Colorectal cancer unmasked: A synergistic AI framework for Hyper-granular image dissection, precision segmentation, and automated diagnosis

Abstract Colorectal cancer (CRC) is the second most common cause of cancer-related mortality worldwide, underscoring the necessity for computer-aided diagnosis (CADx) systems that are interpretable, accurate, and robust. This study presents a practical CADx system that combines Vision Transformers (...

Full description

Saved in:
Bibliographic Details
Main Authors: Akella S. Narasimha Raju, K. Venkatesh, M. Rajababu, Ranjith Kumar Gatla, Shaik Jakeer Hussain, G. Satya Mohan Chowdary, T. Ganga Bhavani, Mohammed Kareemullah, Sameer Algburi, Ali Majdi, Ahmed M. Abdulhadi, Wahaj Ahmad Khan
Format: Article
Language:English
Published: BMC 2025-07-01
Series:BMC Medical Imaging
Subjects:
Online Access:https://doi.org/10.1186/s12880-025-01826-7
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849235625445687296
author Akella S. Narasimha Raju
K. Venkatesh
M. Rajababu
Ranjith Kumar Gatla
Shaik Jakeer Hussain
G. Satya Mohan Chowdary
T. Ganga Bhavani
Mohammed Kareemullah
Sameer Algburi
Ali Majdi
Ahmed M. Abdulhadi
Wahaj Ahmad Khan
author_facet Akella S. Narasimha Raju
K. Venkatesh
M. Rajababu
Ranjith Kumar Gatla
Shaik Jakeer Hussain
G. Satya Mohan Chowdary
T. Ganga Bhavani
Mohammed Kareemullah
Sameer Algburi
Ali Majdi
Ahmed M. Abdulhadi
Wahaj Ahmad Khan
author_sort Akella S. Narasimha Raju
collection DOAJ
description Abstract Colorectal cancer (CRC) is the second most common cause of cancer-related mortality worldwide, underscoring the necessity for computer-aided diagnosis (CADx) systems that are interpretable, accurate, and robust. This study presents a practical CADx system that combines Vision Transformers (ViTs) and DeepLabV3 + to accurately identify and segment colorectal lesions in colonoscopy images.The system addresses class balance and real-world complexity with PCA-based dimensionality reduction, data augmentation, and strategic preprocessing using recently curated CKHK-22 dataset comprising more than 14,000 annotated images of CVC-ClinicDB, Kvasir-2, and Hyper-Kvasir. ViT, ResNet-50, DenseNet-201, and VGG-16 were used to quantify classification performance. ViT achieved best-in-class accuracy (97%), F1-score (0.95), and AUC (92%) in test data. The DeepLabV3 + achieved segmentation state-of-the-art for tasks of localisation with 0.88 Dice Coefficient and 0.71 Intersection over Union (IoU), ensuring sharp delineation of areas that are malignant. The CADx system accommodates real-time inference and served through Google Cloud for information that accommodates scalable clinical implementation. The image-level segmentation effectiveness is evidenced by comparison with visual overlay and expert-manually deliminated masks, and its precision is illustrated by computation of precision, recall, F1-score, and AUC. The hybrid strategy not only outperforms traditional CNN strategies but also overcomes important clinical needs such as detection early, balance of highly disparate classes, and clear explanation. The proposed ViT–DeepLabV3 + system establishes a basis for advanced AI support to colorectal diagnosis by utilizing self-attention strategies and learning with different scales of context. The system offers a high-capacity, reproducible computerised colorectal cancer screening and monitoring solution and can be best deployed where resources are scarce, and it can be highly desirable for clinical deployment.
format Article
id doaj-art-85f79619672e461c8a78e34fc0709bb3
institution Kabale University
issn 1471-2342
language English
publishDate 2025-07-01
publisher BMC
record_format Article
series BMC Medical Imaging
spelling doaj-art-85f79619672e461c8a78e34fc0709bb32025-08-20T04:02:44ZengBMCBMC Medical Imaging1471-23422025-07-0125113510.1186/s12880-025-01826-7Colorectal cancer unmasked: A synergistic AI framework for Hyper-granular image dissection, precision segmentation, and automated diagnosisAkella S. Narasimha Raju0K. Venkatesh1M. Rajababu2Ranjith Kumar Gatla3Shaik Jakeer Hussain4G. Satya Mohan Chowdary5T. Ganga Bhavani6Mohammed Kareemullah7Sameer Algburi8Ali Majdi9Ahmed M. Abdulhadi10Wahaj Ahmad Khan11Department of Computing Technologies, School of Computing, SRM Institute of Science and TechnologyDepartment of Networking and Communications, School of Computing, SRM Institute of Science and TechnologyDepartment of Information Technology (IT), Aditya UniversityDepartment of Computer Science and Engineering (Data Science), Institute of Aeronautical EngineeringDepartment of Computer Science and Engineering (AI & ML), Institute of Aeronautical EngineeringDepartment of IT, Pragati Engineering College (A)Department of CS, Pragati Engineering College (A)Department of Mechanical Engineering, Graphic Era (Deemed to be University)Al-Kitab UniversityDepartment of Buildings and Construction Techniques Engineering, College of Engineering, Al-Mustaqbal UniversityAl-Safwa University CollegeInstitute of Technology, Dire-Dawa UniversityAbstract Colorectal cancer (CRC) is the second most common cause of cancer-related mortality worldwide, underscoring the necessity for computer-aided diagnosis (CADx) systems that are interpretable, accurate, and