ISLES'24: Improving final infarct prediction in ischemic stroke using multimodal imaging and clinical data

2024-08-20

arXiv.org

AI Summary

This study addresses the clinical challenge of inaccurate prediction of final infarct volume following acute ischemic stroke—a limitation impeding critical decisions including patient transfer, endovascular thrombectomy eligibility, and prognostic assessment. To this end, we propose the first multimodal prediction framework integrating acute-phase non-contrast CT, subacute-phase MRI, and heterogeneous clinical tabular data. Methodologically, we introduce a standardized, clinically realistic benchmark challenge encompassing medical image segmentation (CT/MRI), heterogeneous clinical data fusion, and a unified evaluation protocol. Our contributions are threefold: (1) the first publicly available, fully reproducible, and comprehensive multimodal platform for infarct evolution prediction; (2) significantly improved accuracy in final infarct volume estimation, enabling precise core and penumbra delineation; and (3) quantitative support for inter-hospital transfer decisions, reperfusion therapy optimization, and individualized outcome prediction.

Technology Category

• Machine Learning: Multimodal Learning
• Computer Vision: Multi-modal Vision
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Search and Retrieval-Augmented AI: Retrieval-Augmented Generation (RAG) and multi-modal RAG
• Web Mining and Content Analysis: Mining multimedia, multimodal, multilingual, cross-lingual Web data

Abstract

Accurate estimation of core (irreversibly damaged tissue) and penumbra (salvageable tissue) volumes is essential for ischemic stroke treatment decisions. Perfusion CT, the clinical standard, estimates these volumes but is affected by variations in deconvolution algorithms, implementations, and thresholds. Core tissue expands over time, with growth rates influenced by thrombus location, collateral circulation, and inherent patient-specific factors. Understanding this tissue growth is crucial for determining the need to transfer patients to comprehensive stroke centers, predicting the benefits of additional reperfusion attempts during mechanical thrombectomy, and forecasting final clinical outcomes. This work presents the ISLES'24 challenge, which addresses final post-treatment stroke infarct prediction from pre-interventional acute stroke imaging and clinical data. ISLES'24 establishes a unique 360-degree setting where all feasibly accessible clinical data are available for participants, including full CT acute stroke imaging, sub-acute follow-up MRI, and clinical tabular data. The contributions of this work are two-fold: first, we introduce a standardized benchmarking of final stroke infarct segmentation algorithms through the ISLES'24 challenge; second, we provide insights into infarct segmentation using multimodal imaging and clinical data strategies by identifying outperforming methods on a finely curated dataset. The outputs of this challenge are anticipated to enhance clinical decision-making and improve patient outcome predictions. All ISLES'24 materials, including data, performance evaluation scripts, and leading algorithmic strategies, are available to the research community following url{https://isles-24.grand-challenge.org/}.

Problem

• Predict final brain infarct volumes using multimodal imaging and clinical data
• Improve accuracy for guiding acute ischemic stroke treatment decisions
• Establish standardized benchmark for evaluating infarct prediction models

Innovation

• Multimodal nnU-Net-based architecture for infarct prediction
• Utilizes acute CT and sub-acute MRI imaging data
• Integrates structured clinical information for analysis