PROPER COLLIMATION EF FECT ON RA DI ATION DOSE AND IM …

PROPER COLLIMATION EF FECT ON RA DI A TION DOSEAND IM AGE QUAL ITY IN THO RACIC SPINE RA DI OG RA PHY

by

Anamaria PAZANIN 1, Damijan SKRK 2, Nika ZALOKAR 1, and Nejc MEKIS1*

1 Med i cal Im ag ing and Ra dio ther apy De part ment, Fac ulty of Health Sci ences,Uni ver sity of Ljubljana, Ljubljana, Slovenia

2 Slovenian Ra di a tion Pro tec tion Ad min is tra tion, Ljubljana, Slovenia

Sci en tific pa perhttps://doi.org/10.2298/NTRP2002181P

The pur pose of this re search was to de ter mine the im pact of collimation in tho racic spine ra di -og ra phy on pa tient ex po sure and im age qual ity. The study was per formed on 84 pa tients re -ferred to tho racic spine ra di og ra phy. Pa tients were ran domly di vided into two equal groups of 42. The first group was im aged ac cord ing to the stan dard collimation pro to col used in one ofthe hos pi tals in Croatia while the sec ond group was im aged by ap ply ing “op ti mal”collimation, im age field size was in di vid u ally collimated for each pa tient or ac cord ing to thegreat est im age field collimation de picted in pro fes sional lit er a ture. For each pa tient bodymass in dex, im age field size, ex po sure con di tions and dose area prod uct were noted and ab -sorbed doses by or gans were cal cu lated, im age qual ity was as sessed. There were no sta tis ti -cally sig nif i cant dif fer ences in BMI be tween the two groups of pa tients. With the op ti malcollimation the size of the im ag ing field in the anteroposterior pro jec tion was re duced by45 % ( p < 0.001) and in the lat eral pro jec tion by 41 % (p < 0.001). The study also showed re -duced val ues of DAP for anteroposterior pro jec tion by 34 % ( p = 0.007) and for lat eral pro -jec tion by 23 % ( p = 0.040). The mean ab sorbed dose to the se lected or gans de creased by26 % in the anteroposterior pro jec tion and by 28 % in the lat eral pro jec tion. In ad di tion, theop ti mal collimation pro to col im proved im age qual ity by 13 % in anteroposterior pro jec tion.No dif fer ences in im age qual ity were found in lat eral pro jec tion. By car ry ing out this re searchwe have dem on strated that op ti mal collimation in tho racic spine im ag ing has a strong in flu -ence on pa tient ex po sure to ra di a tion and has a pos i tive im pact on im age qual ity.

Key words: tho racic spine ra di og ra phy, collimation, dose re duc tion, im age qual ity

IN TRO DUC TION

The Eu ro pean Com mis sion pub li ca tion [1, 2] onthe ex po sure of Eu ro pean pop u la tion to ra dio log i cal pro -ce dures re ports that gen eral ra di og ra phy ex am i na tionsare the most com mon in all coun tries (76.8 to 97 %), fol -lowed by com puted to mog ra phy (CT) (0.7 up to 16.7 %), flu o ros copy (0.9 to 13.9 %) and interventional ra di ol ogy(from 0.03 to 2.7 %). The most com mon pro ce dures ingen eral ra di og ra phy are those of the chest/tho rax, cer vi -cal spine, tho racic spine, lum bar spine, mam mog ra phy,ab do men and pel vis and hip [1, 3, 4].

In gen eral ra di og ra phy, tho racic spine im ag ing isone of the seven pro ce dures with the high est ef fec tivedose. The high est ef fec tive dose in plain ra di og ra phy isob tained dur ing lum bar spine ra di og ra phy (0.898mSv), fol lowed by pel vis and hip im ag ing (0.709 mSv)and tho racic spine im ag ing (0.636 mSv) [1].

Since the most sen si tive or gans with the high esttis sue weight ing fac tor (0.12), such as breasts, lungs,

co lon, stom ach and bone mar row [5], are or may be af -fected by the pri mary beam dur ing tho racic spinera di og ra phy, it is nec es sary to en sure an ad e quate im -age with the low est pos si ble ex po sure of the pa tient.To put it briefly, it is nec es sary to ap ply the as low aslow est rea son ably achiev able (ALARA) prin ci ple [5].

A collimator is a de vice that lim its the ra di a tionout put of an X-ray ma chine to the nar row est pos si blesize based on the im ag ing area and re fer ral di ag no sis[5]. Collimation has di rect ef fect on the vol ume of thepa tient's body ex posed to the pri mary beam of ion iz ing ra di a tion since the beam is formed by proper po si tion -ing of collimator blades [6]. Since the vol ume of ir ra -di a tion of the pa tient is re duced by ap ply ing ac cu rateand tight collimation, less scat ter ra di a tion is pro duced which im proves im age qual ity (IQ). By re duc ing thesize of the pri mary beam, the dose re ceived by the pa -tient can also be re duced [7]. Thus, if poor collimationis pres ent, the ir ra di ated area is en larged re sult ing indose in crease and more scat ter, which neg a tively af -fects IQ [8].

