3
Editorial Image Segmentation Techniques for Healthcare Systems Orazio Gambino , 1 Vincenzo Conti , 2 Sergio Galdino , 3 Cesare Fabio Valenti , 4 and Wellington Pinheiro dos Santos 5 1 Dept. of Engineering, University of Palermo, Palermo, Italy 2 Faculty of Engineering and Architecture, University of Enna KORE, Enna, Italy 3 Polytechnique School of Pernambuco, University of Pernambuco, Recife, Brazil 4 Dept. of Mathematics and Informatics, University of Palermo, Palermo, Italy 5 Dept. of Biomedical Engineering, Federal University of Pernambuco, Recife, Brazil Correspondence should be addressed to Orazio Gambino; [email protected] Received 16 January 2019; Accepted 16 January 2019; Published 2 April 2019 Copyright © 2019 Orazio Gambino et al. is is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. e present special issue of the Journal of Healthcare En- gineering collects articles written by researchers scattered around the world who belong to the academic and industrial environments. e papers of this special issue have been selected by a rigorous peer-reviewing process with the support of at least two reviewers per paper, along with the opinion written in the final decision by a component of the editorial staff. Different methods on biomedical image segmentation dedicated to healthcare systems have been developed regarding, for example, the fields of machine learning, deformable models, fuzzy models, and so on. Such methods have been applied on different biomedical image modalities (MRI, CT, mammograms, optical coherence tomography, and others) of various anatomical districts, such as the brain, thyroid, lung, and breast. J. Gauci et al. present an automatic approach to determine the tempera- ture of the body’s extremities by means of thermal images in diabetic patients. e approach is based on morphological operations, and geometric transformations are aimed at automatically extracting the required data from 44 pre- defined regions of interest. e method is tested on data from 395 participants. A correct extraction in around 90% of the images was achieved. A. Bougacha et al. investigate a novel classification method for 3D multimodal MRI glio- blastoma tumor characterization. ey propose the seg- mentation problem as a linear mixture model (LMM). First, they provide a nonnegative matrix M from every MRI slice in every segmentation process’ step (matrix used as an input for the first segmentation process to extract the edema re- gion from T2 and FLAIR modalities); after that, they extract edema’s region from T1c modality, generate the matrix M, and segment the necrosis, the enhanced tumor, and the nonenhanced tumor regions. In the segmentation process, they apply a rank-two NMF clustering. C. Wang et al. quantify the subregional alveolar bone changes during or- thodontic tooth movement with a novel method. Ortho- dontic tooth movement (OTM) is the result of the region- specific bone modeling under a load. Quantification of this change in the alveolar bone around a tooth is a basic re- quirement to understand the mechanism of orthodontics. ey have used 12 Sprague-Dawley (SD) rats as an ortho- dontic model, and one side of the first upper molar has been used to simulate OTM. e alveolar bone around the mesial root has been reconstructed from in vivo micro-CT images and separated from other parts of the alveolar bone with two semicylinder filters. e amount and rate of OTM, bone mineral density (BMD), and bone volume (BV) around the root have been calculated and compared at 5 time points. C. L. Toledo Peral et al. present an application for skin macules characterization and could be the background of a future diagnosis-assistance-tool for educational and preventive assistance technology purposes, based on a three-stage segmentation and characterization algorithm used to clas- sify vascular, petechiae, trophic changes, and trauma mac- ules from digital photographs of the lower limbs. First, in order to find the skin region, a logical multiplication is Hindawi Journal of Healthcare Engineering Volume 2019, Article ID 2723419, 2 pages https://doi.org/10.1155/2019/2723419

Editorial ImageSegmentationTechniquesforHealthcareSystemsdownloads.hindawi.com/journals/jhe/2019/2723419.pdf · editorial staff. Different methods on biomedical image segmentation

  • Upload
    others

  • View
    4

  • Download
    0

Embed Size (px)

Citation preview

Page 1: Editorial ImageSegmentationTechniquesforHealthcareSystemsdownloads.hindawi.com/journals/jhe/2019/2723419.pdf · editorial staff. Different methods on biomedical image segmentation

EditorialImage Segmentation Techniques for Healthcare Systems

Orazio Gambino ,1 Vincenzo Conti ,2 Sergio Galdino ,3 Cesare Fabio Valenti ,4

and Wellington Pinheiro dos Santos 5

1Dept. of Engineering, University of Palermo, Palermo, Italy2Faculty of Engineering and Architecture, University of Enna KORE, Enna, Italy3Polytechnique School of Pernambuco, University of Pernambuco, Recife, Brazil4Dept. of Mathematics and Informatics, University of Palermo, Palermo, Italy5Dept. of Biomedical Engineering, Federal University of Pernambuco, Recife, Brazil

Correspondence should be addressed to Orazio Gambino; [email protected]

Received 16 January 2019; Accepted 16 January 2019; Published 2 April 2019

Copyright © 2019 Orazio Gambino et al. $is is an open access article distributed under the Creative Commons AttributionLicense, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work isproperly cited.

