBone DensitometryDoses for Cosmic Ray Exposure for AircrewDose and Volume Specifications for Reporting Intracavi-tary Therapy in Gynecology

Dosimetry Systems for Radiation ProtectionElastic Scattering of Electrons and PositronsImage Quality and Patient Exposure in CTMammography------Assessment of Image QualityMeasurement Quality Assurance for IonizingRadiation

Prescribing, Recording, and Reporting Conformal PhotonBeam Therapy

Prescribing, Recording, and Reporting Proton BeamTherapy

Requirements for Radiological SamplingROC Analysis

International Commission on Radiation Units and Measurements 2005

In addition, the ICRU is evaluating the possibi-lity of expanding its program to encompass non-ionizing radiation, particularly the quantities andunits aspects.The Commission continually reviews radiation

science with the aim of identifying areas where thedevelopment of guidance and recommendations canmake an important contribution.

THE ICRUS RELATIONSHIP WITH OTHERORGANIZATIONS

In addition to its close relationship with the ICRP,the ICRU has developed relationships with otherorganizations interested in the problems of radiationquantities, units, and measurements. Since 1955,the ICRU has had an official relationship with theWorld Health Organization (WHO), whereby theICRU is looked to for primary guidance in mattersof radiation units and measurements and, in turn,the WHO assists in the worldwide dissemination ofthe Commissions recommendations. In 1960, theICRU entered into consultative status with theInternational Atomic Energy Agency (IAEA). TheCommission has a formal relationship with the Uni-ted Nations Scientific Committee on the Effects ofAtomic Radiation (UNSCEAR), whereby ICRUobservers are invited to attend annual UNSCEARmeetings. The Commission and the InternationalOrganization for Standardization (ISO) informallyexchange notifications of meetings, and the ICRUis formally designated for liaison with two of theISO technical committees. The ICRU also corres-ponds and exchanges final reports with the followingorganizations:

Bureau International de Metrologie LegaleBureau International des Poids et MesuresEuropean CommissionCouncil for International Organizations of Medical

SciencesFood and Agriculture Organization of the United NationsInternational Committee of PhotobiologyInternational Council of Scientific UnionsInternational Electrotechnical CommissionInternational Labor OfficeInternational Organization for Medical PhysicsInternational Radiation Protection AssociationInternational Union of Pure and Applied PhysicsUnited Nations Educational, Scientific and Cultural

Organization

The Commission has found its relationship withall of these organizations fruitful and of substantial

benefit to the ICRU program. Relations with theseother international bodies do not affect the basicaffiliation of the ICRU with the InternationalSociety of Radiology.

OPERATING FUNDS

In recent years, principal financial support hasbeen provided by the European Commission, theNational Cancer Institute of the U.S. Departmentof Health andHuman Services and the InternationalAtomic Energy Agency. In addition, during the last10 years, financial support has been received fromthe following organizations:

Belgian Nuclear Research CentreCanadian Nuclear Safety CommissionEastman Kodak CompanyElectricite de FranceFuji Medical SystemsHitachi, Ltd.International Radiation Protection AssociationInternational Society of RadiologyIon Beam ApplicationsItalian Radiological AssociationJapan Industries Association of Radiological SystemsJapanese Society of Radiological TechnologyMDS NordionNational Institute of Standards and TechnologyNederlandse Vereniging voor RadiologiePhilips Medical Systems, IncorporatedRadiation Research SocietySiemensVarian

In addition to the direct monetary support pro-vided by these organizations, many organizationsprovide indirect support for the Commissions pro-gram. This support is provided in many forms,including, among others, subsidies for (1) the timeof individuals participating in ICRU activities,(2) travel costs involved in ICRU meetings, and (3)meeting facilities and services.In recognition of the fact that its work is made

possible by the generous support provided by all ofthe organizations supporting its program, the Com-mission expresses its deep appreciation.

Andre WambersieChairman, ICRU

Brussels, Belgium

PATIENT DOSIMETRY FOR X-RAYS USED IN MEDICAL IMAGING


PATIENT DOSIMETRY FOR X RAYS USED INMEDICAL IMAGING

CONTENTS

PREFACE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

GLOSSARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

ABSTRACT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

EXECUTIVE SUMMARY . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

1 INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

1.1 Evolution of radiation dosimetry in medical x-ray imaging . . . . . . . . . . . . . . . 91.2 Risks for the patient in radiological imaging and relevant

dosimetric quantities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111.2.1 Acute deterministic effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111.2.2 Late effects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

1.2.2.1 Cancer induction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121.2.2.2 Late effects in normal tissues . . . . . . . . . . . . . . . . . . . . . . . . . . . 121.2.2.3 Impairment of mental development . . . . . . . . . . . . . . . . . . . . . . . 121.2.2.4 Genetic risk . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

1.2.3 Relevant dosimetric quantities and dosimetric procedures . . . . . . . . . . . . . . . 121.2.4 Required accuracy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

1.3 Dosimetry in radiology: relevant quantities . . . . . . . . . . . . . . . . . . . . . . . . . 131.3.1 Calibration at the Standards Laboratory . . . . . . . . . . . . . . . . . . . . . . . . . 131.3.2 From air kerma free-in-air to absorbed dose in water in patient or phantom . . . . 131.3.3 Air kerma-area product (KAP) and dose----area product (DAP) . . . . . . . . . . . . . 141.3.4 Reporting patient irradiation in radiological imaging . . . . . . . . . . . . . . . . . 14

1.3.4.1 Radiological parameters of the exposure . . . . . . . . . . . . . . . . . . . . 141.3.4.2 Air kerma----area product (KAP) or dose----area product (DAP) . . . . . . . . 151.3.4.3 Monte Carlo computation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151.3.4.4 Phantoms and in vivo measurements. . . . . . . . . . . . . . . . . . . . . . . 15

1.3.5 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161.4 Need for harmonization of quantities and terminology . . . . . . . . . . . . . . . . . 171.5 The two purposes of patient dosimetry . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18

1.5.1 To set and check standards of good practice . . . . . . . . . . . . . . . . . . . . . . . 181.5.2 To assist in assessing detriment or harm . . . . . . . . . . . . . . . . . . . . . . . . . 18

1.6 Relationship between patient dose and image quality . . . . . . . . . . . . . . . . . . 181.7 Scope of the report . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

2 SPECIFICATION OF X-RAY BEAMS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

2.1 Photon spectrum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212.2 Half-value layer . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222.3 X-ray tube voltage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232.4 Total filtration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 242.5 X-ray tube output . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

3 QUANTITIES AND UNITS FORMEASUREMENT AND CALCULATION IN MEDICALX-RAY IMAGING . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25

3.1 Basic dosimetric quantities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 253.2 Application-specific quantities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

3.2.1 Incident air kerma and incident air kerma rate . . . . . . . . . . . . . . . . . . . . . 283.2.2 Entrance-surface air kerma and entrance-surface air kerma rate . . . . . . . . . . . 293.2.3 Air kerma----area product and air kerma----area product rate . . . . . . . . . . . . . . 293.2.4 Air kerma----length product . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293.2.5 CT air-kerma index free-in-air . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 293.2.6 CT air-kerma index in the standard CT dosimetry phantoms . . . . . . . . . . . . . 303.2.7 Weighted CT air-kerma index and normalized weighted CT air-kerma index . . . 303.2.8 CT air kerma-length product . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30

3.3 Risk-related quantities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303.3.1 Absorbed dose in relation to deterministic effects . . . . . . . . . . . . . . . . . . . . 313.3.2 Absorbed dose for assessment of stochastic effects (organ dose) . . . . . . . . . . . . 313.3.3 Equivalent dose and effective dose . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

3.4 Dose-conversion coefficients for assessment of organ and tissue doses . . . . . . . 323.5 Quantities recommended for establishment and use of diagnostic

reference levels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333.5.1 Incident air kerma and entrance-surface air kerma . . . . . . . . . . . . . . . . . . . 34

3.5.1.1 Mean mammary glandular dose . . . . . . . . . . . . . . . . . . . . . . . . . 343.5.2 Incident air kerma rate and entrance-surface air kerma rate . . . . . . . . . . . . . 343.5.3 Air kerma----area product . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 343.5.4 CT Air kerma----length product, PDL,CT . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

4 MEASUREMENTMETHODS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

4.1 Quality assurance of dosimeters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 354.1.1 Calibration of dosimeters in terms of air kerma free-in-air . . . . . . . . . . . . . . . 364.1.2 Calibration of air kerma----area product meters . . . . . . . . . . . . . . . . . . . . . . 374.1.3 Calibration of thermoluminescent dosimeters . . . . . . . . . . . . . . . . . . . . . . 38