A. Pazanin, et al.: Proper Collimation Ef fect on Ra di a tion Dose and Im age ...Nu clear Tech nol ogy & Ra di a tion Pro tec tion: Year 2020, Vol. 35, No. 2, pp. 181-188 181

* Cor re spond ing au thor; e-mail: [email protected]

There fore, know ing and ap ply ing the cen tral rayand collimation for each in di vid ual ex am i na tion is ofgreat im por tance [9]. In clin i cal set ting, as al readynoted, there is a lack of guide lines on proper use ofcollimation. Karami and Zabihzadeh [7] did a re search in the area of lum bar spine ra di og ra phy and found very poor collimation prac tice on a sam ple of 830 ra dio -graphs of lum bar spine in anteroposterior pro jec tion.Pre cisely, the im age field size was 1.26 times largerthan the op ti mal and con se quently, the most sen si tiveor gans (co lon, breast, go nads) were near or di rectlywithin the pri mary beam with out jus ti fi ca tion. Itcaused higher doses for pa tients, harm ful health ef -fects and lower IQ. No sim i lar re search in the re -viewed lit er a ture was found for tho racic spine.

This study aims to dis cover the im pact ofcollimation on ra di a tion dose (DAP and or gan ab -sorbed dose) and IQ in tho racic spine ra di og ra phy.

MA TE RI ALS AND METH ODS

This re search was a pro spec tive study us ing anex per i men tal ap proach. The study was con ducted intwo phases on 84 pa tients (70 fe male and 14 male) re -ferred to tho racic spine ra di og ra phy in two pro jec tions (anteroposterior – AP and lat eral – LAT), who wereran domly di vided into two groups of 42 pa tients each.The pa tient sam ple size was cal cu lated us ing G*power 3.1 anal y sis tool based on the pre lim i nary data of 20pa tients.

The study was con ducted due to the dis cov erythat the hos pi tal's stan dard collimation pro to col wasnot in ac cor dance with pro fes sional lit er a ture. Thestan dard collimation pro to col was not strictly de ter -mined and adapted in di vid u ally to each pa tient. Inmost cases the ALARA prin ci ple was not con sid eredwhen it came to collimation and the im age field sizequite of ten re mained the same as the size de ter minedby the X-ray sys tem. The pri mary beam was notcollimated prop erly, and the pa tients were be lieved tobe over-ir ra di ated dur ing pro jec tion ra di og ra phy oftho racic spine.

In the first phase, 42 pa tients were im aged in bothpro jec tions ac cord ing to the stan dard collimation pro -to col used in the hos pi tal while in the sec ond phase, 42pa tients were im aged with the so called op ti malcollimation pro to col men tioned in the pro fes sional lit -er a ture [6, 9]. Each pa tient was placed in su pine po si -tion on the ex am i na tion ta ble in the AP pro jec tion andon the left side in the LAT pro jec tion. In the AP pro jec -tion the ver ti cal part of the cen tral ray was po si tioned tothe midline of the body and the hor i zon tal ray in themid dle of the ster num. The field size was sup posed tobe collimated ap prox i mately to the width of the tho racic ver te brae (12 cm). In the LAT pro jec tion the ver ti calplane of the cen tral ray was po si tioned 6-8 cm an te ri -orly from the pos te rior (back) bor der of the pa tient andthe hor i zon tal plane at the same height as in the AP pro -jec tion. Ver ti cally, the pri mary beam was sup posed tobe collimated so that it in cluded the rib cage and hor i -zon tally to the pos te rior (skin) bor der of the back [6].

The max i mum size of the op ti mally collimated fieldwas 18 cm × 43 cm, if pos si ble [10]. An ex am ple of op -ti mal collimation in AP and LAT pro jec tion are pre -sented on a sam ple fig. 1 of an an thro po mor phic phan -tom.

All pa tients' heights and weights were mea suredand re corded in or der to cal cu late their body mass in -dex (BMI). The ex po sure and field sizes for both pro -jec tions and both collimation pro to cols are pre sentedin tabs. 1 and 2.

This re search was ap proved by the Hos pi tal'sMed i cal Eth ics Com mit tee. All the par tic i pants werein formed about the pur pose of the study and havegiven their writ ten con sents.

