Bladder lesion detection using EfficientNet and hybrid attention transformer through attention transformation

Abstract Bladder cancer diagnosis is a challenging task because of its intricacy and variation of tumor features. Moreover, morphological similarities of the cancerous cells make manual diagnosis time-consuming. Recently, machine learning and deep learning methods have been utilized to diagnose blad...

Full description

Saved in:
Bibliographic Details
Main Authors: Poonam Sharma, Bhisham Sharma, Dhirendra Prasad Yadav, Deepti Thakral, Julian L. Webber
Format: Article
Language:English
Published: Nature Portfolio 2025-05-01
Series:Scientific Reports
Subjects:
Online Access:https://doi.org/10.1038/s41598-025-02767-5
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849731804270952448
author Poonam Sharma
Bhisham Sharma
Dhirendra Prasad Yadav
Deepti Thakral
Julian L. Webber
author_facet Poonam Sharma
Bhisham Sharma
Dhirendra Prasad Yadav
Deepti Thakral
Julian L. Webber
author_sort Poonam Sharma
collection DOAJ
description Abstract Bladder cancer diagnosis is a challenging task because of its intricacy and variation of tumor features. Moreover, morphological similarities of the cancerous cells make manual diagnosis time-consuming. Recently, machine learning and deep learning methods have been utilized to diagnose bladder cancer. However, manual feature requirements for machine learning and the high volume of data for deep learning make them less reliable for real-time application. This study developed a hybrid model using CNN (Convolutional Neural Network) and less attention-based ViT (Vision Transformer) for bladder lesion diagnosis. Our hybrid model contains two blocks of the inceptionV3 to extract spatial features. Furthermore, the global co-relation of the features is achieved using hybrid attention modules incorporated in the ViT encoder. The experimental evaluation of the model on a dataset consisting of 17,540 endoscopic images achieved an average accuracy, precision and F1-score of 97.73%, 97.21% and 96.86%, respectively, using a 5-fold cross-validation strategy. We compared the results of the proposed method with CNN and ViT-based methods under the same experimental condition, and we achieved much better performance than our counterparts.
format Article
id doaj-art-94425f4551684df7b8a41754f3c30fd5
institution DOAJ
issn 2045-2322
language English
publishDate 2025-05-01
publisher Nature Portfolio
record_format Article
series Scientific Reports
spelling doaj-art-94425f4551684df7b8a41754f3c30fd52025-08-20T03:08:25ZengNature PortfolioScientific Reports2045-23222025-05-0115111710.1038/s41598-025-02767-5Bladder lesion detection using EfficientNet and hybrid attention transformer through attention transformationPoonam Sharma0Bhisham Sharma1Dhirendra Prasad Yadav2Deepti Thakral3Julian L. Webber4Chitkara University Institute of Engineering and Technology, Chitkara UniversityCentre of Research Impact and Outcome, Chitkara UniversityDepartment of Computer Engineering & Applications, G.L.A. UniversityDepartment of Computer Science and Technology, Manav Rachna UniversityDepartment of Electronics and Communication Engineering, Kuwait College of Science and Technology (KCST)Abstract Bladder cancer diagnosis is a challenging task because of its intricacy and variation of tumor features. Moreover, morphological similarities of the cancerous cells make manual diagnosis time-consuming. Recently, machine learning and deep learning methods have been utilized to diagnose bladder cancer. However, manual feature requirements for machine learning and the high volume of data for deep learning make them less reliable for real-time application. This study developed a hybrid model using CNN (Convolutional Neural Network) and less attention-based ViT (Vision Transformer) for bladder lesion diagnosis. Our hybrid model contains two blocks of the inceptionV3 to extract spatial features. Furthermore, the global co-relation of the features is achieved using hybrid attention modules incorporated in the ViT encoder. The experimental evaluation of the model on a dataset consisting of 17,540 endoscopic images achieved an average accuracy, precision and F1-score of 97.73%, 97.21% and 96.86%, respectively, using a 5-fold cross-validation strategy. We compared the results of the proposed method with CNN and ViT-based methods under the same experimental condition, and we achieved much better performance than our counterparts.https://doi.org/10.1038/s41598-025-02767-5Bladder CancerDeep learningTransformer encoderSelf-AttentionHybrid attentionVision transformer
spellingShingle Poonam Sharma
Bhisham Sharma
Dhirendra Prasad Yadav
Deepti Thakral
Julian L. Webber
Bladder lesion detection using EfficientNet and hybrid attention transformer through attention transformation
Scientific Reports
Bladder Cancer
Deep learning
Transformer encoder
Self-Attention
Hybrid attention
Vision transformer
title Bladder lesion detection using EfficientNet and hybrid attention transformer through attention transformation
title_full Bladder lesion detection using EfficientNet and hybrid attention transformer through attention transformation
title_fullStr Bladder lesion detection using EfficientNet and hybrid attention transformer through attention transformation
title_full_unstemmed Bladder lesion detection using EfficientNet and hybrid attention transformer through attention transformation
title_short Bladder lesion detection using EfficientNet and hybrid attention transformer through attention transformation
title_sort bladder lesion detection using efficientnet and hybrid attention transformer through attention transformation
topic Bladder Cancer
Deep learning
Transformer encoder
Self-Attention
Hybrid attention
Vision transformer
url https://doi.org/10.1038/s41598-025-02767-5
work_keys_str_mv AT poonamsharma bladderlesiondetectionusingefficientnetandhybridattentiontransformerthroughattentiontransformation
AT bhishamsharma bladderlesiondetectionusingefficientnetandhybridattentiontransformerthroughattentiontransformation
AT dhirendraprasadyadav bladderlesiondetectionusingefficientnetandhybridattentiontransformerthroughattentiontransformation
AT deeptithakral bladderlesiondetectionusingefficientnetandhybridattentiontransformerthroughattentiontransformation
AT julianlwebber bladderlesiondetectionusingefficientnetandhybridattentiontransformerthroughattentiontransformation