Adapting Classification Neural Network Architectures for Medical Image Segmentation Using Explainable AI

Segmentation neural networks are widely used in medical imaging to identify anomalies that may impact patient health. Despite their effectiveness, these networks face significant challenges, including the need for extensive annotated patient data, time-consuming manual segmentation processes and res...

Full description

Saved in:
Bibliographic Details
Main Authors: Arturs Nikulins, Edgars Edelmers, Kaspars Sudars, Inese Polaka
Format: Article
Language:English
Published: MDPI AG 2025-02-01
Series:Journal of Imaging
Subjects:
Online Access:https://www.mdpi.com/2313-433X/11/2/55
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849718444622086144
author Arturs Nikulins
Edgars Edelmers
Kaspars Sudars
Inese Polaka
author_facet Arturs Nikulins
Edgars Edelmers
Kaspars Sudars
Inese Polaka
author_sort Arturs Nikulins
collection DOAJ
description Segmentation neural networks are widely used in medical imaging to identify anomalies that may impact patient health. Despite their effectiveness, these networks face significant challenges, including the need for extensive annotated patient data, time-consuming manual segmentation processes and restricted data access due to privacy concerns. In contrast, classification neural networks, similar to segmentation neural networks, capture essential parameters for identifying objects during training. This paper leverages this characteristic, combined with explainable artificial intelligence (XAI) techniques, to address the challenges of segmentation. By adapting classification neural networks for segmentation tasks, the proposed approach reduces dependency on manual segmentation. To demonstrate this concept, the Medical Segmentation Decathlon ‘Brain Tumours’ dataset was utilised. A ResNet classification neural network was trained, and XAI tools were applied to generate segmentation-like outputs. Our findings reveal that GuidedBackprop is among the most efficient and effective methods, producing heatmaps that closely resemble segmentation masks by accurately highlighting the entirety of the target object.
format Article
id doaj-art-ffee84e99d7f4962bff7ebc677914e8e
institution DOAJ
issn 2313-433X
language English
publishDate 2025-02-01
publisher MDPI AG
record_format Article
series Journal of Imaging
spelling doaj-art-ffee84e99d7f4962bff7ebc677914e8e2025-08-20T03:12:22ZengMDPI AGJournal of Imaging2313-433X2025-02-011125510.3390/jimaging11020055Adapting Classification Neural Network Architectures for Medical Image Segmentation Using Explainable AIArturs Nikulins0Edgars Edelmers1Kaspars Sudars2Inese Polaka3Faculty of Computer Science, Information Technology and Energy, Riga Technical University, LV-1048 Riga, LatviaFaculty of Computer Science, Information Technology and Energy, Riga Technical University, LV-1048 Riga, LatviaInstitute of Electronics and Computer Science, LV-1006 Riga, LatviaFaculty of Computer Science, Information Technology and Energy, Riga Technical University, LV-1048 Riga, LatviaSegmentation neural networks are widely used in medical imaging to identify anomalies that may impact patient health. Despite their effectiveness, these networks face significant challenges, including the need for extensive annotated patient data, time-consuming manual segmentation processes and restricted data access due to privacy concerns. In contrast, classification neural networks, similar to segmentation neural networks, capture essential parameters for identifying objects during training. This paper leverages this characteristic, combined with explainable artificial intelligence (XAI) techniques, to address the challenges of segmentation. By adapting classification neural networks for segmentation tasks, the proposed approach reduces dependency on manual segmentation. To demonstrate this concept, the Medical Segmentation Decathlon ‘Brain Tumours’ dataset was utilised. A ResNet classification neural network was trained, and XAI tools were applied to generate segmentation-like outputs. Our findings reveal that GuidedBackprop is among the most efficient and effective methods, producing heatmaps that closely resemble segmentation masks by accurately highlighting the entirety of the target object.https://www.mdpi.com/2313-433X/11/2/55medical imagingclassification modelsimage segmentationexplainable artificial intelligenceneural networks
spellingShingle Arturs Nikulins
Edgars Edelmers
Kaspars Sudars
Inese Polaka
Adapting Classification Neural Network Architectures for Medical Image Segmentation Using Explainable AI
Journal of Imaging
medical imaging
classification models
image segmentation
explainable artificial intelligence
neural networks
title Adapting Classification Neural Network Architectures for Medical Image Segmentation Using Explainable AI
title_full Adapting Classification Neural Network Architectures for Medical Image Segmentation Using Explainable AI
title_fullStr Adapting Classification Neural Network Architectures for Medical Image Segmentation Using Explainable AI
title_full_unstemmed Adapting Classification Neural Network Architectures for Medical Image Segmentation Using Explainable AI
title_short Adapting Classification Neural Network Architectures for Medical Image Segmentation Using Explainable AI
title_sort adapting classification neural network architectures for medical image segmentation using explainable ai
topic medical imaging
classification models
image segmentation
explainable artificial intelligence
neural networks
url https://www.mdpi.com/2313-433X/11/2/55
work_keys_str_mv AT artursnikulins adaptingclassificationneuralnetworkarchitecturesformedicalimagesegmentationusingexplainableai
AT edgarsedelmers adaptingclassificationneuralnetworkarchitecturesformedicalimagesegmentationusingexplainableai
AT kasparssudars adaptingclassificationneuralnetworkarchitecturesformedicalimagesegmentationusingexplainableai
AT inesepolaka adaptingclassificationneuralnetworkarchitecturesformedicalimagesegmentationusingexplainableai