4.2 Measurement methods for specific dosimetric quantities . . . . . . . . . . . . . . . . 394.2.1 Dosimeters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

4.2.1.1 Ionization chamber dosimeters . . . . . . . . . . . . . . . . . . . . . . . . . . 404.2.1.2 Thermoluminescent dosimeters . . . . . . . . . . . . . . . . . . . . . . . . . 404.2.1.3 Scintillation dosimeters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 404.2.1.4 Film dosimeters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

4.2.2 Incident air kerma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 414.2.3 Entrance-surface air kerma . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 424.2.4 Air kerma----area product . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 424.2.5 CT air-kerma index and CT air-kerma index in the standard CT head

and body dosimetry phantoms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 444.2.5.1 Pencil ionization chamber dosimeter . . . . . . . . . . . . . . . . . . . . . . 444.2.5.2 Thermoluminescent dosimeters . . . . . . . . . . . . . . . . . . . . . . . . . 47

4.3 Features of measurements on patients and measurements withphysical phantoms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47


4.4 Skin dose determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 484.4.1 Direct measurement of the maximum skin dose . . . . . . . . . . . . . . . . . . . . . 48

4.4.1.1 Skin dose measurements on patients with thermoluminescentdosimeters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

4.4.1.2 Skin dose measurements on patients with scinillation dosimeters . . . . . 504.4.1.3 Skin dose measurements on patients with film dosimeters . . . . . . . . . 50

4.4.2 Derivation of the skin dose from the air kerma----area product PKA . . . . . . . . . . 504.4.3 Derivation of the skin dose directly from the radiological parameters

of the exposure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

5 METHODS FOR DETERMINING ORGAN AND TISSUE DOSES . . . . . . . . . . . . . . . . . . . . . . . 55

5.1 Dose measurements in physical phantoms . . . . . . . . . . . . . . . . . . . . . . . . . . 555.2 Monte Carlo radiation transport calculations . . . . . . . . . . . . . . . . . . . . . . . 56

5.2.1 Main features of the Monte Carlo technique . . . . . . . . . . . . . . . . . . . . . . . 565.2.2 Main features of the computational models of the human body . . . . . . . . . . . . 56

5.2.2.1 Mathematical phantoms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 565.2.2.2 Special features of the active bone marrow . . . . . . . . . . . . . . . . . . . 575.2.2.3 Voxel phantoms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 58

5.2.3 Uncertainties in Monte Carlo organ-dose calculations . . . . . . . . . . . . . . . . . 585.2.4 Comparison of conversion coefficients calculated at different institutes . . . . . . . 595.2.5 Comparison of measured and calculated organ doses . . . . . . . . . . . . . . . . . . 59

5.2.5.1 Adult phantoms: organs in the x-ray field . . . . . . . . . . . . . . . . . . . 595.2.5.2 Adult phantoms: organs outside the x-ray field . . . . . . . . . . . . . . . . 605.2.5.3 Adult phantoms: active bone marrow . . . . . . . . . . . . . . . . . . . . . . 605.2.5.4 Paediatric phantoms: head and neck . . . . . . . . . . . . . . . . . . . . . . 605.2.5.5 Paediatric phantoms: whole body . . . . . . . . . . . . . . . . . . . . . . . . 605.2.5.6 Adult phantoms: CT. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61

5.2.6 Sources of data on dose-conversion coefficients . . . . . . . . . . . . . . . . . . . . . . 61

6 CONCLUSIONS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63

APPENDIX A BACKSCATTER FACTORS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 65

APPENDIX B HANDBOOKS PRODUCED BY THE CENTER FOR DEVICESAND RADIOLOGICAL HEALTH (CDRH) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69

APPENDIX C REPORTS PRODUCED BY THE GERMAN NATIONAL RESEARCHCENTER FOR ENVIRONMENT AND HEALTH (GSF) . . . . . . . . . . . . . . . . . . . . . 79

APPENDIX D REPORTS PRODUCED BY THE HEALTH PROTECTION AGENCY (HPA) (FORMERLYNATIONAL RADIOLOGICAL PROTECTION BOARD) (NRPB) . . . . . . . . . . . . . . . . 87

APPENDIX E REVIEW OFMONTE CARLO CALCULATIONS FOR ASSESSMENT OF MEANGLANDULAR DOSE IN MAMMOGRAPHY . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93

APPENDIX F PCXMC ------ A PC-BASED MONTE CARLO PROGRAM FOR CALCULATINGPATIENT DOSES IN MEDICAL X-RAY EXAMINATIONS . . . . . . . . . . . . . . . . . . . 99

REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 103

CONTENTS


PREFACE

The mission and main objective of the Interna-tional Commission on Radiation Units andMeasure-ments (ICRU) are to develop a coherent system ofradiological quantities and units that is acceptedworldwide and applied in all fields where ionizingradiation is used. The ICRU also develops recom-mendations on how to measure radiation-relatedquantities to ensure a reliable exchange of results.In addition, within the framework of this mission,

there is often a need for the definition of new termsand concepts that could be adopted universally. Theultimate goal of the ICRU is to improve harmoniza-tion in the concepts and themethods used to describeand to report radiation applications, and therebyfacilitate the exchange of information between cen-tres using radiation in medicine, science, and indus-try. The foundation of the ICRU by the FirstCongress of Radiology in 1925 was to solve exactlythis harmonization problem.The present Report is the first report published

by the ICRU that deals with patient dosimetry forx rays used in diagnostic medical imaging. Theimpetus for this report derives from the broad andsystemic application of x rays for diagnostic andinterventional imaging. The increasing number ofpatients that benefits from radiology and theincreasing number and types of procedures that areapplied to these patients have resulted in a dramaticincrease of the population dose, which, in developedcountries, often exceeds the natural radiation levels.The situation in developing countries will sooner orlater exhibit the same trend (UNSCEAR, 2000).The relation between image quality and patient

dose has always been a matter of concern for theradiology community. To initiate the production ofobjective information, the European Commissionconducted several trials for various types of exam-ination currently performed in diagnostic radiology.As an example, a first trial, involving 24 radiologydepartments from 10 European countries(1987/1988) compared entrance doses for PA chestradiography. The entrance surface doses rangedfrom 0.03 to 12 mGy, i.e., a ratio of about 400 (Macciaet al., 1989). A second larger study (1991) involved83 radiology departments from 16 countries. As

an example, in the second study, for PA chest radio-graphs, the mean entrance doses measured inthe participating departments ranged from 0.1 to0.5 mGy, i.e., a ratio of 5 between the maximumand the minimum mean doses (EC, 1996a). Thisratio was worse in the first study. Most interestingis the fact that there was no correlation between thequality of the image and the dose to the patient. Thequality of the images was evaluated by the informa-tion content of the film as assessed by a team ofexperienced radiologists. Several other studies ofthis kind were initiated (e.g., for breast, lumbarspine), and quality criteria were established (CEC,1990; EC, 1996b; ICRU, 2003).A particular source of concern is that among the

different examinations some modern CT proceduresthat are remarkably powerful in their diagnosticcapabilities deliver significant doses to large regionsof the body. The diagnostic power of the radiologicalprocedures to solve medical issues is of course thefirst priority. Because of the increasing dosesdelivered to an increasing number of patients, how-ever, it becomes important and timely to optimizethe technical conditions, i.e., to reduce the patientexposure for the same quality of diagnostic informa-tion. This is simply common sense and is in agree-ment with the recommendations of the radiationprotection commissions and agencies, and also ofnational and international authorities.An additional issue is the recent development and

rapid growth of interventional radiology, especiallyin cardiology. The exposures are high for the patient(and possibly also for the radiologists), and the num-ber of reported cases of acute tissue reactions1 withdifferent severity is increasing. The need for accur-ate dosimetry becomes critical especially for theskin, which is one of the tissues at highest risk.For patient dosimetry in radiology, the required

accuracy depends on the clinical situation and thedose range involved. It is, in general, much lowerthan the accuracy required in radiation therapy. Inany case the quantities measured should always be

1The ICRP has recently proposed (2005) to replace theterm deterministic effects by tissue reactions.


