A Robust Ensemble Algorithm for Ischemic Stroke Lesion Segmentation: Generalizability and Clinical Utility Beyond the ISLES Challenge
Abstract
Diffusion-weighted MRI (DWI) is essential for stroke diagnosis, treatment decisions, and prognosis. However, image and disease variability hinder the development of generalizable AI algorithms with clinical value. We address this gap by presenting a novel ensemble algorithm derived from the 2022 Ischemic Stroke Lesion Segmentation (ISLES) challenge. ISLES'22 provided 400 patient scans with ischemic stroke from various medical centers, facilitating the development of a wide range of cutting-edge segmentation algorithms by the research community. Through collaboration with leading teams, we combined top-performing algorithms into an ensemble model that overcomes the limitations of individual solutions. Our ensemble model achieved superior ischemic lesion detection and segmentation accuracy on our internal test set compared to individual algorithms. This accuracy generalized well across diverse image and disease variables. Furthermore, the model excelled in extracting clinical biomarkers. Notably, in a Turing-like test, neuroradiologists consistently preferred the algorithm's segmentations over manual expert efforts, highlighting increased comprehensiveness and precision. Validation using a real-world external dataset (N=1686) confirmed the model's generalizability. The algorithm's outputs also demonstrated strong correlations with clinical scores (admission NIHSS and 90-day mRS) on par with or exceeding expert-derived results, underlining its clinical relevance. This study offers two key findings. First, we present an ensemble algorithm (https://github.com/Tabrisrei/ISLES22_Ensemble) that detects and segments ischemic stroke lesions on DWI across diverse scenarios on par with expert (neuro)radiologists. Second, we show the potential for biomedical challenge outputs to extend beyond the challenge's initial objectives, demonstrating their real-world clinical applicability.
Cite
@article{arxiv.2403.19425,
title = {A Robust Ensemble Algorithm for Ischemic Stroke Lesion Segmentation: Generalizability and Clinical Utility Beyond the ISLES Challenge},
author = {Ezequiel de la Rosa and Mauricio Reyes and Sook-Lei Liew and Alexandre Hutton and Roland Wiest and Johannes Kaesmacher and Uta Hanning and Arsany Hakim and Richard Zubal and Waldo Valenzuela and David Robben and Diana M. Sima and Vincenzo Anania and Arne Brys and James A. Meakin and Anne Mickan and Gabriel Broocks and Christian Heitkamp and Shengbo Gao and Kongming Liang and Ziji Zhang and Md Mahfuzur Rahman Siddiquee and Andriy Myronenko and Pooya Ashtari and Sabine Van Huffel and Hyun-su Jeong and Chi-ho Yoon and Chulhong Kim and Jiayu Huo and Sebastien Ourselin and Rachel Sparks and Albert Clèrigues and Arnau Oliver and Xavier Lladó and Liam Chalcroft and Ioannis Pappas and Jeroen Bertels and Ewout Heylen and Juliette Moreau and Nima Hatami and Carole Frindel and Abdul Qayyum and Moona Mazher and Domenec Puig and Shao-Chieh Lin and Chun-Jung Juan and Tianxi Hu and Lyndon Boone and Maged Goubran and Yi-Jui Liu and Susanne Wegener and Florian Kofler and Ivan Ezhov and Suprosanna Shit and Moritz R. Hernandez Petzsche and Bjoern Menze and Jan S. Kirschke and Benedikt Wiestler},
journal= {arXiv preprint arXiv:2403.19425},
year = {2025}
}