Hybrid Neural Networks for Precise Hydronephrosis Classification Using Deep Learning

<h3 dir="ltr">Objective</h3><p dir="ltr">To develop and evaluate a deep learning framework for automatic kidney and fluid segmentation in renal ultrasound images, aiming to enhance diagnostic accuracy and reduce variability in hydronephrosis assessment. </p&g...

Full description

Saved in:
Bibliographic Details
Main Author: Abdus Salam (1918387) (author)
Other Authors: Mansura Naznine (21399893) (author), Muhammad E.H. Chowdhury (17151154) (author), Saidanvar Agzamkhodjaev (22928887) (author), Ali Tekin (22155874) (author), Santiago Vallasciani (10583165) (author), Elias Ramírez-Velázquez (20487203) (author), Tariq O. Abbas (11247771) (author)
Published: 2025
Subjects:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1864513525503754240
author Abdus Salam (1918387)
author2 Mansura Naznine (21399893)
Muhammad E.H. Chowdhury (17151154)
Saidanvar Agzamkhodjaev (22928887)
Ali Tekin (22155874)
Santiago Vallasciani (10583165)
Elias Ramírez-Velázquez (20487203)
Tariq O. Abbas (11247771)
author2_role author
author
author
author
author
author
author
author_facet Abdus Salam (1918387)
Mansura Naznine (21399893)
Muhammad E.H. Chowdhury (17151154)
Saidanvar Agzamkhodjaev (22928887)
Ali Tekin (22155874)
Santiago Vallasciani (10583165)
Elias Ramírez-Velázquez (20487203)
Tariq O. Abbas (11247771)
author_role author
dc.creator.none.fl_str_mv Abdus Salam (1918387)
Mansura Naznine (21399893)
Muhammad E.H. Chowdhury (17151154)
Saidanvar Agzamkhodjaev (22928887)
Ali Tekin (22155874)
Santiago Vallasciani (10583165)
Elias Ramírez-Velázquez (20487203)
Tariq O. Abbas (11247771)
dc.date.none.fl_str_mv 2025-12-09T03:00:00Z
dc.identifier.none.fl_str_mv 10.1016/j.urology.2025.08.005
dc.relation.none.fl_str_mv https://figshare.com/articles/journal_contribution/Hybrid_Neural_Networks_for_Precise_Hydronephrosis_Classification_Using_Deep_Learning/30971659
dc.rights.none.fl_str_mv CC BY 4.0
info:eu-repo/semantics/openAccess
dc.subject.none.fl_str_mv Health sciences
Health services and systems
Information and computing sciences
Machine learning
Deep Learning
Kidney Segmentation
Fluid Segmentation
Renal Ultrasound Imaging
Hydronephrosis Assessment
dc.title.none.fl_str_mv Hybrid Neural Networks for Precise Hydronephrosis Classification Using Deep Learning
dc.type.none.fl_str_mv Text
Journal contribution
info:eu-repo/semantics/publishedVersion
text
contribution to journal
description <h3 dir="ltr">Objective</h3><p dir="ltr">To develop and evaluate a deep learning framework for automatic kidney and fluid segmentation in renal ultrasound images, aiming to enhance diagnostic accuracy and reduce variability in hydronephrosis assessment. </p><h3 dir="ltr">Methods</h3><p dir="ltr">A dataset of 1731 renal ultrasound images, annotated by four experienced urologists, was used for model training and evaluation. The proposed framework integrates a DenseNet201 backbone, Feature Pyramid Network (FPN), and Self-Organizing Neural Network (SelfONN) layers to enable multi-scale feature extraction and improve spatial precision. Several architectures were tested under identical conditions to ensure a fair comparison. Segmentation performance was assessed using standard metrics, including the Dice coefficient, precision, and recall. The framework also supported hydronephrosis classification using the fluid-to-kidney area ratio, with a threshold of 0.213 derived from prior literature. </p><h3 dir="ltr">Results</h3><p dir="ltr">The model achieved strong segmentation performance for kidneys (Dice: 0.92, precision: 0.93, recall: 0.91) and fluid regions (Dice: 0.89, precision: 0.90, recall: 0.88), outperforming baseline methods. The classification accuracy for detecting hydronephrosis reached 94%, based on the computed fluid-to-kidney ratio. Performance was consistent across varied image qualities, reflecting the robustness of the overall architecture. </p><h3 dir="ltr">Conclusion</h3><p dir="ltr">This study presents an automated, objective pipeline for analyzing renal ultrasound images. The proposed framework supports high segmentation accuracy and reliable classification, facilitating standardized and reproducible hydronephrosis assessment. Future work will focus on model optimization and incorporating explainable AI to enhance clinical integration.</p><h2 dir="ltr">Other Information</h2><p dir="ltr">Published in: Urology<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.urology.2025.08.005" target="_blank">https://dx.doi.org/10.1016/j.urology.2025.08.005</a></p>