clearly identified. The relevant quantities to bedetermined are those most closely related to the bio-logical effects or risks of such effects. Presently, theavailable data establish correlation between the bio-logical effects and absorbed dose at the point or inthe volume of interest.For the low doses delivered by most of the current

procedures in diagnostic radiology, cancer induction(stochastic effect) is considered to be the main risk.In contrast to diagnostic radiology, at the high dosesdelivered, for example, during interventional radi-ology, acute effects become the major source of con-cern (deterministic effect). Late effects resultingfrom acute effects constitute particularly dangerouspre-cancerous lesions. This is well established forthe skin and is now being investigated for the rectumafter radiotherapy of prostate and cervix tumors.Previous ICRU Reports have dealt with patient

dosimetry for external beam therapy with photons(ICRU Report 42 1987; ICRU Report 50, 1993; ICRUReport 62, 1999; ICRU Report 64, 2001), with elec-trons (ICRU Report 71, 2004a), protons (ICRUReport 59, 1998b), and neutrons (ICRU Report 45,1989a), for brachytherapy, and for b-ray applica-tions (ICRU 38, 1985a; ICRU Report 58, 1997;ICRU Report 72, 2004b). ICRU Reports 32 (1979)and 67 (2002) dealt with patient dosimetry indiagnostic nuclear medicine procedures.Some aspects of the dosimetry of x rays generated

at tube voltages ranging from 5 to 150 kV were dis-cussed in ICRU Report 17 (1970) but were not direc-ted specifically at patient dosimetry in diagnostic orinterventional radiology.In the field of radiation protection, the ICRU has

also provided advice on the determination of doseequivalents from sources of radiation external tothe body (ICRU Report 39, 1985b; ICRU Report 43,1988; ICRU Report 47, 1992a; ICRU Report 57,1998a). The advice involved the use of operationalquantities suitable for practical measurements forthe evaluation of occupational exposures. In particu-lar, these operational quantities introduced by theICRU facilitate an adequate and conservative estim-ate of effective dose, that is, the protection quantitydefined by the International Commission on Radio-logical Protection (ICRP) for use in its system ofradiological protection (ICRP Publication 60, 1991a).Exposure to ionizing radiation due to medical

x-ray imaging entails the well-defined irradiationof localized parts of the body. Diverging beams of

x rays, emitted from a point source and character-ized by half-value layers of 0.310 mm of Al, arecollimated to penetrate the volume of interest. Theuse of different irradiation conditions, in terms ofincident radiation quality and beam geometry inrelation to the patients body, has led to the develop-ment of specific dosimetric methods and the defini-tion of appropriate quantities quantities differentfrom those used for occupational and environmentalexposures.Not surprisingly, the exposure conditions

assumed in deriving the relationships between theeffective dose and the operational quantities foroccupational and environmental exposures are notappropriate for patient dosimetry in medicalimaging. In the first case one is dealing with wholebody irradiation by broad beams of photons or neut-rons while, in the second case, strictly collimatedbeams are used resulting in partial-body patientirradiations.Whereas some of the dosimetric concepts and tech-

niques used in radiotherapy have been successfullyemployed in medical x-ray imaging, additional dosi-metric quantities and measurement methods arerequired for patient dosimetry for procedures suchas fluoroscopy, CT, and mammography. Conversioncoefficients are often used in practice to relate dir-ectly measurable quantities to doses to differentcritical organs or at specific reference points. Whendeterministic effects are considered a possibility,doses to the more heavily irradiated sites of thebody need to be critically evaluated.The present Report provides a detailed framework

of recommendations for assessing patient dose inradiological imaging. Moreover, this framework issuitable for the accurate, harmonized exchange ofinformation as well as to provide an assessment toavoid or reduce the severity of tissue reactions. Thisreport will be soon followed by a second one, exclus-ively focused upon CT dosimetry and its image qual-ity. The recent development and rapid growth of CTapplications, and the specific issues that are raised,deserve a special ICRU Report.

Andre WambersiePaul M. DeLuca

Johannes ZoeteliefSeptember 2005

PATIENT DOSIMETRY FOR X RAYS USED IN MEDIUM IMAGING

2


GLOSSARY

AEC automatic exposure controlAP anteroposterior view: x rays enter from the front of a patientBM bone marrowCC cranio-caudal view: x rays enter on the top of the head of a patientCT computed tomographyCTDI computed tomography dose indexDICOM digital imaging and communications in medicineDRL diagnostic reference levelFID focal spot-to-image receptor distanceFOV field of viewFSD focal spot-to-surface distanceGI gastro-intestinalGSD genetically significant doseHVL half-value layerLAT lateral incidence of radiationLAO left anterior oblique view: x rays enter right rear side of patient and form an image on the left front

sideL LAT left lateral view: x rays enter from the right side of patient and form an image on the left sideLPO left posterior oblique view: x rays enter right front side of patient and form an image on the left rear

side (All oblique views lie in a transverse plane and form a 45 angle with the AP/PA axis.)LSJ lumbosacral jointML medio-lateralMRI magnetic resonance imagingOBL obliquePA posteroanterior view: x rays enter rear of patientPMMA polymethylmethacrylate plasticRAO right anterior oblique view: x rays enter left rear side of patient and form an image on the right front

sideR LAT right lateral view: x rays enter left side of patient and form an image on the right sideRPO right posterior oblique view: x rays enter left front side of patient and form an image on the right

rear sideTLD thermoluminescent dosimeterTMJ temporomandibular joint



ABSTRACT

This report presents specifications of x-ray beamsand quantities, and units for dose measurement andcalculation in medical x-ray imaging, includingapplication-specific quantities, and new symbols. Itaddresses measurement methods for normalizationquantities and for quantities recommended for theestablishment and use of diagnostic reference levels.It presents methods for determining organ and tis-sue doses as well as doses in localized regions oforgans and tissues, including detailed information

on dose-conversion coefficients for x-ray imagingfields. This is the first ICRU report dealing withmethods for patient dosimetry of x rays used in med-ical imaging. Whereas some of the dosimetric con-cepts and techniques used in radiotherapy have beensuccessfully employed for medical imaging using xrays, additional dosimetric quantities and measure-ment methods are required for patient dosimetryassociated with procedures such as radiography, flu-oroscopy, CT, and mammography.



EXECUTIVE SUMMARY

In Section 1 it is emphasized that medicalimaging was virtually the first application ofx rays. The health risks of irradiation becameevident only later. Appropriate quantities to meas-ure the amount of irradiation of an object had to bedeveloped, leading to quantities like exposure, ab-sorbed dose, and kerma. It is furthermore stressedthat there is a need for harmonization of quantitiesand terminology for different applications in med-ical imaging using x rays. The two purposes ofpatient dosimetry of x rays used in medical imagingare to set and check standards of good practice andto assess the risks of detriment or harm. Imagequality is stressed to be of paramount importancein medical imaging but is the subject of other ICRUReports.Specifications of x-ray beams used for medical

imaging are presented in Section 2. It is recommen-ded to characterize the radiation quality of x-raybeams used for medical imaging by a combinationof various parameters, including first and secondhalf-value layer, HVL1 and HVL2, the ratio ofHVL1 and HVL2, the tube voltage, and the totalfiltration. In most cases a combination of three ofthese parameters will be sufficient for character-ization. The radiation intensity is also an import-ant characteristic of an x-ray tube (includingfiltration). For this purpose the x-ray tube outputis defined.Quantities and units for dose measurement

and calculation in medical x-ray imaging aredealt with in Section 3. Relevant basic dosimetricqualities are presented in first instance. Severalapplication-specific qualities have been found use-ful for measurements in medical x-ray imaging,but ambiguity remains in the names of quanti-ties and their use. Therefore, application-specificquantities and new symbols are defined. Concern-ing risk-related quantities, mean organ and tissuedoses are defined as well as absorbed dose to themore heavily irradiated regions of tissues in rela-tion to deterministic effects. Dose-conversion coef-ficients relate the specified dosimetric quantities toa normalization quality. Both types of dosimetric

quantity are discussed in Section 4. Quantities arealso recommended for establishment and use ofdiagnostic reference levels (DRLs). It should benoted that the recommendations made here of dosi-metric quantities for CT are of a provisionalnature.Methods for assessing the patient dose of x rays

used in medical imaging are presented in Section 4.Such methods are required for the determination ofnormalization quantities used in dose-conversioncoefficients and of quantities recommended forestablishment and use of DRLs. Measurements ofthese quantities rely mainly on the use of ionizationchambers or solid-state dosimeters, including TLDs.For the low- and medium-energy x rays used inmedical imaging, the air kerma free-in-air is thedesired quality for calibration. Examples are givenof x-ray beam qualities recommended for calibration,but it is noted that an international code of practicefor patient dosimetry in diagnostic radiology is pres-ently being developed by the International AtomicEnergy Agency, including practical details ofcalibrations.Methods for determining organ and tissue doses

are the subject of Section 5. It is concluded thatassessment of organ and tissue doses, as well asdoses to the most heavily irradiated regions of thebody, mainly relies on Monte Carlo calculations.Therefore, specific information is provided on theapplication of Monte Carlo calculations of radiationtransport as employed for patient dosimetry in med-ical x-ray imaging. This section includes comparis-ons of dose-conversion coefficients calculated atdifferent institutes as well as comparisons of dosemeasurements and calculations. It is noted that pro-cedures for medical x-ray imaging vary from countryto country. Consequently, dose-conversion coeffi-cients calculated by various authors commonlyrefer to national or regional imaging procedures.For similar exposure conditions, similar results areobtained from calculations at different institutesand from measurements. When a dose-conversioncoefficient is needed for a specific situation, thebest approach is to select a value from the available


data based on similarities in exposure conditions(projection, view, and radiation quality) and apatient model.Appendix A presents information on backscatter

factors for irradiation conditions relevant for dosi-metry for medical imaging using x rays. AppendicesB, C, and D provide dose-conversion coefficients that

reflect the differences in medical imaging using xrays in the USA, Germany, and the UK, respect-ively. Appendix E treats dose-conversion coefficientsfor mammography. Appendix F describes a PC-based Monte Carlo program for calculating patientdoses in radiography (excludingmammography) andin fluoroscopy.


