Repository logo
 

On unsupervised methods for medical image segmentation: Investigating classic approaches in breast cancer dce-mri

cam.issuedOnline2021-12-24
dc.contributor.authorMilitello, C
dc.contributor.authorRanieri, A
dc.contributor.authorRundo, L
dc.contributor.authorD’angelo, I
dc.contributor.authorMarinozzi, F
dc.contributor.authorBartolotta, TV
dc.contributor.authorBini, F
dc.contributor.authorRusso, G
dc.contributor.orcidMilitello, Carmelo [0000-0003-2249-9538]
dc.contributor.orcidRundo, Leonardo [0000-0003-3341-5483]
dc.contributor.orcidMarinozzi, Franco [0000-0002-4872-2980]
dc.contributor.orcidBartolotta, Tommaso Vincenzo [0000-0002-8808-379X]
dc.contributor.orcidBini, Fabiano [0000-0002-5641-1189]
dc.contributor.orcidRusso, Giorgio [0000-0003-1493-1087]
dc.date.accessioned2022-01-07T16:48:27Z
dc.date.available2022-01-07T16:48:27Z
dc.date.issued2022
dc.date.updated2022-01-07T16:48:26Z
dc.description.abstract<jats:p>Unsupervised segmentation techniques, which do not require labeled data for training and can be more easily integrated into the clinical routine, represent a valid solution especially from a clinical feasibility perspective. Indeed, large-scale annotated datasets are not always available, undermining their immediate implementation and use in the clinic. Breast cancer is the most common cause of cancer death in women worldwide. In this study, breast lesion delineation in Dynamic Contrast Enhanced MRI (DCE-MRI) series was addressed by means of four popular unsupervised segmentation approaches: Split-and-Merge combined with Region Growing (SMRG), k-means, Fuzzy C-Means (FCM), and spatial FCM (sFCM). They represent well-established pattern recognition techniques that are still widely used in clinical research. Starting from the basic versions of these segmentation approaches, during our analysis, we identified the shortcomings of each of them, proposing improved versions, as well as developing ad hoc pre- and post-processing steps. The obtained experimental results, in terms of area-based—namely, Dice Index (DI), Jaccard Index (JI), Sensitivity, Specificity, False Positive Ratio (FPR), False Negative Ratio (FNR)—and distance-based metrics—Mean Absolute Distance (MAD), Maximum Distance (MaxD), Hausdorff Distance (HD)—encourage the use of unsupervised machine learning techniques in medical image segmentation. In particular, fuzzy clustering approaches (namely, FCM and sFCM) achieved the best performance. In fact, for area-based metrics, they obtained DI = 78.23% ± 6.50 (sFCM), JI = 65.90% ± 8.14 (sFCM), sensitivity = 77.84% ± 8.72 (FCM), specificity = 87.10% ± 8.24 (sFCM), FPR = 0.14 ± 0.12 (sFCM), and FNR = 0.22 ± 0.09 (sFCM). Concerning distance-based metrics, they obtained MAD = 1.37 ± 0.90 (sFCM), MaxD = 4.04 ± 2.87 (sFCM), and HD = 2.21 ± 0.43 (FCM). These experimental findings suggest that further research would be useful for advanced fuzzy logic techniques specifically tailored to medical image segmentation.</jats:p>
dc.identifier.doi10.17863/CAM.79797
dc.identifier.eissn2076-3417
dc.identifier.issn2076-3417
dc.identifier.urihttps://www.repository.cam.ac.uk/handle/1810/332351
dc.languageen
dc.language.isoeng
dc.publisherMDPI AG
dc.publisher.urlhttp://dx.doi.org/10.3390/app12010162
dc.subjectmedical image segmentation
dc.subjectbreast cancer
dc.subjectpattern recognition
dc.subjectmachine learning
dc.subjectclinical feasibility
dc.subjectmagnetic resonance imaging
dc.subjectcomputer-assisted segmentation
dc.titleOn unsupervised methods for medical image segmentation: Investigating classic approaches in breast cancer dce-mri
dc.typeArticle
dcterms.dateAccepted2021-12-19
prism.issueIdentifier1
prism.publicationNameApplied Sciences (Switzerland)
prism.volume12
pubs.funder-project-idItalian MISE (GeSeTon project)
rioxxterms.licenseref.urihttps://creativecommons.org/licenses/by/4.0/
rioxxterms.versionVoR
rioxxterms.versionofrecord10.3390/app12010162

Files

Original bundle
Now showing 1 - 3 of 3
No Thumbnail Available
Name:
additional-files.zip
Size:
1.01 MB
Format:
ZIP file
Description:
Supporting information
Licence
https://creativecommons.org/licenses/by/4.0/
No Thumbnail Available
Name:
applsci-12-00162.xml
Size:
11.86 KB
Format:
Extensible Markup Language
Description:
Bibliographic metadata
Licence
https://creativecommons.org/licenses/by/4.0/
Loading...
Thumbnail Image
Name:
applsci-12-00162.pdf
Size:
1.33 MB
Format:
Adobe Portable Document Format
Description:
Published version
Licence
https://creativecommons.org/licenses/by/4.0/