Multiclass ensemble framework for enhanced prostate gland Segmentation: Integrating Self-ONN decoders with EfficientNet

<p dir="ltr"><u>Digital pathology</u> relies on the morphological architecture of prostate glands to recognize cancerous tissue. <u>Prostate cancer</u> (PCa) originates in walnut shaped prostate gland in the <u>male reproductive system</u>. <u&g...

Full description

Saved in:
Bibliographic Details
Main Author: Md. Shaheenur Islam Sumon (22045907) (author)
Other Authors: Muhammad E.H. Chowdhury (17151154) (author), Enamul Hoque Bhuiyan (22803500) (author), Md. Sohanur Rahman (8430495) (author), Muntakim Mahmud Khan (22803503) (author), Israa Al-Hashimi (18131374) (author), Adam Mushtak (22048013) (author), Sohaib Bassam Zoghoul (22803506) (author)
Published: 2025
Subjects:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1864513532044771328
author Md. Shaheenur Islam Sumon (22045907)
author2 Muhammad E.H. Chowdhury (17151154)
Enamul Hoque Bhuiyan (22803500)
Md. Sohanur Rahman (8430495)
Muntakim Mahmud Khan (22803503)
Israa Al-Hashimi (18131374)
Adam Mushtak (22048013)
Sohaib Bassam Zoghoul (22803506)
author2_role author
author
author
author
author
author
author
author_facet Md. Shaheenur Islam Sumon (22045907)
Muhammad E.H. Chowdhury (17151154)
Enamul Hoque Bhuiyan (22803500)
Md. Sohanur Rahman (8430495)
Muntakim Mahmud Khan (22803503)
Israa Al-Hashimi (18131374)
Adam Mushtak (22048013)
Sohaib Bassam Zoghoul (22803506)
author_role author
dc.creator.none.fl_str_mv Md. Shaheenur Islam Sumon (22045907)
Muhammad E.H. Chowdhury (17151154)
Enamul Hoque Bhuiyan (22803500)
Md. Sohanur Rahman (8430495)
Muntakim Mahmud Khan (22803503)
Israa Al-Hashimi (18131374)
Adam Mushtak (22048013)
Sohaib Bassam Zoghoul (22803506)
dc.date.none.fl_str_mv 2025-07-30T09:00:00Z
dc.identifier.none.fl_str_mv 10.1016/j.compbiomed.2025.110459
dc.relation.none.fl_str_mv https://figshare.com/articles/journal_contribution/Multiclass_ensemble_framework_for_enhanced_prostate_gland_Segmentation_Integrating_Self-ONN_decoders_with_EfficientNet/30819533
dc.rights.none.fl_str_mv CC BY 4.0
info:eu-repo/semantics/openAccess
dc.subject.none.fl_str_mv Biomedical and clinical sciences
Oncology and carcinogenesis
Engineering
Biomedical engineering
Health sciences
Health services and systems
Information and computing sciences
Artificial intelligence
Machine learning
Peripheral zone
Transitional zone
Self-ONN
MRI
Semantic segmentation
Deep learning
dc.title.none.fl_str_mv Multiclass ensemble framework for enhanced prostate gland Segmentation: Integrating Self-ONN decoders with EfficientNet
dc.type.none.fl_str_mv Text
Journal contribution
info:eu-repo/semantics/publishedVersion
text
contribution to journal
description <p dir="ltr"><u>Digital pathology</u> relies on the morphological architecture of prostate glands to recognize cancerous tissue. <u>Prostate cancer</u> (PCa) originates in walnut shaped prostate gland in the <u>male reproductive system</u>. <u>Deep learning</u> (DL) pipelines can assist in identifying these regions with advanced <u>segmentation techniques</u> which are effective in diagnosing and treating <u>prostate diseases</u>. This facilitates early detection, targeted biopsy, and accurate treatment planning, ensuring consistent, reproducible results while minimizing human error. Automated segmentation techniques trained on MRI datasets can aid in monitoring disease progression which leads to clinical support by developing patient-specific models for personalized medicine. In this study, we present multiclass <u>segmentation models</u> designed to localize the prostate gland and its zonal regions—specifically the <u>peripheral zone</u> (PZ), transition zone (TZ), and the whole gland—by combining EfficientNetB4 encoders with Self-organized <u>Operational Neural Network</u> (Self-ONN)-based decoders. Traditional <u>convolutional neural networks</u> (CNNs) rely on linear neuron models, which limit their ability to capture the complex dynamics of biological neural systems. In contrast, Operational Neural Networks (ONNs), particularly Self-ONNs, address this limitation by incorporating nonlinear and adaptive operations at the neuron level. We evaluated various encoder-decoder configurations and identified that the combination of an EfficientNet-based encoder with a Self-ONN-based decoder yielded the best performance. To further enhance <u>segmentation accuracy</u>, we employed the STAPLE method to ensemble the top three performing models. Our approach was tested on the large-scale, recently updated PI-CAI Challenge dataset using 5-fold cross-validation, achieving Dice scores of 95.33 % for the whole gland and 92.32 % for the combined PZ and TZ regions. These advanced segmentation techniques significantly improve the quality of PCa diagnosis and treatment, contributing to better patient care and outcomes.</p><h2>Other Information</h2><p dir="ltr">Published in: Computers in Biology and Medicine<br>License: <a href="http://creativecommons.org/licenses/by/4.0/" target="_blank">http://creativecommons.org/licenses/by/4.0/</a><br>See article on publisher's website: <a href="https://dx.doi.org/10.1016/j.compbiomed.2025.110459" target="_blank">https://dx.doi.org/10.1016/j.compbiomed.2025.110459</a></p>
eu_rights_str_mv openAccess
id Manara2_26bd8d701fa26a1afb387b0abf76091f
identifier_str_mv 10.1016/j.compbiomed.2025.110459
network_acronym_str Manara2