Equip ment

The mea sure ments were per formed at the Ra di ol -ogy De part ment of one of the Cro atian hos pi tals that usethe CR Fujifilm im ag ing sys tem (FCR Clear View CS;soft ware ver sion V3.6) on a Siemens Multix/Vertix de -

A. Pazanin, et al.: Proper Collimation Ef fect on Ra di a tion Dose and Im age ...182 Nu clear Tech nol ogy & Ra di a tion Pro tec tion: Year 2020, Vol. 35, No. 2, pp. 181-188

Fig ure 1. Ex am ple of op ti mal collimated im ag ing fieldfor AP and LAT pro jec tion of tho racic spine

vice (tube type: OPTI 150/30/50HC) (fig. 1). Prior to and dur ing the study Qual ity Con trol (QC) test ing was per -formed. The con ducted tests were the fol low ing: tubevolt age ac cu racy and reproducibility, half value layer,tube out put (mGy/mAs), lin ear ity (tube cur rent) and vari -a tion with tube volt age, au to matic ex po sure con trol(AEC) test ing, dose area prod uct (DAP) me ter test ingand im age re cep tor test ing (spa tial res o lu tion, con trastres o lu tion and dy namic range). The mea sured re sultswere ac cept able re gard ing the stan dards [11]. The size ofthe im age re ceiver was 35 cm × 43 cm. The grid ra tio was 12:1, with 40 lines per cm, fo cus-de tec tor dis tance was115 cm and the to tal beam fil tra tion was 2.5 mm Al.

Im age qual ity

The im ages were as sessed by two ra di ol o gistsand one ra di og ra pher with more than 4 years of ex pe ri -ence in a blind ran dom ized study. The as sess mentswere made on the same di ag nos tic mon i tor (EIZORadiForce GX340 21.3'') by ap ply ing ViewDEX[12-14] im age soft ware. Two folders were cre ated,one con tained 84 ra dio graphs in AP pro jec tion and theother 84 ra dio graphs in LAT pro jec tion.

All ra dio graphs were as sessed on a 3-point scaleac cord ing to the cri te ria listed in Eu ro pean guide lines.The rat ings on the scale were: score 1 – di ag nos ti callyin suf fi cient ra dio graph, score 2 – di ag nos ti cally goodra dio graph, and score 3 – di ag nos ti cally per fect ra dio -graph.

Ac cord ing to the rec om men da tions based on theguide lines in the doc u ment IQ and Dose Man age mentFor Dig i tal Ra di og ra phy [15], the cri te ria for an op ti -mal im age that ap ply for tho racic spine im ag ing are asfol lows.

The cri te ria for AP pro jec tion:– Com plete im ag ing of tho racic spine, in clud ing

Th1.– Vi su ally sharp im ag ing in a sin gle line of the up per

and lower-plate sur face in the cen tred beam area.– Vi su ally sharp im ag ing of the pedicle, spinous

pro cesses and costovertebral joints.

The cri te ria for LAT pro jec tion:– Com plete im ag ing of the tho racic spine from Th2

down to the thoracolumbar junc tion.– Vi su ally sharp im ag ing in a sin gle line of the up per

and lower-plate sur faces in the cen tred beam area. – Visu ali sa tion of the intervertebral spaces and

intervertebral joints in the cen tred beam area. – Vi su ally sharp im ag ing of the cor ti cal and

trabecular struc tures.The scores for each cri te rion were then de ter -

mined by a vot ing sys tem where the most com mongrade (mode) for the im age was set as the grade for that spe cific cri te rion. If eval u a tors gave dif fer ent scores(1, 2, and 3) for a spe cific cri te rion, then the score wasset as di ag nos ti cally good (score 2). Af ter that, the sum was cal cu lated for each im age and pre sented as the to -tal score for the im age.

Or gan ab sorbed dose cal cu la tions

Doses ab sorbed by se lected radiosensitive or -gans that lie close to or within the pri mary beam (theor gans are listed at the end of this chap ter) were cal cu -lated by us ing the Monte Carlo sim u la tion pro gramPCXMC 2.0 (STUK, Ra di a tion and Nu clear SafetyAu thor ity in Fin land) [16].

Pa tient height and weight, im age field size, ex -po sure pa ram e ters and DAP val ues were used for cal -cu la tions. DAP was mea sured by us ing a built-in DAPme ter (VacuDAP com pact; VacuTec, Ger many), cal i -brated prior to the study.

Cal cu la tions were based on the math e mat i calprob a bil ity of in ter ac tions be tween pho tons and pa -tient's body, such as pho to elec tric ef fect, co her entscat ter and in co her ent scat ter. Dur ing the sim u la tion,the max i mum en ergy of pho tons was set (100 keV),and the num ber of tracked pho ton par ti cles was1000000 [16]. The data on im age field size, ex po surecon di tions, weight and height were used for each ra -dio graph/pa tient in di vid u ally. For each pa tient first the


Ta ble 1. Ex po sure and collimation pro to cols for AP pro jec tion

AP pro jec tion Cur rent collimation pro to col Op ti mal collimation pro to col

Tube volt age range [kV] 77-96 75-81

Time cur rent prod uct range [mAs] 4.88-62.70 (avg. 25.34) 10.30-63.80 (avg. 30.08)

