Cape Town - 2026 ISMRM-ISMRT Annual Meeting and Exhibition - Decoding Stroke with Imaging and AI: New Frontiers in Cerebrovascular Research

Decoding Stroke with Imaging and AI: New Frontiers in Cerebrovascular Research

Traditional Poster

Neuro A

Tuesday, 12 May 2026

Traditional Posters | Exhibition Hall

08:20 - 09:15

Session Number: 470-02

No CME/CE Credit

Similar Sessions

This traditional poster session focuses on innovative imaging and computational approaches to improve the detection, characterization, and prognosis of stroke and cerebrovascular injury. Presentations span advanced MRI techniques (e.g., diffusion models, APTw imaging, vascular fingerprinting, susceptibility imaging), physiological and biochemical markers of ischemia, and machine learning models for outcome prediction. Several studies also introduce synthetic data generation and mechanistic modeling to support robust algorithm development. Together, these posters highlight how quantitative imaging and AI-driven methods can deepen understanding of stroke pathophysiology and support more precise clinical decision-making.

		470-02-079. Predicting Longitudinal Language Deficits with Connectome‑Thresholded Graph Neural Networks on Brain Connectivity Qiao Deng, Yuming Zhong, Sai Kam Hui The Chinese University of Hong Kong, Shatin, Hong Kong Impact: CT‑GNN enables earlier, patient-specific prognostication of post‑stroke language trajectories.
		470-02-080. Deep Learning and Brain Parcellation–Based Machine Learning Model for Prognosis after Acute Ischemic Stroke Tung-Yang Lee, Chun-Jung Juan, Shao-Chieh Lin, Chia-Ching Chang, Cheng-Hsuan Juan, Cheng-En Juan, Ya-Hui Li, Yi-Jui Liu Feng Chia University, Taichung, Taiwan Impact: Integrating deep learning–based lesion segmentation, brain parcellation, and machine-learning classification with demographic information enables objective and accurate 3-month outcome prediction after acute ischemic stroke, enhancing precision medicine and supporting personalized, evidence-based decision-making in stroke management.
		470-02-081. Clinical Value of Rapid Amide Proton Transfer-weighted Imaging in Detecting Ischemic Penumbra in Acute Cerebral Infarction Jimei Tang, Zhiqiang Chang, Zhiwei Shen Xing’an League People’s Hospital, Inner Mongolia, China Impact: Currently, the use of APTw technology to assess acute ischemic stroke is predominantly concentrated in animal studies, with clinical research focusing on patients with a disease duration of less than 24 hours being rather rare.
		470-02-082. Extended MR Vascular Fingerprinting for the study of brain lesions: implementation and tests in a stroke model Maitê Marçal, Thomas Coudert, Aurélien Delphin, Daniil Kirdyashkin, Antoine Barrier, Emmanuel BARBIER, Benjamin Lemasson, Thomas Christen Univ. Grenoble Alpes, INSERM, U1216, Grenoble Institute Neurosciences, GIN, Grenoble, France Impact: MR-vascular Fingerprinting using realistic geometries and advanced simulations models allows the simultaneous evaluation of multiple microvascular properties. The proposed extended framework better characterizes brain microvasculature, and could potentially improve stroke management.
		470-02-083. Evaluation of pH Changes in Brain Infarction with MTRasym and MRAPTR in APTw MRI and its Influencing Factors Huan Li, Tanoj Bahadur Singh, Ting Liu, Tianhao Wang, Danni Huang, Yin Wu, Hao dong Qin, Jianzhong Yin Haikou People's Hospital, haikou, China Impact: This research tackles the clinical challenge of standardizing APTw-MRI in ischemic stroke. The findings confirm the higher diagnostic performance of MRAPTR and advance the clinical translation of APTw-MRI by defining its limitations and reliability.
		470-02-084. Synthetic Stroke Lesion Generation Based on Mechanistic Principles and Label to Image Methodology Aksel Leknes, Ayo Zahra, Ketil Oppedal, Kathinka Kurz, Martin Kurz, Thomas Lindner, Soffien Ajmi, Muriel Bruchhage University of Stavanger, STAVANGER, Norway Impact: Developing synthetic stroke images can be a basis for robust stroke segmentation models across modalities and contrasts. Additionally, the mechanistic approach of lesion generation can serve as a basis for descriptive and predictive stroke analysis tools.
		470-02-085. Alterations in Brain Iron in the Basal Ganglia of Acute Ischemic Stroke Patients: A Susceptibility Source Separation Imaging Jie Yang, Shenghai Zhou, Hao Feng, Meining Chen, Wei Sheng, Hui Zhang, Jianquan Zhong First People's Hospital of Zigong City, Zigong, China Impact: Susceptibility source separation technology may provide a new advantageous tool for detecting abnormal brain iron deposition in the basal ganglia of AIS patients.
		470-02-086. White matter hyperintensity burden as a predictor of white matter damage Hyeong-Geol Shin, Sarvin Sasannia, Mykola Matsyuk, Shimeng Wang, Jinwei Zhang, Xu Li, Filip Szczepankiewicz, Jerry Prince, Richard Leigh, Linda Knutsson, Peter van Zijl, Paul Nyquist Johns Hopkins University, Baltimore, United States of America Impact: Cross-sectional WMH burden better captures ongoing white matter degeneration (measured by diffusion MRI), outperforming BBB permeability indicators and WMH burden progression history, highlighting the potential of a simple volumetric parameter as a clinically relevant biomarker of small vessel disease–related injury.
		470-02-087. Automatic Quantification of Ischemic Core, Penumbra, and Perfusion–Diffusion Mismatch using Perfusion and Diffusion MRI Anup Singh, Ankit Kandpal, Rakesh Kumar Gupta Indian Institute of Technology, Delhi, India Impact: This study presents an endogenous contrast-based, rapid, and fully automated deep-learning framework that segments ischemic core and penumbra and quantifies perfusion–diffusion mismatch from multiparametric MRI, enabling accurate and reproducible stroke assessment to support real-time clinical decision-making.