robust. This study presents a practical CADx system that combines Vision Transformers (ViTs) and DeepLabV3 + to accurately identify and segment colorectal lesions in colonoscopy images.The system addresses class balance and real-world complexity with PCA-based dimensionality reduction, data augmentation, and strategic preprocessing using recently curated CKHK-22 dataset comprising more than 14,000 annotated images of CVC-ClinicDB, Kvasir-2, and Hyper-Kvasir. ViT, ResNet-50, DenseNet-201, and VGG-16 were used to quantify classification performance. ViT achieved best-in-class accuracy (97%), F1-score (0.95), and AUC (92%) in test data. The DeepLabV3 + achieved segmentation state-of-the-art for tasks of localisation with 0.88 Dice Coefficient and 0.71 Intersection over Union (IoU), ensuring sharp delineation of areas that are malignant. The CADx system accommodates real-time inference and served through Google Cloud for information that accommodates scalable clinical implementation. The image-level segmentation effectiveness is evidenced by comparison with visual overlay and expert-manually deliminated masks, and its precision is illustrated by computation of precision, recall, F1-score, and AUC. The hybrid strategy not only outperforms traditional CNN strategies but also overcomes important clinical needs such as detection early, balance of highly disparate classes, and clear explanation. The proposed ViT–DeepLabV3 + system establishes a basis for advanced AI support to colorectal diagnosis by utilizing self-attention strategies and learning with different scales of context. The system offers a high-capacity, reproducible computerised colorectal cancer screening and monitoring solution and can be best deployed where resources are scarce, and it can be highly desirable for clinical deployment.https://doi.org/10.1186/s12880-025-01826-7Diagnosis of colorectal cancerVision transformersHyper-granular image analysisDeepLabV3+Automatic segmentation
spellingShingle Akella S. Narasimha Raju
K. Venkatesh
M. Rajababu
Ranjith Kumar Gatla
Shaik Jakeer Hussain
G. Satya Mohan Chowdary
T. Ganga Bhavani
Mohammed Kareemullah
Sameer Algburi
Ali Majdi
Ahmed M. Abdulhadi
Wahaj Ahmad Khan
Colorectal cancer unmasked: A synergistic AI framework for Hyper-granular image dissection, precision segmentation, and automated diagnosis
BMC Medical Imaging
Diagnosis of colorectal cancer
Vision transformers
Hyper-granular image analysis
DeepLabV3+
Automatic segmentation
title Colorectal cancer unmasked: A synergistic AI framework for Hyper-granular image dissection, precision segmentation, and automated diagnosis
title_full Colorectal cancer unmasked: A synergistic AI framework for Hyper-granular image dissection, precision segmentation, and automated diagnosis
title_fullStr Colorectal cancer unmasked: A synergistic AI framework for Hyper-granular image dissection, precision segmentation, and automated diagnosis
title_full_unstemmed Colorectal cancer unmasked: A synergistic AI framework for Hyper-granular image dissection, precision segmentation, and automated diagnosis
title_short Colorectal cancer unmasked: A synergistic AI framework for Hyper-granular image dissection, precision segmentation, and automated diagnosis
title_sort colorectal cancer unmasked a synergistic ai framework for hyper granular image dissection precision segmentation and automated diagnosis
topic Diagnosis of colorectal cancer
Vision transformers
Hyper-granular image analysis
DeepLabV3+
Automatic segmentation
url https://doi.org/10.1186/s12880-025-01826-7
work_keys_str_mv AT akellasnarasimharaju colorectalcancerunmaskedasynergisticaiframeworkforhypergranularimagedissectionprecisionsegmentationandautomateddiagnosis
AT kvenkatesh colorectalcancerunmaskedasynergisticaiframeworkforhypergranularimagedissectionprecisionsegmentationandautomateddiagnosis
AT mrajababu colorectalcancerunmaskedasynergisticaiframeworkforhypergranularimagedissectionprecisionsegmentationandautomateddiagnosis
AT ranjithkumargatla colorectalcancerunmaskedasynergisticaiframeworkforhypergranularimagedissectionprecisionsegmentationandautomateddiagnosis
AT shaikjakeerhussain colorectalcancerunmaskedasynergisticaiframeworkforhypergranularimagedissectionprecisionsegmentationandautomateddiagnosis
AT gsatyamohanchowdary colorectalcancerunmaskedasynergisticaiframeworkforhypergranularimagedissectionprecisionsegmentationandautomateddiagnosis
AT tgangabhavani colorectalcancerunmaskedasynergisticaiframeworkforhypergranularimagedissectionprecisionsegmentationandautomateddiagnosis
AT mohammedkareemullah colorectalcancerunmaskedasynergisticaiframeworkforhypergranularimagedissectionprecisionsegmentationandautomateddiagnosis
AT sameeralgburi colorectalcancerunmaskedasynergisticaiframeworkforhypergranularimagedissectionprecisionsegmentationandautomateddiagnosis
AT alimajdi colorectalcancerunmaskedasynergisticaiframeworkforhypergranularimagedissectionprecisionsegmentationandautomateddiagnosis
AT ahmedmabdulhadi colorectalcancerunmaskedasynergisticaiframeworkforhypergranularimagedissectionprecisionsegmentationandautomateddiagnosis
AT wahajahmadkhan colorectalcancerunmaskedasynergisticaiframeworkforhypergranularimagedissectionprecisionsegmentationandautomateddiagnosis