Source-to-im age re cep tor dis tance (SID) [cm] 115 115

Im age width [cm] 17.9-35.0 (avg. 24.78) 10.7-187.3 (avg. 13.71)

Im age height [cm] 34.4-43.0 (avg. 42.22) 35.0-43 (avg. 41.5)

Ta ble 2. The ex po sure and collimation pro to cols for the LAT pro jec tion

LAT pro jec tion Cur rent collimation pro to col Op ti mal collimation pro to col

Tube volt age range [kV] 81-90 81-87.5

Time cur rent prod uct range [mAs] 5.26-77.10 (avg. 71.39) 7.56-58.80 (avg. 21.36)

Source-to-im age re cep tor dis tance (SID) [cm] 115 115

Im age width [cm] 20.6-35.0 (avg. 29.29) 13.3-21.9 (avg. 17.43)

Im age height [cm] 43 37.9-43.0 (avg. 42.88)

weight and height were inputted in the PCXMC 2.0pro gram so that the phan tom used was the same size asthe pa tient. Af ter the pro gram had changed the size ofthe pa tient the im ag ing field size that was col lated foreach pa tient was set in the pro gram and then the po si -tion ing of the im ag ing field was done based on the pa -tient's ra dio graphic im ages. The or gan ab sorbed dosecal cu la tions were per formed for each pa tient ac cord -ing to Monte Carlo sim u la tion. The an ode an gle of12°, ex act tube volt age used (kV), fil tra tion and mea -sured DAP value were set in the pro gram for each pa -tient sep a rately.

The po si tion ing of the im ag ing field was done sothat the first and the last ribs were shown in the im ageas in the ex am ple shown in fig. 2.

The av er age or gan ab sorbed dose was cal cu lated for the fol low ing radiosensitive or gans ly ing in the vi -cin ity or within the pri mary beam: ac tive bone mar -row, ad re nals, gall blad der, heart, liver, lymph nodes,lungs, oe soph a gus, pan creas, stom ach, thy roid glandand thy mus. In the fe male pop u la tion, the breast dosewas also ob served. The av er age or gan dose is the av er -age cal cu lated from all the av er age dose cal cu lated foreach in di vid ual or gan.

Sta tis ti cal anal y sis

The mea sure ments were ana lysed with the IBMSPSS STA TIS TICS ver sion 25 (IBM cor po ra tion,USA). Shapiro – Wilk test was used to check the nor maldis tri bu tion of the sam ple. In the case of nor mally dis trib -uted data, T-test for in de pend ent sam ples was used forcom par ing the dif fer ences be tween data. Oth er wise,when the data were non-nor mally dis trib uted, anon-para met ric ver sion of T-test: Mann Whit ney U-testwas used. The re sults were pre sented in the form of ta bles and in graphic form with a boxplot chart. The sig nif i -cance of p < 0.05 was used for all the tests.

RE SULTS

For each pro jec tion, AP and LAT, a to tal of 84BMI, im age field sizes, DAP, av er age or gan doses and252 IQ as sess ments were col lected. The re sults of allthe listed val ues for AP pro jec tion are sum ma rized intab. 3 and for the LAT pro jec tion in tab. 4.

Ab sorbed doses by se lected or gans are pre sented in tab. 5 for the AP pro jec tion and tab. 6 for LAT pro -jec tion.

Or gans that re ceived the high est dose dur ing theAP pro jec tion of tho racic spine ra di og ra phy were thefol low ing: thy mus, heart, stom ach, lungs and liver;while dur ing the LAT pro jec tion, lungs and liver re -ceived the high est dose.

The re sults re gard ing the IQ are pre sented as av -er age for each cri te rion, figs. 3 and 4.

DIS CUS SION

The aim of the study was to op ti mize thecollimation pro to col for tho racic spine ra di og ra phy inAP and LAT pro jec tion in one of Cro atian hos pi talsand to eval u ate its im pact on pa tient dose (in clud ingDAP, ab sorbed dose to se lected radiosensitive or gans)and IQ.

Ac cord ing to the BMI re sults ob tained in the be -gin ning there were no sta tis ti cally sig nif i cant dif fer -ences be tween cur rent and op ti mal collimation, there -fore, all the other val ues could be com pared sincepa tient size would not af fect the re sults.

The re sults of op ti mal im age field collimationcom pared to the collimation in line with the stan dardpro to col used in the hos pi tal showed re duc tion of pri -mary beam size in AP pro jec tion by 474.7 cm² (45 %)and in LAT pro jec tion by 512.1 cm² (41 %). Pri marybeam size re duc tion caused DAP re duc tion, more spe -cif i cally in AP pro jec tion by 34 % and in LAT pro jec -tion by 23 %. This data is of great im por tance as in suf -fi cient collimation has been iden ti fied as the larg estand most com mon cause of un nec es sary pa tient dose


Fig ure 2. Ex am ple of im age field po si tion ing in PCXMC2.0 pro gram for AP and LAT pro jec tion of tho racic spine

load. In com par i son with DRL given in Eu ro peanCom mis sion re port [2] we found lower DAP val ues inour study. Be fore field size collimation op ti mi za tion,the DAP value was by 29 % and 53 % lower in AP andLAT pro jec tion of the tho racic spine than the DRLgiven in the re port, re spec tively, while af ter op ti mi za -tion, the DAP value was by 53 % and 64 % lower in theAP and LAT pro jec tion, re spec tively.