network_name_str Manara2
oai_identifier_str oai:figshare.com:article/30819533
publishDate 2025
repository.mail.fl_str_mv
repository.name.fl_str_mv
repository_id_str
rights_invalid_str_mv CC BY 4.0
spelling Multiclass ensemble framework for enhanced prostate gland Segmentation: Integrating Self-ONN decoders with EfficientNetMd. Shaheenur Islam Sumon (22045907)Muhammad E.H. Chowdhury (17151154)Enamul Hoque Bhuiyan (22803500)Md. Sohanur Rahman (8430495)Muntakim Mahmud Khan (22803503)Israa Al-Hashimi (18131374)Adam Mushtak (22048013)Sohaib Bassam Zoghoul (22803506)Biomedical and clinical sciencesOncology and carcinogenesisEngineeringBiomedical engineeringHealth sciencesHealth services and systemsInformation and computing sciencesArtificial intelligenceMachine learningPeripheral zoneTransitional zoneSelf-ONNMRISemantic segmentationDeep learning<p dir="ltr"><u>Digital pathology</u> relies on the morphological architecture of prostate glands to recognize cancerous tissue. <u>Prostate cancer</u> (PCa) originates in walnut shaped prostate gland in the <u>male reproductive system</u>. <u>Deep learning</u> (DL) pipelines can assist in identifying these regions with advanced <u>segmentation techniques</u> which are effective in diagnosing and treating <u>prostate diseases</u>. This facilitates early detection, targeted biopsy, and accurate treatment planning, ensuring consistent, reproducible results while minimizing human error. Automated segmentation techniques trained on MRI datasets can aid in monitoring disease progression which leads to clinical support by developing patient-specific models for personalized medicine. In this study, we present multiclass <u>segmentation models</u> designed to localize the prostate gland and its zonal regions—specifically the <u>peripheral zone</u> (PZ), transition zone (TZ), and the whole gland—by combining EfficientNetB4 encoders with Self-organized <u>Operational Neural Network</u> (Self-ONN)-based decoders. Traditional <u>convolutional neural networks</u> (CNNs) rely on linear neuron models, which limit their ability to capture the complex dynamics of biological neural systems. In contrast, Operational Neural Networks (ONNs), particularly Self-ONNs, address this limitation by incorporating nonlinear and adaptive operations at the neuron level. We evaluated various encoder-decoder configurations and identified that the combination of an EfficientNet-based encoder with a Self-ONN-based decoder yielded the best performance. To further enhance <u>segmentation accuracy</u>, we employed the STAPLE method to ensemble the top three performing models. Our approach was tested on the large-scale, recently updated PI-CAI Challenge dataset using 5-fold cross-validation, achieving Dice scores of 95.33 % for the whole gland and 92.32 % for the combined PZ and TZ regions. These advanced segmentation techniques significantly improve the quality of PCa diagnosis and treatment, contributing to better patient care and outcomes.</p><h2>Other Information</h2><p dir="ltr">Published in: Computers in Biology and Medicine<br>License: <a href="http://creativecommons.org/licenses/by/4.0/" target="_blank">http://creativecommons.org/licenses/by/4.0/</a><br>See article on publisher's website: <a href="https://dx.doi.org/10.1016/j.compbiomed.2025.110459" target="_blank">https://dx.doi.org/10.1016/j.compbiomed.2025.110459</a></p>2025-07-30T09:00:00ZTextJournal contributioninfo:eu-repo/semantics/publishedVersiontextcontribution to journal10.1016/j.compbiomed.2025.110459https://figshare.com/articles/journal_contribution/Multiclass_ensemble_framework_for_enhanced_prostate_gland_Segmentation_Integrating_Self-ONN_decoders_with_EfficientNet/30819533CC BY 4.0info:eu-repo/semantics/openAccessoai:figshare.com:article/308195332025-07-30T09:00:00Z
spellingShingle Multiclass ensemble framework for enhanced prostate gland Segmentation: Integrating Self-ONN decoders with EfficientNet
Md. Shaheenur Islam Sumon (22045907)
Biomedical and clinical sciences
Oncology and carcinogenesis
Engineering
Biomedical engineering
Health sciences
Health services and systems
Information and computing sciences
Artificial intelligence
Machine learning
Peripheral zone
Transitional zone
Self-ONN
MRI
Semantic segmentation
Deep learning
status_str publishedVersion
title Multiclass ensemble framework for enhanced prostate gland Segmentation: Integrating Self-ONN decoders with EfficientNet
title_full Multiclass ensemble framework for enhanced prostate gland Segmentation: Integrating Self-ONN decoders with EfficientNet
title_fullStr Multiclass ensemble framework for enhanced prostate gland Segmentation: Integrating Self-ONN decoders with EfficientNet
title_full_unstemmed Multiclass ensemble framework for enhanced prostate gland Segmentation: Integrating Self-ONN decoders with EfficientNet
title_short Multiclass ensemble framework for enhanced prostate gland Segmentation: Integrating Self-ONN decoders with EfficientNet
title_sort Multiclass ensemble framework for enhanced prostate gland Segmentation: Integrating Self-ONN decoders with EfficientNet
topic Biomedical and clinical sciences
Oncology and carcinogenesis
Engineering
Biomedical engineering
Health sciences
Health services and systems
Information and computing sciences
Artificial intelligence
Machine learning
Peripheral zone
Transitional zone
Self-ONN
MRI
Semantic segmentation
Deep learning