8


1 INTRODUCTION

1.1 EVOLUTION OF RADIATION DOSIMETRYIN MEDICAL X-RAY IMAGING

The need for accurate measurement of x rays wasimmediately evident in the early days of x-ray use inmedicine. This was not only because of the observedbiological effects induced by the exposure to x rays butwas also because of the instability in the generation ofx rays. The firstmeasurement techniquesmade use ofthe blackening of photographic emulsions and colourchanges of chemical compounds. However, no unit ofexposure existed and the development of instru-mentation was, therefore, not very straightforward.In 1905 at the Rontgen Congress in Berlin, a com-

mittee was founded to define a unit for the measure-ment of Roentgen rays. Only after the suggestion byVillard (1908) to make use of the change in electricalconductivity of air by irradiation with x rays for theirquantification was a promising direction of experi-mental attempts created. This resulted in 1925 inthe adoption, by the German Rontgen Society, of therontgen, R, as the unit of x-ray exposure. The sameyear, at the First International Congress of Radi-ology in London, a joint meeting of the Physics andRadiology Sections was devoted to a discussion oninternational standards for x-ray work. It resulted inthe appointment of the International X-ray UnitCommittee, with the mission to establish a uniformx-ray standard of intensity and an x-ray unit. ThisCommittee became later on the International Com-mission on Radiation Units and Measurements(ICRU) (Taylor, 1958). The Second InternationalCongress of Radiology in Stockholm (1928) recom-mended the definition of the Rontgen, which is inessence the unit proposed by the German physicists,in which the term Rontgen-ray dose was replaced bythe term quantity of radiation. The initial definitionof the unit was rephrased by the ICRU in 1931 and1937: the essence of the unit was not changed but thephysical quantity of which the rontgen is the unithad not been clearly defined (Allisy, 1995).Early measurements with ionization chambers in

diagnostic radiology were difficult for two reasons:

(i) The first problem was the lack of sensitivity andthe strong energy dependence of the instruments.

(ii) The second was the lack of standards and ref-erences to assist the interpretation of the res-ults.

For further developments of x-ray dosimetry theelaboration of an adequate standard dosimeter wasof major importance (Kustner, 1924) and led to thedevelopment of sophisticated and precise instru-mentation based on ionization of air. The way todose quantification and understanding of biologicaland technical aspects of x-ray diagnosis was opened.Patient and staff dosimetry reflected the impact oftheories on biological effects at the time, resulting inthe use of derived quantities. Skin dosimetry withthe entrance skin exposure, related to cosmetic anddeterministic damage, was used to discuss limits ofpatient exposure (Braun et al., 1928). For radiationdosimetry, in the time period between the late 1920sand the early 1950s the quantity exposure wasreplaced by the quantity of absorbed dose (Taylor,1990).In the 1950s interest was focused on the induction

of genetic effects and cancer, mostly leukaemia. Con-sequently, the anatomical sites included most oftenin patient dose studies were the skin, the gonads,and the active bone marrow (BM). The doses at theskin could be directly measured with ionizationchamber dosimeters. The doses to the gonads andthe active BM were taken to be representative ofthe totality of likely radiation effects. The absorbeddose to the male gonads was usually taken to be thesame as that measured directly to the closely adja-cent skin, whereas mean absorbed doses to thefemale gonads or the active BM were related toskin dose using the results of dose measurementsin physical phantoms. A large number of dosimeterswere required to obtain even approximate estimatesof the mean dose to widely distributed tissues suchas the active BM and many different exposure con-ditions were needed to simulate the most commontypes of x-ray examination. Nationwide surveysusing these dosimetric techniques combined withstudies of the numbers of x-ray examinations tookplace in several countries in the period from 1955 upto the early 1970s. Attention was focused on provid-ing estimates of the genetically significant dose


(GSD) (UNSCEAR, 1972; NAS/NRC, 1980) and themean active BM dose to the population.Developments in solid-state dosimetry led to the

increasing use of small unobtrusive thermolumines-cent dosimeters (TLDs) directly attached to thepatient, for the measurement of the entrance surfacedose (ESD), including backscattered radiation, atthe centre of the x-ray beam. Alternatively, exposurewas measured with ionization chambers positionedfree-in-air on the x-ray beam axis. The results wereconverted to the exposure at the patient entrancesurface plane using the inverse-square law. The res-ults of measurements of radiation incident on theskin were variously reported as entrance exposurein rontgens (free-in-air) or as entrance absorbed dosein rads (ICRU, 1961). The material in which theabsorbed dose was measured was air, water, or softtissue. Measurements were made either free-in-airor on the surface of a patient or a phantom. There isapproximate numerical agreement (to within15%)between exposures expressed in rontgens andabsorbed dose to air, water, or soft tissue expressedin rads, for the x-ray spectra used in medicalimaging. However, doses measured with the patientor phantom present can be up to 60 % higher thanthose measured free-in-air at the skin-entranceplane because of the contribution of backscatteredradiation.In the 1960s a special type of large-area transmis-

sion ionization chamber dosimeter became availableto measure the radiation incident on the patient interms of the product of the exposure and the area ofthe x-ray beam in units of R cm2 (Arnal and Pychlau,1961; Morgan, 1961; Cameron, 1971). The chamberintercepted the entire useful beam, irrespective ofthe collimator setting, so that its response was pro-portional to both the area of the beam and the expos-ure. This was thought to provide a more nearlycomplete measure of the total exposure of the patientand hence to be more closely related to the radiationrisk, which depends on the extent of the irradiatedvolume within the patient as well as the exposure atthe centre of the x-ray beam. Moreover, there isconsiderable practical advantage in the flexibilityafforded in the positioning of the chamber owing tothe approximate invariance of the product of expos-ure and beam area for all planes perpendicular tothe beam axis between the beam defining collimat-ors and the patient. However, this device led to theintroduction of a quantity and unit (exposureareaproduct and R cm2), unfamiliar to the practitionersat that time, to add to those already used to expressthe dose to the patient in diagnostic radiology.It was soon appreciated that measurements of the

exposurearea product (or strictly the integral of theexposure over the area of the beam), together with

data on the x-ray spectrum, could be readily conver-ted into the total radiant energy in the beam. Henceit could be used to make estimates of the total energyimparted to the patient or the integral dose(Carlsson, 1963) as it was then known. The integraldose, defined as the mass integral of the absorbeddose over the total body (expressed in units of g rad,or erg), was seen to be more closely related to thelikely biological effects than the exposureareaproduct. It had the considerable advantage overorgan doses that it could be derived directly fromexposurearea product measurements. As long asother details of the exposure conditions were avail-able, it allowed a reliable estimate of the energy thatwas deposited in the patient. This fraction is critic-ally dependent on the x-ray beam quality, the sizeand position of the x-ray beam in the patient, thethickness of the part being x-rayed, and the propor-tion of the beam intercepted by the patient. Becausethe integral dose was often easier and more practicalto estimate than the doses to individual organs,many studies in the 1960s and 1970s reporteddoses to patients in this way.In the 1980s the names and definitions of the

quantities and units used in diagnostic radiology,as in other areas of radiation dosimetry, underwentmajor changes following the publication of revisedrecommendations by the ICRU (1980), which advoc-ated adoption of the International System of Units.As a result, exposure was replaced by air kerma (AK)(unit: joule per kilogram with the special namegray) as the quantity in which dosimeters were cal-ibrated and linked to the national primary stand-ards. Absorbed doses were also expressed in gray,thus improving the numerical agreement betweenthe quantity measured by dosimeters and theabsorbed dose to soft tissue (to within 5% for medicalx rays and most tissue-equivalent materials).Exposurearea product (R cm2) was replaced bykermaarea product (Gy cm2) or (absorbed)dosearea product (DAP) (also Gy cm2) and integraldoses were expressed as the total energy imparted tothe patient in joules.The limitations of the concept of the total energy

imparted to the body and its relation to possibleradiation effects became more apparent when com-putational methods of dosimetry became moreadvanced. They allowed the mean absorbed dosesto individual tissues and organs of relevance to radi-ation protection to be calculated. Using simple geo-metrical models of the body and of typical medicalx-ray beams, Monte Carlo radiation transport codesenabled researchers to calculate organ doses nor-malized to easily measured quantities such as theentrance skin exposure, the ESD, or the DAP for awide range of medical x-ray exposure conditions.