We did not find any sim i larly per formed re -search in the re viewed lit er a ture re gard ing the or ganab sorbed doses to se lected radiosensitive or gans dur -ing the ra di og ra phy of tho racic spine. In our study, theor gans that re ceived the high est dose in AP pro jec tionwere: the thy mus, heart, stom ach, lungs and liver andin LAT pro jec tion, the lungs and the liver. With the op -ti mal collimation, no tice able re duc tions in ab sorbeddoses were achieved dur ing AP pro jec tion for: lungs(64 %), stom ach (61 %), liver (37 %), ac tive bone mar -row (29 %) and lymph nodes (23 %); and in LAT pro -jec tion for: thy roid (46 %), heart (40 %), gall blad der(38 %), stom ach (37 %), liver (23 %), lymph nodes(22 %), lungs (21 %), thy mus (20 %), and pan creas(12 %).

We would like to em pha size the im por tance ofcollimation in the im ag ing of the tho racic spine (bothAP and LAT pro jec tions) re lat ing to the doses re ceivedby the breast. Namely, in the re search we found thatwith op ti mal collimation in each pro jec tion, both AP

and LAT, of tho racic spine, we would pro vide 89 %lower ab sorbed dose to the breast. This re sult is of greatvalue as it is a highly sen si tive or gan to ion iz ing ra di a -tion (weight ing fac tor – 0.12; ICRP 103) [5] and dur ingthis ex am i na tion it is not pos si ble to pro tect this areawith lead pro tec tion. From this we can con clude that op -ti mal use of collimation is a pow er ful tool that can beeas ily used to pre vent un jus ti fied pa tient ex po sure.

As an ad di tional ben e fit of proper collimationand con se quently the re mark able dose re duc tion, Rob -in son et al. [16] stated that it sig nif i cantly re duces therisk of can cer in ci dence.

As the fi nal part of the re search, we in ves ti gatedthe in flu ence of collimation on IQ, as it is cru cial forim ag ing di ag nos tics. In AP pro jec tion of tho racicspine the im prove ment of IQ by 13 % (p = 0.001) wasfound when the op ti mal collimation pro to col was usedin com par i son with the stan dard one used in the hos pi -tal, while in the LAT pro jec tion no sta tis ti cally sig nif i -cant dif fer ence (p = 0.079) in IQ was found whencollimation pro to col men tioned in pro fes sional lit er a -ture was used. That means that better IQ was achievedin the AP pro jec tion by im ple ment ing the op ti malcollimation pro to col and ap pro pri ate po si tion ing. Ourcon clu sions are in ac cor dance with re sults of Karamiand Zabihzadeh [10] and Mileti} [17], stat ing that op -ti mal collimation also con trib utes to IQ im prove mentsince greater collimation re duces the ef fect of scat -tered ra di a tion.


Ta ble 4. Re sults of the re search for LAT pro jec tion

Vari able Collimation Mean Stan dardde vi a tion Me dian Min i mum Max i mum Dif fer ence p-value

BMI*Stan dard 26.9 6.0 25.0 18.8 46.9

0.1 0.549Op ti mal 26.8 4.3 25.9 18.6 35.6

Im age field size [cm2]Stan dard 1259.5 210.7 1243.1 885.8 1505.0

512.1 (41 %) <0.001Op ti mal 747.4 59.4 765.4 571.9 941.7

DAP [µGym²]*Stan dard 80.2 54.6 59.8 23.7 243.2

18.7 (23 %) 0.040Op ti mal 61.5 43.3 48.8 19.7 184.8

Avg. or gan ab sorbed dose [µGy]Stan dard 96.3 57.8 77.2 31.5 281.3

28 % /Op ti mal 70.9 46.3 57.1 25.1 210.1

To tal IQ eval u a tion*Stan dard 7.8 1.2 8.0 4.0 12.0

15 % 0.079Op ti mal 8.6 1.8 8.0 5.0 12.0

*non-para met ric test was used to cal cu late the p value

Ta ble 3. Re sults of the re search for AP pro jec tion

Vari able Collimation Mean Stan dardde vi a tion Me dian Min i mum Max i mum Dif fer ence p-value

