---
title: "Graph-based End-to-End Deep Learning for CT Multitask Stroke Diagnosis"
id: "medrxiv-11-clinically-generalisable-end-to-end-graph-learning-for-ct-image-based-multitask"
canonical_url: "https://medichelpline.com/clinical-feed/medrxiv-11-clinically-generalisable-end-to-end-graph-learning-for-ct-image-based-multitask"
content_type: "clinical_feed_article"
specialty: "Radiology"
source_name: "medRxiv (Clinical Preprints)"
source_url: "https://www.medrxiv.org/content/10.64898/2026.08.26.26360026v1?rss=1"
published_at: "2026-08-31T12:00:00.000Z"
evidence_level: "Verified Feed"
license: "CC-BY-NC-4.0 / Informational Use"
---
# Graph-based End-to-End Deep Learning for CT Multitask Stroke Diagnosis
## Provenance & Clinical Metadata
- **Canonical URL:** https://medichelpline.com/clinical-feed/medrxiv-11-clinically-generalisable-end-to-end-graph-learning-for-ct-image-based-multitask
- **Specialty:** [Radiology](https://medichelpline.com/clinical-feed/radiology.md)
- **Primary Source:** medRxiv (Clinical Preprints)
- **Source URL:** [Original Journal Publication](https://www.medrxiv.org/content/10.64898/2026.08.26.26360026v1?rss=1)
- **Published At:** 2026-08-31T12:00:00.000Z
- **Evidence Rating:** Verified Feed
## Executive GIST (TL;DR)
- Stroke remains a leading cause of death and long-term disability; rapid diagnosis is limited by shortages of trained radiologists, especially in resource-constrained settings. - The authors developed the Intelligent Integrated Stroke Diagnosis System (IISDS), an end-to-end deep learning framework built on **StrokeGNN**, a graph-based architecture that combines 3D contextual feature extraction with U-Net–based 2D lesion segmentation for non-contrast CT scans. - IISDS performs three coordinated tasks in a single pipeline: **stroke subtype classification**, **lesion segmentation**, and **lesion volume estimation**. - The team collected and curated a multi-centre dataset named **BGD-ISD**, comprising 1,507 CT scans from 597 stroke cases across six hospitals and medical centres in Bangladesh, developed through collaboration among AI researchers, neurologists, radiologists and clinicians. - Across BGD-ISD and multiple publicly available datasets, IISDS achieved improvements over existing approaches: ≥0.011 increase in **Dice** score for segmentation, ≥0.3 reduction in average symmetric surface distance (**ASSD**) for lesion volume estimation error, and ≥0.018 increase in classification **AUC**. - The results support the potential of graph-based deep learning to enable clinically generalisable, automated, and scalable **CT**-based stroke diagnosis to aid rapid clinical decision-making in settings with limited radiological expertise. - Ethical approvals were obtained from Southeast University, Bangladesh and International Medical College, Bangladesh; authors declared no competing interests. - Data and code are planned to be made available at the project GitHub repository upon paper acceptance, per the authors’ statement.
## Clinical Analysis & Structured Key Points
Clinically Generalisable End-to-End Graph Learning for CT Image-Based Multitask Stroke Diagnosis | medRxiv Skip to main content Clinically Generalisable End-to-End Graph Learning for CT Image-Based Multitask Stroke Diagnosis View ORCID Profile Zhicheng Lu , View ORCID Profile Shahadat Uddin , View ORCID Profile Sergio Uribe , Sam White , View ORCID Profile Rodrigo Tomazini Martins , View ORCID Profile Shayne Chau , Abu Syed Md. Mosaddek , Md. Siddiqul Islam , Nusratun Nahar , View ORCID Profile A.K.M. Azad , K. M. Nazmul Hossain , Habib Sadat Choudhury , K. M. Rakibul Hasan , Nabil Mosaddek , Samia Rahman , Md. Mostaque Hossain , K. M. Mehedi Hasan Sizar , View ORCID Profile Claudio Angione , View ORCID Profile Pietro Lio , Md Tauhidul Islam , View ORCID Profile Mohammad Ali Moni doi: https://doi.org/10.64898/2026.08.26.26360026 Zhicheng Lu 1 Charles Sturt University; Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Zhicheng Lu Shahadat Uddin 2 The University of Sydney; Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Shahadat Uddin Sergio Uribe 3 Monash University; Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Sergio Uribe Sam White 4 Adelaide University; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Rodrigo Tomazini Martins 5 Mater Misericordiae Hospital; Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Rodrigo Tomazini Martins Shayne Chau 1 Charles Sturt University; Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Shayne Chau Abu Syed Md. Mosaddek 6 Uttara Adhunik Medical College; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Md. Siddiqul Islam 7 American International University-Bangladesh; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Nusratun Nahar 8 Southeast University; Find this author on Google Scholar Find this author on PubMed Search for this author on this site A.K.M. Azad 9 Imam Mohammad Ibn Saud Islamic University; Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for A.K.M. Azad K. M. Nazmul Hossain 10 Jahangirnagar University; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Habib Sadat Choudhury 11 International Medical College; Find this author on Google Scholar Find this author on PubMed