eu_rights_str_mv openAccess
id Manara2_fef7b87dd280bc4f93411c93034205bd
identifier_str_mv 10.1016/j.urology.2025.08.005
network_acronym_str Manara2
network_name_str Manara2
oai_identifier_str oai:figshare.com:article/30971659
publishDate 2025
repository.mail.fl_str_mv
repository.name.fl_str_mv
repository_id_str
rights_invalid_str_mv CC BY 4.0
spelling Hybrid Neural Networks for Precise Hydronephrosis Classification Using Deep LearningAbdus Salam (1918387)Mansura Naznine (21399893)Muhammad E.H. Chowdhury (17151154)Saidanvar Agzamkhodjaev (22928887)Ali Tekin (22155874)Santiago Vallasciani (10583165)Elias Ramírez-Velázquez (20487203)Tariq O. Abbas (11247771)Health sciencesHealth services and systemsInformation and computing sciencesMachine learningDeep LearningKidney SegmentationFluid SegmentationRenal Ultrasound ImagingHydronephrosis Assessment<h3 dir="ltr">Objective</h3><p dir="ltr">To develop and evaluate a deep learning framework for automatic kidney and fluid segmentation in renal ultrasound images, aiming to enhance diagnostic accuracy and reduce variability in hydronephrosis assessment. </p><h3 dir="ltr">Methods</h3><p dir="ltr">A dataset of 1731 renal ultrasound images, annotated by four experienced urologists, was used for model training and evaluation. The proposed framework integrates a DenseNet201 backbone, Feature Pyramid Network (FPN), and Self-Organizing Neural Network (SelfONN) layers to enable multi-scale feature extraction and improve spatial precision. Several architectures were tested under identical conditions to ensure a fair comparison. Segmentation performance was assessed using standard metrics, including the Dice coefficient, precision, and recall. The framework also supported hydronephrosis classification using the fluid-to-kidney area ratio, with a threshold of 0.213 derived from prior literature. </p><h3 dir="ltr">Results</h3><p dir="ltr">The model achieved strong segmentation performance for kidneys (Dice: 0.92, precision: 0.93, recall: 0.91) and fluid regions (Dice: 0.89, precision: 0.90, recall: 0.88), outperforming baseline methods. The classification accuracy for detecting hydronephrosis reached 94%, based on the computed fluid-to-kidney ratio. Performance was consistent across varied image qualities, reflecting the robustness of the overall architecture. </p><h3 dir="ltr">Conclusion</h3><p dir="ltr">This study presents an automated, objective pipeline for analyzing renal ultrasound images. The proposed framework supports high segmentation accuracy and reliable classification, facilitating standardized and reproducible hydronephrosis assessment. Future work will focus on model optimization and incorporating explainable AI to enhance clinical integration.</p><h2 dir="ltr">Other Information</h2><p dir="ltr">Published in: Urology<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.urology.2025.08.005" target="_blank">https://dx.doi.org/10.1016/j.urology.2025.08.005</a></p>2025-12-09T03:00:00ZTextJournal contributioninfo:eu-repo/semantics/publishedVersiontextcontribution to journal10.1016/j.urology.2025.08.005https://figshare.com/articles/journal_contribution/Hybrid_Neural_Networks_for_Precise_Hydronephrosis_Classification_Using_Deep_Learning/30971659CC BY 4.0info:eu-repo/semantics/openAccessoai:figshare.com:article/309716592025-12-09T03:00:00Z
spellingShingle Hybrid Neural Networks for Precise Hydronephrosis Classification Using Deep Learning
Abdus Salam (1918387)
Health sciences
Health services and systems
Information and computing sciences
Machine learning
Deep Learning
Kidney Segmentation
Fluid Segmentation
Renal Ultrasound Imaging
Hydronephrosis Assessment
status_str publishedVersion
title Hybrid Neural Networks for Precise Hydronephrosis Classification Using Deep Learning
title_full Hybrid Neural Networks for Precise Hydronephrosis Classification Using Deep Learning
title_fullStr Hybrid Neural Networks for Precise Hydronephrosis Classification Using Deep Learning
title_full_unstemmed Hybrid Neural Networks for Precise Hydronephrosis Classification Using Deep Learning
title_short Hybrid Neural Networks for Precise Hydronephrosis Classification Using Deep Learning
title_sort Hybrid Neural Networks for Precise Hydronephrosis Classification Using Deep Learning
topic Health sciences
Health services and systems
Information and computing sciences
Machine learning
Deep Learning
Kidney Segmentation
Fluid Segmentation
Renal Ultrasound Imaging
Hydronephrosis Assessment