$e present special issue of the Journal of Healthcare En-gineering collects articles written by researchers scatteredaround the world who belong to the academic and industrialenvironments. $e papers of this special issue have beenselected by a rigorous peer-reviewing process with thesupport of at least two reviewers per paper, along with theopinion written in the final decision by a component of theeditorial staff. Different methods on biomedical imagesegmentation dedicated to healthcare systems have beendeveloped regarding, for example, the fields of machinelearning, deformable models, fuzzy models, and so on. Suchmethods have been applied on different biomedical imagemodalities (MRI, CT, mammograms, optical coherencetomography, and others) of various anatomical districts,such as the brain, thyroid, lung, and breast. J. Gauci et al.present an automatic approach to determine the tempera-ture of the body’s extremities by means of thermal images indiabetic patients. $e approach is based on morphologicaloperations, and geometric transformations are aimed atautomatically extracting the required data from 44 pre-defined regions of interest. $e method is tested on datafrom 395 participants. A correct extraction in around 90% ofthe images was achieved. A. Bougacha et al. investigate anovel classification method for 3D multimodal MRI glio-blastoma tumor characterization. $ey propose the seg-mentation problem as a linear mixture model (LMM). First,they provide a nonnegative matrix M from every MRI slicein every segmentation process’ step (matrix used as an input

for the first segmentation process to extract the edema re-gion from T2 and FLAIR modalities); after that, they extractedema’s region from T1c modality, generate the matrix M,and segment the necrosis, the enhanced tumor, and thenonenhanced tumor regions. In the segmentation process,they apply a rank-two NMF clustering. C. Wang et al.quantify the subregional alveolar bone changes during or-thodontic tooth movement with a novel method. Ortho-dontic tooth movement (OTM) is the result of the region-specific bone modeling under a load. Quantification of thischange in the alveolar bone around a tooth is a basic re-quirement to understand the mechanism of orthodontics.$ey have used 12 Sprague-Dawley (SD) rats as an ortho-dontic model, and one side of the first upper molar has beenused to simulate OTM. $e alveolar bone around the mesialroot has been reconstructed from in vivo micro-CT imagesand separated from other parts of the alveolar bone with twosemicylinder filters. $e amount and rate of OTM, bonemineral density (BMD), and bone volume (BV) around theroot have been calculated and compared at 5 time points. C.L. Toledo Peral et al. present an application for skin maculescharacterization and could be the background of a futurediagnosis-assistance-tool for educational and preventiveassistance technology purposes, based on a three-stagesegmentation and characterization algorithm used to clas-sify vascular, petechiae, trophic changes, and trauma mac-ules from digital photographs of the lower limbs. First, inorder to find the skin region, a logical multiplication is

HindawiJournal of Healthcare EngineeringVolume 2019, Article ID 2723419, 2 pageshttps://doi.org/10.1155/2019/2723419

Page 2: Editorial ImageSegmentationTechniquesforHealthcareSystemsdownloads.hindawi.com/journals/jhe/2019/2723419.pdf · editorial staff. Different methods on biomedical image segmentation

performed on two skin masks obtained from color spacetransformations. $en, in order to locate the lesion region,illumination enhancement is performed using a chromaticmodel color space, followed by a principal componentanalysis grayscale transformation. Finally, characteristics ofeach type of macule are considered and classified; mor-phologic properties (area, axes, perimeter, and solidity),intensity properties, and a set of shade indices (red, green,blue, and brown) are proposed as a measure to obviate skincolor differences among subjects. E. Pociask et al. present amethod aimed at the automated lumen segmentation onoptical coherence tomography images. $e adopted ap-proach is composed of a preprocessing phase (artifacts re-moval: speckle noise, circular rings, and guide wire) andmorphological operations. $e method is tested on 667images of different patients from the Medical University ofSilesia, and it has been compared with other ones by usingobjective measures. R. Zhang et al. propose an improvedfuzzy connectedness (FC) method for three-dimensional(3D) liver vessel segmentation on computed tomography(CT) images. In particular, a novel method to define thefuzzy affinity function of FC is presented, and an improvedfilter based on adaptive sigmoid filtering is proposed. $eproposed approach is evaluated in 40 cases of clinical CTvolumetric images from public image repositories. Y. Gaoet al. proposed a novel technique to perform the optic discdetection on retinal images. $e elaboration pipeline in-volves an initial rough segmentation based on saliencydetection and largest object selection to define the optic discinitial contour. Subsequently, a method based on a de-formable model refines the result. $e effectiveness of suchan approach has been evaluated on a public retinal imagerepository. A. Cruz-Bernal et al. propose a method for theautomated detection of malignant calcifications on mam-mography images. $e approach is based on the analysis ofthe cluster prominence (cp) feature histogram. Indeed, thecalcifications on the mammography are characterized byhigh occurrences in the histogram. $erefore, the Vander-monde interpolation is used to obtain a function whichmodels the cp histogram, and a KNN classifier finalizes themethod. Z. Kong et al. propose a method to segment MRbrain images by means of a convolutional neural network.Initially, a wavelet technique is used to extract the contoursof different tissues such as skull, cerebrospinal fluid (CSF),grey matter (GM), and white matter (WM). $erefore, aconvolutional neural network refines the segmentation re-sults. B. Khagi and G.-R. Kwon proposed a method tosegment MR brain images using SegNet, a convolutionalneural network. Such a CNN is trained by using pre-segmented MR brain images of the OASIS free dataset. Z. Z.Wang et al. described a simple, yet effective approach tosegment three-dimensional livers in computed tomographyimaging. $is research field is of particular interest becausethe described task requires considerable experience by theclinician and it is subject to personal interpretation. $eproposed method is based on multiple thresholds throughslope difference distribution, on Gibbs energy minimizationto reduce inhomogeneity and on proper mathematicalmorphology operations to refine the segmented components,