BMI*Stan dard 26.9 6.0 25.0 18.8 46.9

0.1 0.549Op ti mal 26.8 4.3 25.9 18.6 35.6

Im age field size [cm2]Stan dard 1046.2 168.2 1029.9 729.3 1505.0

474,6 (45 %) <0.001Op ti mal 571.5 93.4 579.0 374.5 743.9

DAP [µGym²]*Stan dard 92.0 53.4 81.6 13.1 234.0

30.9 (34 %) 0.007Op ti mal 61.2 31.5 54.0 15.0 135.1

Avg. or gan ab sorbed dose [µGy]Stan dard 334.4 182.6 304.0 59.4 840.4

26 % /Op ti mal 265.1 132.8 236.3 68.9 631.3

To tal IQ eval u a tion*Stan dard 6.8 1.7 6.5 3.0 9.0

13 % 0.003Op ti mal 7.8 1.4 9.0 5.0 9.0

*non-para met ric test was used to cal cu late the p value

It should also be noted that IQ was eval u ated by 2ra di ol o gists and 1 radiologic tech nol o gist, and the rat -ings of the as ses sors dif fered. How ever, mean IQ scores were higher for each cri te rion in the group with op ti malcollimation, as well as in AP and LAT pro jec tion. In or -der to avoid the sub jec tive eval u a tion of in di vid ualeval u a tors and to achieve greater ac cu racy and ob jec -tiv ity of this part of the re search, a com par i son could bemade of ob jec tive qual ity as sess ment cri te ria, such assig nal-to-noise ra tio and con trast-to-noise ra tio.

CON CLU SIONS

By op ti mal collimation in tho racic spine ra di og -ra phy, the im age field size, DAP, and the mean ab -sorbed dose to the se lected or gans was on the av er agere duced by 43 %, 29 %, and 27 %, re spec tively, forboth AP and LAT pro jec tion. In ad di tion, the op ti malcollimation pro to col im proved im age qual ity by 13 %in AP pro jec tion. No dif fer ences in im age qual ity were found in LAT pro jec tion.

The re sults of the re search prove that there isplenty of room for im prove ment in clin i cal prac tice re -

gard ing dose re duc tion and im age qual ity by tak ingsim ple steps like proper use of collimation guide lineswhich can eas ily im prove the out comes of ra di og ra -phy pro ce dures that are among those with the high estpa tient ra di a tion dose.

AU THORS' CON TRI BU TIONS

Prep a ra tions of the re search plan, ViewDEX forim age eval u a tion were made by N. Mekiš and D. Škrk.Data col lec tion and prep a ra tion for data anal y sis werecon trib uted by A. Pažanin. Data anal y sis was made byA. Pažanin, and N. Zalokar. All the au thors have con -trib uted to ar ti cle prep a ra tion.

REF ER ENCES

[1] ***, Eu ro pean Com mis sion, Med i cal Ra di a tion Ex -po sure of the Eu ro pean Pop u la tion, Part 1/2, Ra di a -tion Pro tec tion, 180 (2014), 2, pp. 1-181

[2] ***, Eu ro pean Com mis sion, Di ag nos tic Ref er enceLev els in Thirty-six Eu ro pean Coun tries, Part 2/2,Ra di a tion Pro tec tion, 180 (2014), pp. 1-73

[3] Resnik, A., et al., Pel vis Im ag ing: Achiev ing DoseRe duc tion with Dif fer ent Pa tient Po si tions, NuclTechnol Radiat, 34 (2019), 4, pp. 375-83


Ta ble 5. Spe cific ab sorbed or gan dose in AP pro jec tion of the tho racic spine

Or gan Op ti malcollimation Mean ± stdev. [µGy] Mean er ror [%] Me dian [µGy] Min i mum [µGy] Max i mum [µGy] p-value