Search for this author on this site K. M. Rakibul Hasan 12 Paediatric Neuro Care & Research (PNR); Find this author on Google Scholar Find this author on PubMed Search for this author on this site Nabil Mosaddek 13 Military Institute of Science and Technology (MIST); Find this author on Google Scholar Find this author on PubMed Search for this author on this site Samia Rahman 14 Popular Diagnostic Centre; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Md. Mostaque Hossain 11 International Medical College; Find this author on Google Scholar Find this author on PubMed Search for this author on this site K. M. Mehedi Hasan Sizar 15 IbnaSina Diagnostic & Imaging Center; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Claudio Angione 16 Teesside University; Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Claudio Angione Pietro Lio 17 University of Cambdridge; Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Pietro Lio Md Tauhidul Islam 18 Stanford University; Find this author on Google Scholar Find this author on PubMed Search for this author on this site Mohammad Ali Moni 19 University of Queensland Find this author on Google Scholar Find this author on PubMed Search for this author on this site ORCID record for Mohammad Ali Moni For correspondence: m.moni{at}uq.edu.au Abstract Info/History Metrics Supplementary material Data/Code Preview PDF Abstract Stroke remains a leading cause of mortality and long-term disability worldwide, yet rapid diagnosis is often limited by the shortage of trained radiologists, particularly in resource-constrained settings. Automated analysis of CT imaging offers a potential solution, but existing methods often struggle to achieve clinically generalisable performance while jointly addressing multiple diagnostic tasks. Here we present the Intelligent Integrated Stroke Diagnosis System IISDS, an end-to-end deep learning framework built upon StrokeGNN, a graph-based architecture that integrates 3D contextual feature extraction with U-Net-based 2D lesion segmentation to enable comprehensive stroke analysis from non-contrast CT scans. IISDS performs stroke subtype classification, lesion segmentation and lesion volume estimation within a unified pipeline. To develop and validate the system, we collected and curated BGD-ISD through a collaboration between AI researchers, neurologists, radiologists and clinicians, resulting in a large multi-centre dataset comprising 1,507 CT scans from 597 stroke cases acquired across six hospitals and medical centres in Bangladesh. Across BGD-ISD and multiple publicly available datasets, IISDS achieves state-of-the-art performance on all tasks, improving segmentation accuracy by ≥0.011 Dice score, reducing lesion volume estimation error by ≥0.3 average symmetric surface distance (ASSD), and increasing classification performance by ≥0.018 area under the receiver operating characteristic curve (AUC) compared with existing approaches. These results demonstrate the potential of graph-based deep learning to enable clinically generalisable, automated and scalable stroke diagnosis from CT imaging, supporting rapid clinical decision-making, particularly in healthcare environments with limited access to expert radiological interpretation. Competing Interest Statement The authors have declared no competing interest. Author Declarations I confirm all relevant ethical guidelines have been followed, and any necessary IRB and/or ethics committee approvals have been obtained. Yes The details of the IRB/oversight body that provided approval or exemption for the research described are given below: Southeast University, Bangladesh gave ethical approval for this work. International Medical College, Bangladesh gave ethical approval for this work. I confirm that all necessary patient/participant consent has been obtained and the appropriate institutional forms have been archived, and that any patient/participant/sample identifiers included were not known to anyone (e.g., hospital staff, patients or participants themselves) outside the research group so cannot be used to identify individuals. Yes I understand that all clinical trials and any other prospective interventional studies must be registered with an ICMJE-approved registry, such as ClinicalTrials.gov. I confirm that any such study reported in the manuscript has been registered and the trial registration ID is provided (note: if posting a prospective study registered retrospectively, please provide a statement in the trial ID field explaining why the study was not registered in advance). Yes I have followed all appropriate research reporting guidelines, such as any relevant EQUATOR Network research reporting checklist(s) and other pertinent material, if applicable. Yes Data Availability All data produced will be available online at https://github.com/Zhicheng-Lu/stroke_ct upon the acceptance of the paper. Funder Information Declared Australian Commonwealth Funding Copyright The copyright holder for this preprint is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. It is made available under a CC-BY-NC-ND 4.0 International license . Back to top Previous Next Posted August 31, 2026. Download PDF Supplementary Material Data/Code Email Thank