represented by their spline contours. $is heuristic was finetuned experimentally on a variety of real cases on publicdatasets. C.-Y. Lee et al. introduced a neural network ap-proach to deal with sonograms in the case of breast cancer.Potential applications to reduce the amount of noise andintensity inhomogeneity are quite evident in this particularfield. $e proposed technique has been validated on two realdatasets and comparing different networkmodels.$e imagesare passed to a stack denoised autoencoder-based classifier toenhance the contrast of the ultrasound signal. $e extractionof the most promising features and the comparison withmanually labeled data allowed to train the system, thusgaining a final accuracy equal to 85%. J. Hai et al. haveaddressed the problem of automatic breast cancer segmen-tation through fully convolutional networks, able to detectabstract features from the input data. Moreover, a non-standard multiresolution approach led to definition of ageneral technique, independent of the tumor size. Indeed,while the usual MRA analysis is efficient for compressionpurposes, the redundant methodology chosen by the authorsretains the original resolution of the mammograms, andtherefore, it is particularly suited for their automatic seg-mentation. Experimental results are presented on a vastproprietary dataset of real clinical cases. E. Dandıl proposed atechnique for the automatic detection and classification oflung cancer in computed tomography. A preliminary phaseidentifies candidate pulmonary nodules, which are segmentedby self-organizing maps, investigated by principal componentanalysis and classified due to a probabilistic neural network.$e computer-aided diagnostic system achieves high accu-racy, sensitivity, and specificity on a publicly available dataset.W. Tan et al. have defined an interesting technique for theidentification of pulmonary vascular structures in chestcomputed tomography. $ey proved that the region growingwith maximum between-class variance segmentation isenough to select the regions of interest. $e vascular com-ponents are therefore fully located through the fast marchingmethod. Experimental comparisons were done against otherwell-known methods, and the accuracy of this technique isabout 90%.

Conflicts of Interest

$e editors declare that they have no conflicts of interestregarding the publication of this special issue.

Acknowledgments

$e lead guest editor and his editorial staff express theirgratefulness both to all the reviewers for their precioussupport and to the authors who decided to publish theirworks in this special issue.

Orazio GambinoVincenzo ContiSergio Galdino

Cesare Fabio ValentiWellington Pinheiro dos Santos

2 Journal of Healthcare Engineering

Page 3: Editorial ImageSegmentationTechniquesforHealthcareSystemsdownloads.hindawi.com/journals/jhe/2019/2723419.pdf · editorial staff. Different methods on biomedical image segmentation

International Journal of

AerospaceEngineeringHindawiwww.hindawi.com Volume 2018

RoboticsJournal of

Hindawiwww.hindawi.com Volume 2018

Hindawiwww.hindawi.com Volume 2018

Active and Passive Electronic Components

VLSI Design

Hindawiwww.hindawi.com Volume 2018

Hindawiwww.hindawi.com Volume 2018

Shock and Vibration

Hindawiwww.hindawi.com Volume 2018

Civil EngineeringAdvances in

Acoustics and VibrationAdvances in

Hindawiwww.hindawi.com Volume 2018

Hindawiwww.hindawi.com Volume 2018

Electrical and Computer Engineering

Journal of

Advances inOptoElectronics

Hindawiwww.hindawi.com

Volume 2018

Hindawi Publishing Corporation http://www.hindawi.com Volume 2013Hindawiwww.hindawi.com

The Scientific World Journal

Volume 2018

Control Scienceand Engineering

Journal of

Hindawiwww.hindawi.com Volume 2018

Hindawiwww.hindawi.com

Journal ofEngineeringVolume 2018

SensorsJournal of

Hindawiwww.hindawi.com Volume 2018

International Journal of

RotatingMachinery

Hindawiwww.hindawi.com Volume 2018

Modelling &Simulationin EngineeringHindawiwww.hindawi.com Volume 2018

Hindawiwww.hindawi.com Volume 2018

Chemical EngineeringInternational Journal of Antennas and

Propagation

International Journal of

Hindawiwww.hindawi.com Volume 2018

Hindawiwww.hindawi.com Volume 2018

Navigation and Observation

International Journal of

Hindawi

www.hindawi.com Volume 2018

Advances in

Multimedia

Submit your manuscripts atwww.hindawi.com