Ac tive bonemar row

No 91.7 ± 45.2 0.3 86.3 17.6 214.60.004

Yes 64.9 ± 27.6 0.3 60.4 18.1 142.4

Ad re nalsNo 91.0 ± 39.7 1.8 83.2 23.1 194.2

0.188Yes 79.5 ± 29.3 1.8 71.5 21.4 176.0

Gall blad derNo 210.7 ± 138.5 2.2 171.2 50.6 698.6

0.083Yes 151.3 ± 82.1 2.2 147.2 6.6 325.5

HeartNo 650.9 ± 342.2 0.6 619.7 119.5 1759.0

0.816Yes 657.8 ± 292.1 0.5 588.0 212.5 1429.3

LiverNo 284.1 ± 137.3 0.5 258.7 55.0 564.9

<0.001Yes 178.8 ± 76.2 0.5 161.8 33.2 378.7

Lymph nodesNo 147.6 ± 70.8 0.5 136.8 43.7 309.3

0.028Yes 113.1 ± 48.2 0.5 100.0 25.5 243.0

LungsNo 307.0 ± 149.9 0.5 256.1 58.6 593.1

<0.001Yes 110.0 ± 41.9 0.6 105.8 30.2 234.9

Oe soph a gusNo 177.4 ± 81.4 2.3 169.8 40.7 425.2

0.865Yes 173.8 ± 66.9 1.9 158.5 57.2 382.4

Pan creasNo 223.7 ± 111.3 1.7 204.4 0.1 509.9

0.133Yes 191.6 ± 79.8 1.6 165.2 27.3 400.0

Stom achNo 365.7 ± 196.8 1.1 287.8 75.2 824.3

<0.001Yes 143.2 ± 72.34 1.3 132.5 13.1 296.7

Thy roidNo 197.7 ± 162.7 4.0 171.7 17.5 713.7

0.844Yes 214.6 ± 230.2 3.5 136.5 13.9 1235.7

Thy musNo 1131.4 ± 629.4 1.7 1096.4 195.3 3164.3

0.201Yes 1313.7 ± 658.0 1.3 1192.9 422.9 2861.6

Breasts*No 468.2 ± 268.6 0.7 411.3 75.2 954.9

<0.001Yes 53.6 ± 21.6 1.4 51.2 14.1 101.2

* The or gan dose was cal cu lated only for fe male pa tients (n = 70)

[4] Šegota, D., et al., Es tab lish ment of Lo cal Di ag nos ticRef er ence Lev els for Typ i cal Ra di og ra phy Ex am i na -tions in the West Re gion of Croatia, Nucl TechnolRadiat, 34 (2019), 1, pp. 102-6

[5] ***, In ter na tional Com mis sion on Ra dio log i cal pro -tec tion, The 2007 Rec om men da tions of the In ter na -tional Com mis sion on Ra dio log i cal Pro tec tion, ICRPpub li ca tion, 103 (2007), Elsevier, pp. 259-263 ISBN978 07020 30482

[6] Medi}, M., Mekiš, N., Diagnostical Ra di og ra phy Meth -ods: Skel e tal Ra di og ra phy (in Slovenian) Diagnosti~neRadiološke Metode: Skeletna Diagnostika), Fac ulty ofHealth Sci ences, Ljubljana, 2018

[7] Karami, V., Zabihzadeh, M., Beam Collimation Dur -ing Lum bar Spine Ra di og ra phy: a Ret ro spec tiveStudy, Jour nal of Bio med i cal Phys ics & En gi neer ing,7 (2017), 2, pp. 101-106


Ta ble 6. Spe cific ab sorbed or gan dose in LAT pro jec tion of the tho racic spine

Or gan Op ti malcollimation

Mean ± stdev.[µGy]

Mean er ror [%] Me dian [µGy] Min i mum[µGy]

Max i mum[µGy] p-value

Ac tive bonemar row

No 82.5 ± 47.6 0.3 69.2 29.3 211.10.458

Yes 77.4 ± 48.1 0.3 62.6 28.2 210.0

Ad re nalsNo 80.5 ± 38.7 1.6 67.6 34.4 193.2

0.145Yes 97.8 ± 56.2 1.6 81.9 35.1 246.4

Gall blad derNo 42.0 ± 20.2 3.8 35.4 12.8 94.7

<0.001Yes 25.9 ± 16.9 4.3 20.6 7.7 84.1

HeartNo 99.0 ± 49.1 1.2 80.7 31.8 234.7

<0.001Yes 59.6 ± 41.1 1.4 48.3 20.4 176.8

LiverNo 278.3 ± 148.2 0.4 220.0 106.9 757.7

0.006Yes 214.1 ± 150.2 0.5 172.1 75.7 715.7

Lymph nodesNo 44.9 ± 23.1 0.8 37.0 15.6 106.6

0.004Yes 35.0 ± 22.4 0.8 28.4 12.7 98.0

LungsNo 278.3 ± 163.6 0.5 233.4 96.4 735.3

0.018Yes 221.0 ± 152.1 0.5 172.9 75.4 654.5

Oe soph a gusNo 98.9 ± 49.2 2.8 89.3 36.9 249.6

0.900Yes 102.2 ± 60.1 2.6 83.0 39.0 282.6

Pan creasNo 35.7 ± 15.0 3.5 30.6 13.8 73.2

0.037Yes 31.4 ± 18.2 3.5 26.0 12.1 85.5

Stom achNo 10.9 ± 4.3 4.2 9.5 3.7 19.9

<0.001Yes 6.9 ± 3.9 4.8 5.3 2.5 17.7

Thy roidNo 22.5 ± 28.5 9.5 15.1 3.9 188.0

<0.001Yes 12.2 ± 8.6 10.9 10.0 4.7 40.9

Thy musNo 66.2 ± 36.5 5.2 59.6 14.4 192.6

<0.001Yes 26.3 ± 17.3 6.9 21.3 8.3 80.7

Breasts*No 112.1 ± 127.6 1.7 56.2 9.7 600.5

<0.001Yes 12.3 ± 8.4 2.8 10.1 4.6 39.0

[8] Hiles, P. A., et al., IPEM Re port 91 Rec om mendedStan dards for the Rou tine Per for mance Test ing of Di -ag nos tic X-Ray Im ag ing Sys tems, In sti tute of Phys ics and En gi neer ing in Med i cine, York, UK, 2005