10

Better information on the relative radiosensitivitiesof different organs and tissues became available inthe late 1980s (UNSCEAR, 1988; BEIR, 1990; ICRP,1991b). This information combined with knowledgeof the mean absorbed doses to these organs andtissues led to estimates of the total radiation riskfrom the partial body exposures typical of medicalimaging.There is, however, a need for a single dosimetric

quantity related to the total potential health detri-ment to provide a practical tool for optimization ofprotection for the patient. Unless one particularorgan is completely dominant in determining thetotality of likely health effects, it is inconvenient toassess the risk from a list of organ doses. The dose tosome organs may be increased and to othersdecreased by the technique changes under consid-eration. For practical comparisons of radiation risksfor different techniques or procedures, some invest-igators have applied the concept of effective dose aweighted sum of organ doses developed by the ICRPprimarily for use in its system of radiological protec-tion (ICRP, 1991b) tomedical exposures.However,effective dose should not be used, for example byusing the nominal fatality coefficients (ICRP, 1991b),for assessment of detriment from exposure due tomedical x-ray imaging. Such assessments could beinappropriate because of potential differences inhealth status, gender, and age between a particulargroup of patients and the reference population for

whom the ICRP derived the risk coefficients. Fur-thermore, the conditions of a low dose and especiallya low dose rate, assumed by the ICRP, may arguablynot always be met in practice in radiology.

1.2 RISKS FOR THE PATIENT INRADIOLOGICAL IMAGING AND RELEVANTDOSIMETRIC QUANTITIES

The objective of dosimetry in radiological imagingis the quantification of radiation exposure within anapproach to optimize the image quality to absorbeddose ratio. The image quality should be understoodas the relevant information appropriate to theclinical situation. Dosimetry also provides themeans to avoid excessive doses that could imply asignificant risk of induction of deterministic effects,for example, for some cases in interventional radi-ology. The dosimetric quantities and dosimetric pro-tocols relevant in radiological imaging are thosemost closely related to the major (or more frequent)risks for the patient.Most of our knowledge on induced radiobiological

effects is based on the relation between absorbeddose and biological effect. For radiation protectionand therapy applications, the absorbed dose hassometimes to be weighted by appropriate factors topredict the effects or risks (Hall, 2000; Wambersieet al., 2002; Zoetelief et al., 2003a). Internationalrecommendations and national and internationalregulations are also based on absorbed dose orweighted absorbed dose (ICRU Report 60, 1998c;ICRP Publication 60, 1991b; IAEA, 1996a; EC,1997).

Figure 1.1. Acute dermatitis 3 weeks after excessive skinirradiation (several Gy) during an interventional procedure(placement of a porto-cave shunt). The lesion is well delineated(12 cm in diameter) and located at the back of the patient, in theupper right paralumbar area. One can notice the beginning ofre-epithelialization originating from the border of the lesion(arrows 1) but also within the irradiated area. Inside the irradi-ated area, each white spot is a clone resulting from the prolifera-tion of a surviving basal stem cell of the epidermis (only twoclones are shown by arrows 2) [P. Goffette, ImagingDepartment, UCL University Clinics St Luc, Brussels, cited inWambersie et al. (2005), by permission of Oxford UniversityPress].

Figure 1.2. Chronic dermatitis on the skin of the same patient asin Figure 1.1. One year after irradiation, sclerosis of the dermis,numerous telangiectasiae, total loss of elasticity, fixation to thedeep tissue layers, frequent ulcerations [P. Goffette, ImagingDepartment, UCL University Clinics St Luc, Brussels, cited inWambersie et al. (2005), by permission of Oxford UniversityPress].

INTRODUCTION

11

1.2.1 Acute deterministic effects1

In the early years of application of x rays in radi-ology, several cases of acute and severe skin reac-tions were reported. With the improvement of theequipment, the use of safer working procedures, andbetter training, the number and severity of theseaccidents decreased rapidly and even became excep-tional. In recent years, however, with the rapiddevelopment of interventional radiology, acute skinreactions became again a major concern for patientsafety (Figures 1.1 and 1.2).At high dose (>2 Gy), the severity of the lesions

can be predicted from the local absorbed dose in theskin and underlying dermis. Therefore, the skindose in the most heavily irradiated area is usuallythe most relevant quantity to be determined ininterventional radiology. This implies that both themaximum dose as well as the surface irradiatedabove the tolerance dose be evaluated (ICRP Pub-lication 59, 1991b). In interventional radiologicalprocedures, one of the main practical issues is oftenthe identification of the skin region that receives thehighest dose, as discussed in Section 3.

1.2.2 Late effects

1.2.2.1 Cancer induction

Cancer induction is generally considered to be themain risk for patients after radiological imaging.The quantitative approach of the risk assessment isa more complex issue than for acute effects, becausewe are dealing with (rather low) probabilities at lowdoses (UNSCEAR, 2000). The risk of cancer induc-tion in an organ or tissue is assumed to be related tothe average absorbed dose in that organ andstrongly depends on the type of organ or tissue. Inradiation protection, the relative risk of lethal can-cer induction is assumed to be the product of theorgan dose and a weighting factor (wT), whichexpresses the particular susceptibility of that organto cancer induction (Section 3.3; ICRP Publication60, 1991b).For the purpose of radiation protection, i.e., for the

management of radiation risk, the total risk of lethalcancer induction for a person is assumed to berelated to the sum of the weighted organ doses.This is known as the concept of effective dose, mainlydesigned for occupational exposure (ICRP Pub-lication 60, 1991b). The numerical values of theweighting factors were selected for an averageadult population. In radiology, more appropriate

weighting factors could probably be selected takinginto account the characteristics of individual patientsundergoing radiology such as gender and age (e.g.,the variation of susceptibility of the breast for breastcancer induction).Cancer induction is generally considered to be a

stochastic effect with the probability (and not theseverity) of effect increasing with dose. For cancerinduction, however, this traditional distinctionbetween deterministic and stochastic effects is likelyto have limitations. There is indeed evidence thatchronic radiation-induced lesions, such as sclerosisor chronic dermatitis, are particularly dangerouspre-cancerous lesions, although they are clearlydeterministic effects (Boice et al., 1985, 1988; Araiet al., 1991; ICRP Publication 59, 1991a; Wambersieet al., 1995).In addition to the lethal cancers taken into

account in the weighting factors discussed above,the incidence of non-lethal cancers is of importance.

1.2.2.2 Late effects in normal tissues

An increasing number of data indicates that somelate effects in normal tissues, such as the heart,represent a risk equally important as the risk ofcancer induction, for doses larger than several hun-dred mSv (Brenner et al., 2003; Preston et al., 2003;Tubiana et al., 2005).

1.2.2.3 Impairment of mental development

Evidence has been reported that doses such asthose delivered by repeated computed tomography(CT) examinations performed in sub-optimal condi-tions in children may significantly impair mentaldevelopment (Hall, 2002; Hall et al., 2004; Yamadaet al., 2004).

1.2.2.4 Genetic risk

For some radiological procedures involving thegonads, the genetic risk has to be considered, espe-cially for younger patients. A careful analysis, basedon a critical review of the available data, however,concluded that the previous evaluation of the riskwas significantly overestimated. It has now becomepossible to estimate risks for all classes of geneticdiseases (which was not the case until 1993) and thatthe risks are small compared with the baselinerisks (Sankaranarayanan and Chakraborty, 2000;UNSCEAR, 2001).

1.2.3 Relevant dosimetric quantities anddosimetric procedures

On the basis of the discussion of the mainradiation-induced effects and risks, the following

1The ICRP has recently proposed (2005) to replace theterm deterministic effects by tissue reactions.


12

relevant dosimetric quantities and dosimetric pro-cedures can be derived.Absorbed dose, expressed in gray (Gy), is the rel-

evant quantity to be determined and reported inradiological imaging. In practice, it should bedetermined in water (or water equivalent) and simu-lating or using patient geometry, i.e., taking intoaccount scattered radiation and shape of patientcontour.Absorbed dose should be evaluated at the level of:

Skin, because the skin is the most heavilyexposed tissue. Reports on skin dose shouldinclude the site location, maximum dose, andskin surface irradiated above the tolerance levelthat should be specified;

The most heavily irradiated organs, taking intoaccount the specific susceptibility of the organs forcancer induction and late effects as discussedabove.