you for your interest in spreading the word about medRxiv. NOTE: Your email address is requested solely to identify you as the sender of this article. Your Email * Your Name * Send To * Enter multiple addresses on separate lines or separate them with commas. You are going to email the following Clinically Generalisable End-to-End Graph Learning for CT Image-Based Multitask Stroke Diagnosis Message Subject (Your Name) has forwarded a page to you from medRxiv Message Body (Your Name) thought you would like to see this page from the medRxiv website. Your Personal Message CAPTCHA This question is for testing whether or not you are a human visitor and to prevent automated spam submissions. Share Clinically Generalisable End-to-End Graph Learning for CT Image-Based Multitask Stroke Diagnosis Zhicheng Lu , Shahadat Uddin , Sergio Uribe , Sam White , Rodrigo Tomazini Martins , Shayne Chau , Abu Syed Md. Mosaddek , Md. Siddiqul Islam , Nusratun Nahar , A.K.M. Azad , K. M. Nazmul Hossain , Habib Sadat Choudhury , K. M. Rakibul Hasan , Nabil Mosaddek , Samia Rahman , Md. Mostaque Hossain , K. M. Mehedi Hasan Sizar , Claudio Angione , Pietro Lio , Md Tauhidul Islam , Mohammad Ali Moni medRxiv 2026.08.26.26360026; doi: https://doi.org/10.64898/2026.08.26.26360026 Share This Article: Copy Citation Tools Clinically Generalisable End-to-End Graph Learning for CT Image-Based Multitask Stroke Diagnosis Zhicheng Lu , Shahadat Uddin , Sergio Uribe , Sam White , Rodrigo Tomazini Martins , Shayne Chau , Abu Syed Md. Mosaddek , Md. Siddiqul Islam , Nusratun Nahar , A.K.M. Azad , K. M. Nazmul Hossain , Habib Sadat Choudhury , K. M. Rakibul Hasan , Nabil Mosaddek , Samia Rahman , Md. Mostaque Hossain , K. M. Mehedi Hasan Sizar , Claudio Angione , Pietro Lio , Md Tauhidul Islam , Mohammad Ali Moni medRxiv 2026.08.26.26360026; doi: https://doi.org/10.64898/2026.08.26.26360026 Citation Manager Formats BibTeX Bookends EasyBib EndNote (tagged) EndNote 8 (xml) Medlars Mendeley Papers RefWorks Tagged Ref Manager RIS Zotero Tweet Widget Facebook Like Google Plus One Subject Area Radiology and Imaging Subject Areas All Articles Addiction Medicine (613) Allergy and Immunology (895) Anesthesia (325) Cardiovascular Medicine (4765) Dentistry and Oral Medicine (474) Dermatology (406) Emergency Medicine (641) Endocrinology (including Diabetes Mellitus and Metabolic Disease) (1612) Epidemiology (15825) Forensic Medicine (32) Gastroenterology (1194) Genetic and Genomic Medicine (6979) Geriatric Medicine (727) Health Economics (1056) Health Informatics (4968) Health Policy (1424) Health Systems and Quality Improvement (1733) Hematology (579) HIV/AIDS (1333) Infectious Diseases (except HIV/AIDS) (16239) Intensive Care and Critical Care Medicine (1160) Medical Education (663) Medical Ethics (150) Nephrology (720) Neurology (7184) Nursing (363) Nutrition (1072) Obstetrics and Gynecology (1230) Occupational and Environmental Health (999) Oncology (3567) Ophthalmology (1035) Orthopedics (393) Otolaryngology (446) Pain Medicine (464) Palliative Medicine (139) Pathology (702) Pediatrics (1772) Pharmacology and Therapeutics (732) Primary Care Research (758) Psychiatry and Clinical Psychology (5827) Public and Global Health (9678) Radiology and Imaging (2384) Rehabilitation Medicine and Physical Therapy (1431) Respiratory Medicine (1239) Rheumatology (633) Sexual and Reproductive Health (765) Sports Medicine (574) Surgery (765) Toxicology (107) Transplantation (302) Urology (285)
## Related Clinical Research

- [SISYPHUS study protocol: 3D rotational angiography to assess final implant quality in carotid sten](https://medichelpline.com/clinical-feed/medrxiv-8-systematic-study-of-patients-with-atherosclerotic-neurovascular-stenoses.md)
- [Spatially Context-Aware Transformers (SpyCAT) for Anomaly Detection of Subtle Brain MRI Lesions](https://medichelpline.com/clinical-feed/medrxiv-16-spatially-context-aware-transformers-facilitate-modeling-based-anomaly.md)
- [Participation in a low‑intensity health assessment program linked to lower diabetes risk in older](https://medichelpline.com/clinical-feed/plos-medicine-0-association-of-a-multiple-risk-factor-assessment-and-intervention-program-with.md)
- [Cherish Your Heartbeat: AHA launches Charleston campaign to reduce heart disease risk](https://medichelpline.com/clinical-feed/aha-news-1-take-control-of-your-heart-health-new-charleston-campaign-encourages-action.md)
- [Predicting Cardiovascular Autonomic Neuropathy in Type 2 Diabetes: Risk Factors and Nomogram Model](https://medichelpline.com/clinical-feed/pubmed-42763210.md) (DOI: 10.3760/cma.j.cn112137-20260410-00979)

## Navigation
- [← Back to Radiology Feed](https://medichelpline.com/clinical-feed/radiology.md)
- [← All Clinical Specialties](https://medichelpline.com/clinical-feed.md)
## Medical & Regulatory Disclaimer

> [!CAUTION]
> MedicHelpline content is structured for research, educational, and professional discovery purposes. It does not constitute individual medical advice, clinical diagnosis, or treatment recommendations.
> Always verify dosing, contraindications, and regulatory alerts against official product labeling and primary regulatory sources before clinical decision-making.