[9] Long, B. W., et al., Merrill's At las of Ra dio graphicPo si tion ing and Pro ce dures, Elsevier Health Sci -ences, 2018

[10] Ben-Shlomo, A., et al., Ef fec tive Dose Re duc tion inSpine Ra dio graphic Im ag ing by Choos ing the LessRa di a tion-Sen si tive Side of the Body, The Spine Jour -nal, 16 (2016), 4, pp. 558-563

[11] ***, Busch H P 2004 Im age Qual ity and DoseMangement For Dig i tal Ra di og ra phy

[12] Hakansson, M., et al., ViewDEX: an Ef fi cient andEasy-To-Use Soft ware for Ob server Per for manceStud ies, Ra di a tion Pro tec tion Do sim e try, 139 (2010),1-3, pp. 42-51

[13] Borjesson, S., et al., A Soft ware Tool for In creased Ef -fi ciency in Ob server Per for mance Stud ies in Ra di ol -

ogy, Ra di a tion Pro tec tion Do sim e try, 114 (2005), 1-3, pp. 45-52

[14] Svalkvist, A., et al., ViewDEX: a Sta tus Re port, Ra di a -tion Pro tec tion Do sim e try, 169 (2016), 1-4, pp. 38-45.

[15] Chaparian, A., et al., Re duc tion of Ra di a tion Risks inPa tients Un der go ing Some X-Ray Ex am i na tions byUs ing Op ti mal Pro jec tions: A Monte Carlo Pro -gram-Based Math e mat i cal Cal cu la tion, Jour nal ofMed i cal Phys ics/As so ci a tion of Med i cal Phys i cists ofIn dia, 39 (2014), 1, 32

[16] Rob in son, J. B., et al., Does Collimation Af fect Pa -tient Dose in Antero-Pos te rior Thoraco-Lum barSpine? Ra di og ra phy, 23 (2017), 3, pp. 211-215

[17] Meleti}, D., Sceletal Ra di og ra phy (in Cro atian Lan -guage: Skeletna Diagnostika), Glosa, Rijeka, 2008

Re ceived on Feb ru ary 4, 2020Ac cepted on Au gust 4, 2020


Anamarija PA@ANIN, Damijan [KRK, Nika ZALOKAR, Nejc MEKI[

UTICAJ PRAVILNE KOLIMACIJE NA DOZU ZRAÊWA IKVALITET SLIKE PRI RADIOGRAFIJI TORAKALNE KI^ME

Ciq ovog istraìvawa je utvr|ivawe uticaja kolimacije snopa na pacijentnu dozu ikvalitet slike pri radiografiji torakalne ki~me. Studijom su obuhva}ena 84 pacijenta koji subili upu}eni na snimawe torakalne ki~me. Pacijenti su nasumi~no bili podeqeni u dve grupe od po 42. Prva grupa snimana je prema standardnom protokolu za kolimaciju koji se primewuje u jednoj odbolnica u Hrvatskoj dok je druga grupa snimana primenom “optimalne” kolimacije. Veli~ina poqakolimisana je zasebno za svakog pacijenta ili prema kolimaciji za najve}e poqe na osnovu podataka dostupnih u profesionalnoj literaturi. Za svakog pacijenta zabeleèni su indeks telesne mase,veli~ina poqa, parametri ekspozicije i proizvod doze i povr{ine, izra~unata je apsorbovana dozau ogranima i ocewen kvalitet slike. Nije bilo statisti~ki zna~ajne razlike u indeksu telesnemase za obe grupe pacijenata. Sa optimalnom kolimacijom, veli~ina poqa pri anteroposteriornojprojekciji smawena je za 45 % (p < 0.001), a pri lateralnoj projekciji smawena je za 41 % (p = 0.007)pri anteroposteriornoj projekciji i 23 % (p = 0.040) pri lateralnoj projekciji. Sredwaapsorbovana doza u odabranim organima umewena je za 26 % pri anteroposteriornoj projekciji, a za28 % pri lateralnoj projekciji. Dodatno, protokol sa optimalnom kolimacijom poboq{ao jekvalitet slike za 13 % pri anteroposteriornoj projekciji. Pri snimawu u lateralnoj projekcijinisu uo~ene razlike u kvalitetu slike. Ovim istraìvawem pokazali smo da optimalna kolimacija pri radiografiji torakalne ki~me ima zna~ajan uticaj na izlagawe pacijenta zra~ewu i imapozitivni uticaj na kvalitet slike.

Kqu~ne re~i: radiografija torakalne ki~me, kolimacija, smawewe doze, kvalitet slike

Documents

PROPER COLLIMATION EF FECT ON RA DI ATION DOSE AND IM …