Depending on the radiological procedure that isperformed, dose should be evaluated for the follow-ing organs: female breast, heart, thyroid, gonads(depending on age), brain (especially in children),and also dose to the embryo. The maximum doseand the average organ dose should be evaluated.The susceptibility of the female breast for cancerinduction strongly varies with age. The averagedose to the glandular tissue is probably more signi-ficant than the average dose to the organ as a whole.The use of CT raises specific problems that will be

discussed in Section 3.2.

1.2.4 Required accuracy

The required accuracy for dose determinationdepends on the dose level and potential risk. Anaccuracy requirement of 7 %, at an expanded uncer-tainty using a coverage factor of 2 corresponding tothe confidence level of 95 %, is recommended forcomparative risk assessments as well as for qualityassurance (AAPM, 1992). When there is a risk ofdeterministic effects similar accuracy will be needed.For calibration the expanded standard uncertaintyshould not exceed 5 % (AAPM, 1992). These require-ments can, in general, be achieved.An accuracy of 3050 % can be accepted in the

cases where organ doses are low. For comparison,in occupational exposure or radiation protection, anaccuracy of 50 % may also often be accepted in somesituations. In contrast, in radiation therapy, anaccuracy on dose, at the reference point, better than5 % is aimed for in the case of radical treatment.In the case of an accidental acute effect (e.g., skin

burn), sufficient information should be available tothe staff of a department of radiology to allow dose

reconstruction with a reasonable accuracy to predictthe severity of the possible harm.

1.3 DOSIMETRY IN RADIOLOGY: RELEVANTQUANTITIES

1.3.1 Calibration at the StandardsLaboratory

In diagnostic radiology, dosimetry is based oncalibrated instrument(s), usually ionization cham-ber(s). These instruments (used as local referencein the department) are calibrated at NationalStandards Laboratories, or at laboratories withinstruments directly traceable to National Stand-ards Laboratories. The calibration coefficients in thecertificates are expressed in terms of air kerma free-in-air. The manufacturers also express their calibra-tions in these terms. Any change in the situation isunlikely in the foreseeable future.

1.3.2 From air kerma free-in-air toabsorbed dose in water in patient orphantom

Conversion coefficients are used to obtainabsorbed doses in organs or tissues and at selectedclinically relevant points (Figure 1.3). These conver-sion coefficients depend on a number of factors thatare discussed in detail in the present report. Theyare based in part on ICRU Report 57 (ICRU, 1998a).

1.3.3 Air kerma-area product (KAP) anddosearea product (DAP)

The output of an x-ray tube can be monitored witha transmission ionization chamber. The chamber ismounted on radiological tubes, downstream relativeto the collimator (Figure 1.3). The increasing use ofthis monitoring device is partly a consequence of theEC Directive (EC, 1997).The signal of the chamber is calibrated in terms of

KAP (or absorbed DAP), at a specified distance free-in-air. It is expressed in mGym2. As more convenientmultiples, the radiology community would preferGy cm2. The transmission chamber acts as amonitor, like in radiation therapy, and should becalibrated against the reference dosimeter of thedepartment.From the radiological parameters (Section 1.3.4.1),

the readings of the reference chamber, and the KAP(or DAP), it is possible to derive useful information:

Doses at reference points, for example, entrance(skin) dose on the beam axis at a given distance,and at specified points in organs.

Mean organ doses.

INTRODUCTION

13

This, however, requires sophisticatedMonte Carloprograms, which produce conversion coefficientsvalid only for well-defined procedures and withinstrict (geometric) limits that should be specified(Chapter 5; Appendices B, C, D, and F; Struelens,thesis, 2005).As part of a quality assurance program, it is

important to check regularly the reproducibility ofthe response of the transmission chamber with acalibrated dosimeter.A normalization quantity is a dosimetric quantity

that can be readily measured or calculated in theclinical situation. Normalization quantities are usedto derive a specified dosimetric quantity, for example,mean organ dose, using an appropriate conversioncoefficient. To date KAP orDAP, (Section 2), are com-monly used as normalization quantities, but variousother quantities can also be used (Section 3.4).

1.3.4 Reporting patient irradiation inradiological imaging

1.3.4.1 Radiological parameters of the exposure

Radiological parameters, such as tube voltage(kV), tube currentexposure time product (mAs),filtration, exposure time (for fluoroscopy), field size

and position, and irradiated region, should be recor-ded for the standard medical imaging procedures. Inrecent equipment, in particular in direct digital radi-ology systems (flat panel detectors), recording ofthese parameters ismade automatically (Figure 1.4).To the extent that the tube output remains con-

stant relative to these parameters, this set ofinformation allows evaluating doses at referencepoints where previous calibration has been per-formed at the surface or inside a phantom. Informa-tion obtained on the phantom can then be used toevaluate the dose to the patients, but only for routineprocedures and with the necessary care.The ICRU recommends that the radiographic con-

ditions of the irradiations be reported as completelyas possible. When these are carefully reported, it willbe possible later on to apply better conversion coef-ficients when available and to derive more relevantquantities.

1.3.4.2 Air kermaarea product (KAP) ordosearea product (DAP)

KAP or DAP provides a continuous monitoring ofthe x-ray tube output and an indication of theabsorbed dose at reference points at the skin andthe possibility to calculate organ doses for standard

Figure 1.3. Calibration of dosimeters to be used in a diagnostic radiology. (1) The reference dosimeter used in diagnostic radiology iscalibrated (directly or through transfer dosimeters) at a Standards Laboratory in air kerma free-in-air (left). (2) In diagnostic radiology,the dosimeter can be used to measure the entrance surface dose (skin dose) provided appropriate conversion coefficients are applied, i.e.,from air kerma free-in-air to absorbed dose in water including backscatter (right). In addition to skin dose, the chamber can also be used todetermine the dose at any point in the patient. (3) The transmission chamber, regularly calibrated (Section 4.2.4), is used as a monitor. Itprovides the DAP or the air kermaarea product (KAP), depending on how it is calibrated. From the radiological parameters (Section1.3.4.1), the readings of the reference and KAP dosimeters, in principle, organ doses can be computed for any organ using complex MonteCarlo programs. The resulting conversion coefficients are valid within strict limits and specific for each radiological procedure (fromWambersie et al., 2006, by permission of Oxford University Press).


14

procedures. KAP or DAP can be used as normaliza-tion quantities to derive relevant dosimetricquantities, such as organ doses, using conversioncoefficients (Sections 1.3.3 and 3.4).Continuous monitoring of KAP relative to the

radiographic parameters of exposure listed previ-ously (Section 1.3.4.1) improves the confidence inboth approaches.

1.3.4.3 Monte Carlo computation

Sophisticated Monte Carlo programs have beendesigned to calculate doses at specified points oraverage organ doses in phantoms starting from theradiographic parameters and/or the KAP or DAP(Figure 1.5). The results from these approaches areavailable mainly for standard, well-establishedradiological procedures. The use of such MonteCarlo codes needs considerable expertise whenapplied for specific or uncommon procedures.

1.3.4.4 Phantoms and in vivo measurements

Different types of phantoms have been designed(ICRU Report 44, 1989b; ICRU Report 48, 1992b).Transfer of doses derived from measurements using

phantoms to doses for individual patients or patientseries requires appropriate care, but some programsallow the introduction of individual patient para-meters (gender, age, shape, dimensions, organ sizeand location, etc.) (Tapiovaara et al., 1997).Patient measurements may be considered as a

validation (in clinical conditions) of phantom meas-urements to derive the doses. However, in radiologysuch measurements are often not a realistic scenariofor every type of procedure on a systematic base.They are, however, strongly recommended on a lim-ited series of patients for standard procedures.

1.3.5 Discussion

In radiological imaging, one of the dosimetricissues is the fact that several quantities are used toquantify the magnitude of the exposure of thepatient to ionizing radiation. There is some ambigu-ity (and even disagreement) in their names andapplications in radiological procedures (e.g., kermaversus absorbed dose, air versus water, free-in-air orwith backscatter).The numerical values of the different quantities

may be close to each other and within dosimetricuncertainties (even within clinical accuracy

Figure 1.4. Correlation between radiological parameters, transmission monitor readings [air kermaarea product (KAP) or dose areaproduct (DAP), and clinically relevant dosimetric quantities (see arrows 3 and 1, respectively). This correlation should be checked atregular intervals depending on the type of radiological examination, the irradiated organs and level of exposures, and the type of patients(pregnant women, children, etc.]. The radiological parameters include tube voltage (kV), tube currentexposure time product (mAs),filtration, field size (at a given distance), and exposure time for fluoroscopy (Section 1.3.4.1). The clinically relevant dosimetric quantitiesinclude the skin (entrance) dose on the beam axis, skin surface (and regions) irradiated above the tolerance dose for induction ofdeterministic effects at clinically relevant points and/or organs. Left part Regular checks of the correlation between the radiologicalparameters and the readings of the KAP (or DAP) dosimeter (arrows 2) guarantee output stability and the reliability of both dosimetricapproaches. It improves the confidence in the two sets of dosimetric results. Right part In the absence of a KAP (or DAP) meter, regularchecks of the correlation (arrow 3) between the radiological parameters and the readings of the reference ionization chamber, at referencepoints, may also provide some guarantee on the stability and reliability of the dosimetry. This latter procedure, however, should be limitedto simple, routine, and well-defined radiological examinations and is not recommended for examinations of pregnant women and children.Courtesy: A. Wambersie.

INTRODUCTION

15

requirements). This cannot be used as an excuse fornot specifying clearly the involved quantities.There is still debate concerning the relevant

quantity to select in radiology: (i) the standardslaboratories (and manufacturers) calibrate the ion-ization chambers in air kerma free-in-air; (ii) theradiological community is using the concept ofabsorbed dose; (iii) a device informing the practi-tioner of the quantity of radiation produced by theequipment during the radiological procedure (inpractice a transmission chamber) is now imposedby regulation in several countries (EC, 1997).This situation is very similar to the situation in

radiation therapy. In each radiation therapy depart-ment, there is a reference dosimeter (ionizationchamber) calibrated for 60Co gamma rays or a fewMV x rays. Conversion coefficients are available forphoton and electron beams of higher energies (ICRUReport 64, 2001; ICRU Report 71, 2004). A transmis-sion chamber used as a monitor is regularly calib-rated against the reference dosimeter in specifiedconditions.

It is the responsibility of the local medical physicsteam to select and apply the appropriate con-version coefficients and to check the reproducibilityof the responses of the different dosimetricsystems. Similar working procedures and sharingof responsibilities can be applied in a radiologydepartment.The medical and radiobiological community, radi-

ologists, and radiation oncologists currently relatethe biological effects to the absorbed dose. A largeexperience has been built using absorbed dose as aquantification of the magnitude of exposure anddose-effect relationships. It would not be safe orwise to modify this well-established approach.When using ionization chambers calibrated in AK,it is thus important to select the appropriate conver-sion coefficients. In that respect, the ICRU Report 74where this issue is carefully discussed is thus timely.In the European and national regulations and

recommendations, the quantities absorbed dose andweighted absorbed dose (equivalent dose, effectivedose, etc.) are used. This may interfere with the

Figure 1.5. Visual representation of the photon tracks (primary and scattered radiation) and energy distribution (colour code: light gray,2040 kV; medium gray, 4060 kV; dark gray, 6080 kV) of a Monte Carlo simulation for a PA view of a pelvic exposure, as part of anangiographic examination of the lower limbs. The exposure was assumed to be performed at a peak tube voltage of 80 kV, a total filtrationof 4.6 mm Aleq, and a focus-to-skin distance of 55 cm. As model for the patient, the mathematical phantom BODYBUILDER was used.Conversion coefficients can be retrieved between different calculated values, such as a specific organ dose or DAP/KAP. Reproduced withpermission from Dr. L. Struelens.


16

selection of the quantities in the radiological depart-ments (EC, 1997; ICRP Publication 73, 1996).Dosimetry for CT raises specific issues. For CT,

skin dose and dose at a point do not have the samesignificance as for conventional radiography, andspecific indexes thus need to be introduced. The CTdose index (CTDI) has been defined for the assess-ment of organ dose and QA (Sections 3.2 and 4.2.5).Several methods were published for the practicalassessment of CTDI (IEC, 1999; EC 2000).For interventional radiology, where the avoidance

of deterministic effects becomes important, dose-conversion coefficients relating the absorbeddose to the more heavily irradiated site at thesurface of the body to normalization quantities(Section 3.4) are not yet commonly available, exceptfor some coronary procedures. In particular, thepoint (place on skin) where the maximum dose isobtained may not always be known in advance(Section 3.3.1).

1.4 NEED FOR HARMONIZATION OFQUANTITIES AND TERMINOLOGY

Various quantities and terminologies have beenused for the specification of dose on the centralbeam axis at the point where the x-ray beam entersthe patient (or a phantom representing the patient).These include the exposure at skin entrance (ESE),the input radiation exposure, the entrance surfaceair kerma (ESAK), the entrance air kerma, the AK,the ESD, the entrance skin dose (ESD), and theintegral skin dose (ISH and EC, 1998).ICRU Reports 33 and 60 (1980, 1998c) recommen-

ded the use of the International System of Units (SI).In ICRU Report 33 it is stated that the special unitof exposure, the rontgen, should be dropped by1985 and be replaced by the SI unit C/kg. Con-sequently, the approximate numerical equivalenceof exposure, AK, and absorbed dose to air was lost.As a result of the introduction of the SI, exposure hasbeen replaced by air kerma (free-in-air) as thequantity in which dosimeters are calibrated atstandards laboratories.Sometimes it seemed to be the convention that

kerma implies absence of backscatter and absorbeddose implies the presence of backscatter. However,the presence or absence of backscatter cannot bederived from the definitions of kerma and absorbeddose (ICRU, 1998c). Calibrations of dosimeters aregenerally made in terms of air kerma free-in-air.Therefore, it is often assumed that kerma isexpressed only in air. However, the ICRU (1998c)explicitly states that one can refer to a value ofkerma for a specified material at a point in freespace, or inside a different material. Different

names are used in practice for the same quantity,for example, entrance surface air kerma, air kerma,and entrance air kerma. The same abbreviation ESDis used for both entrance surface dose (absorbed dosedetermined free-in-air most likely expressed in air)and entrance skin dose (absorbed dose most likelyexpressed in skin tissue).The kermaarea product is often used for dose

assessment for more complex examinations, includ-ing radiography and fluoroscopy. Although the namedoes not state this explicitly, the kerma is usuallyexpressed in air free-in-air and the backscatter froma patient or a phantom is not to be included.In interventional radiology, the value of the

kermaarea product for a complete examinationhas been used as an indicator for the occurrence ofstochastic effects, whereas information on the max-imum dose at skin entrance (that is, the dose at thelocation of the skin where it is highest) is of import-ance with respect to the possible occurrence of deter-ministic effects. The maximum dose at skin entranceis also referred to as the peak skin dose (Miller et al.,2002) and defined as the highest dose delivered toany portion of the patients skin.For the assessment of organ doses and quality

assurance in CT, the CTDI has been defined as theintegral of the absorbed dose profile along a lineparallel to the axis of rotation of the scanner dividedby the nominal slice thickness (Shope et al., 1981). Inthe literature, different methods can be found for thepractical assessment of CTDI. These include differ-ences in the boundaries of integration, the use of aphantom or measurement free-in-air, and differ-ences in the material in which the absorbed dose isexpressed, for example, polymethylmethacrylate(PMMA) or air.The present situation in patient dosimetry for

medical x-ray imaging clearly indicates the needfor dose quantities recommended for the differentapplications and the need for using the same, self-consistent, system for names, symbols, and units.

1.5 THE TWO PURPOSES OFPATIENT DOSIMETRY

In medical x-ray imaging there are two funda-mental reasons for measuring or estimating thepatient dose. First, measurements provide a meansfor setting and checking standards of good practice,as an aid to the optimization of the radiation protec-tion of the patient and of image quality. Second,estimates of the absorbed dose to tissues and organsin the patient are needed to assess radiationdetriment so that radiological techniques can bejustified and cases of accidental overexposureinvestigated.

INTRODUCTION

17

1.5.1 To set and check standards ofgood practice

It is well known from the results of numeroussurveys that there is considerable variation in thedoses delivered to patients from the same type of x-ray examination conducted in different facilities oreven within a single facility. Without some form ofpatient dose monitoring, it is difficult to know theperformance of an x-ray facility and to judge how itcompares with generally accepted practice. As an aidto the optimization of the radiation protection of thepatient, reference values, variously called referencedose values (EC, 1996c), DRLs (ICRP, 1996), or guid-ance levels (IAEA, 1996a), can be specified for par-ticular x-ray imaging tasks. Local performance canbe checked against these reference values by peri-odic measurement as part of a quality assuranceprogram. For these purposes, clearly defined quant-ities are required, which can be easily measuredwith readily available instruments of sufficient pre-cision and accuracy. Consequently, dosimetricquantities associated with the primary x-ray beam(e.g., AK at a specified point on the beam axis) orclosely related radiation quantities suited to particu-lar applications (e.g., KAP, CT kerma index or CTDI)have become established quantities used to set andcheck standards of good practice.

1.5.2 To assist in assessing detriment orharm

For the justification of a practice or for the invest-igation of cases of accidental over-exposure, theabsorbed dose to the patient should be expressed inamanner that can be directly related to the potentialhealth risks. For stochastic radiation effects, theICRP (1991b) has recommended that the meanabsorbed dose to individual organs or tissues is thequantity of interest. If deterministic effects are con-sidered possible, for example, in some cases of inter-ventional radiology, the absorbed dose to the moreheavily irradiated sites at the surface of the body,such as the skin in the primary beam, is the radi-ation quantity of interest.Organ or tissue doses cannot be directly measured

in patients but can be derived from other directlymeasurable radiation quantities using appropriateconversion coefficients. Extensive tabulations ofsuch conversion coefficients for reference patientsand reference irradiation conditions have been pub-lished, are available in computer readable form, orcan be calculated (Appendices BF). These coeffi-cients relate organ doses to practically measurablenormalization quantities, some of which are thesame as the quantities used for setting and checkingstandards of good practice. Alternatively, organ and

tissue doses can be derived from measurementsinside physical phantoms.

1.6 RELATIONSHIP BETWEEN PATIENTDOSE AND IMAGE QUALITY

It is important to ensure that efforts to reducepatient doses do not also reduce doses to the imagereceptor to such an extent that the quality of theimages is degraded to an unacceptable level. Imagequality can be affected by inadequate doses in fourdistinct ways:

(i) In the non-digital imaging systems used in con-ventional radiography and fluoroscopy, opticaldensity or brightness of the image depends onthe dose and dose rate received by the imagereceptor, respectively. Too low a dose or doserate can result in images that are too faint to beclearly discerned.

(ii) Dose reduction by increasing the tube voltageand thereby allowing a reduction in tube cur-rent or exposure time to maintain the samedose to the image receptor can degrade imagequality by decreasing contrast.

(iii) As medical imaging systems have become moresensitive, needing only lower doses to achieveimages of satisfactory density or brightness,there is an increased likelihood that randomvariations in the photon fluence rate reachingthe image receptor will give a disturbingmottled appearance to the image. This so-called quantum mottle is preferably the dom-inant source of image degradation in sensitivedigital and non-digital imaging systems.

(iv) The sensitivity of the imaging system can oftenbe improved by increasing the thickness of thesensitive layer of the image receptor so that itabsorbs more of the incident x-ray energy. Forthe majority of image receptors that re-emit theabsorbed energy in the form of visible light,thicker sensitive layers result in wider spatialdispersion of the emitted light before the imageis recorded. Greater sensitivity, and hence thepossibility of using lower doses, is consequentlygained at the expense of poorer spatial resolu-tion in the image.

Procedures for checking that doses have not beenreduced to such an extent that inadequate opticaldensity, excessive noise, poor spatial resolution, orlack of contrast prevents reliable diagnosis should bean essential component of x-ray department qualityassurance programs. A range of phantoms that canbe used to assess image quality in diagnostic radi-ology are described in ICRU Report 48 (1992c). More


18

fundamental methods for assessing the quality ofmedical images are published by Doi et al. (1982;1986) and presented in ICRU Reports 41 and54 (1986, 1995). In addition, an ICRU report hasbeen recently published on Chest radiography-Assessment of Image Quality (ICRU Report 70,2003) and one is underway in Mammography Assessment of Image Quality.

1.7 SCOPE OF THE REPORT

The present Report is primarily concerned withmethods of patient dosimetry in medical x-rayimaging. Methods are described for dosimetry per-formed to set or check standards of good practice andas a step towards the assessment of detriment orharm. In many situations the dosimetric quantityof interest can be directly measured. In other cases,however, it is not practicable to measure a quantityof interest directly (e.g., organ or tissue dose).Instead it is obtained indirectly by application of anappropriate conversion coefficient to a quantity thatcan be measured directly.All aspects of patient dosimetry require know-

ledge of the properties of the x-ray beam. The basicfeatures of the x-ray beams used for medical imagingand their specification and measurement are, there-fore, discussed in Section 2. The various dosimetricquantities to be used for patient dosimetry aredefined and discussed in Section 3, and appropriatenotations are introduced. The discussion is dividedinto five parts: basic dosimetric quantities,

application specific quantities, risk related quantit-ies, dose-conversion coefficients, and quantitiesrecommended for establishment and use of DRLs.Section 4 discusses the methodology to be used forthe measurement of specific dose quantities includ-ing incident AK, ESAK, KAP and various quantitiesto be used for CT dosimetry. It considers the choiceand calibration of dosimeters, the practical measure-ment technique, the measurement uncertainty, andthe advantages and disadvantages of phantom-based and patient measurements. The derivation ofbackscatter factors, which can be used to relateincident AK and ESAK to ESD, is treated inAppendix A. The conversion coefficients used toestimate organ and tissue doses are covered in Sec-tion 5. Two approaches to determine the conversioncoefficients are considered: measurements in phys-ical phantoms and computational methods. The lat-ter useMonte Carlo techniques to simulate radiationtransport through computer-based models of thepatient. This is usually the more useful approach.The main features of the computational method andthe associated uncertainties in the conversion coef-ficients are discussed. Results of such calculationsare available from a number of sources for conven-tional projection imaging, CT, and mammography.Conclusions, including recommendations on theselection of the most appropriate conversion coeffi-cient for a particular examination or procedure, arepresented in Section 6. Organ dose-conversion coef-ficients from different sources are considered inAppendices BF.

INTRODUCTION

19


2 SPECIFICATION OF X-RAY BEAMS

X-ray tubes used in medical x-ray imaging employpeak voltages of between 25 and 150 kV to acceler-ate electrons from a heated cathode towards ananode target. X rays are produced either as a resultof radiative energy loss by electrons slowing down inthe target material (bremsstrahlung) or by radiativetransitions of electrons within the atoms of thetarget (characteristic x rays). The former have acontinuous range of energies from that of the mostenergetic electron downwards whereas the latteroccur at discrete energies characteristic of the targetmaterial.The radiation quality of an x-ray beam can be

characterized by the x-ray spectrum. X-ray spectracan be measured by using spectrometers based onscintillation counters, germanium or silicon detect-ors, or by crystal diffraction. These techniques,however, require considerable expertise and aretime-consuming to perform. Therefore, it is recom-mended that the radiation quality of x-ray beamsused for medical imaging be characterized by a com-bination of various parameters. These include firsthalf-value layer (HVL; symbol HVL1); the secondHVL (HVL2); the ratio of HVL1 to HVL2, referred toas the homogeneity coefficient; the tube voltage andthe total filtration. In most cases, the quality of anx-ray beam can be adequately specified by means ofthe combined information on tube voltage, HVL1,and HVL2, or the tube voltage, HVL1, and totalfiltration.The yield is also an important characteristic of an

x-ray tube (including filtration). For this purpose thex-ray tube output is defined in Section 2.5.

2.1 PHOTON SPECTRUM

The shape of the spectrum depends on the voltageapplied to the x-ray tube (kV), the waveform ofthe generator, the target material and angle, andthe amount of inherent and added filtration in thex-ray beam. For most diagnostic examinations, andall interventional procedures, x-ray spectra from atungsten target are used (Figure 2.1). Aluminiumfiltration is generally employed to remove thelow-energy end of the bremsstrahlung spectrum,

which would otherwise be absorbed in the superficialtissues of the patient without contributing to thefinal image. An additional filter of, for example, cop-per may also be used in some situations to furtherharden the spectrum. To prevent the low-energycharacteristic x rays produced in the copper filterfrom reaching the patient the aluminium filtershould be between the copper filter and the patient.There are peaks in the tungsten spectra due to Kaand Kb characteristic radiation for peak tube vol-tages >69.5 kV, corresponding to the energy of the

0 20 40 60 80PHOTON ENERGY; E /keV

REL

ATI

VEN

UMB

ERO

FPH

OTO

NS

0 20 40 60 80PHOTON ENERGY; E /keV

REL

ATI

VEN

UMB

ERO

FPH

OTO

NS

Figure 2.1. Typical photon fluence x-ray spectra used for medicalimaging (IPEM, 1997). Top tube voltage: 80 kV constant poten-tial, tungsten anode; filtration: 3 mm Al, anode angle 16; HVL1:2.98 mm Al. Bottom tube voltage: 28 kV constant potential,molybdenum anode; filtration: 0.03 mm Mo, anode angle 12,b

Documents

Icru Report 74