Meeting Banner
Abstract #0016

Prediction of Tissue Recovery in Chronic Stroke Using Adaptive Models and Acute MR Information

Hassan Bagher-Ebadian 1,2 , Marie Luby 3 , James R Ewing 2,4 , Panayiotis Mitsias 4 , and Hamid Soltanian-Zadeh 1,5

1 Radiology, Henry Ford Hospital, Detroit, MI, United States, 2 Physics, Oakland University, Rochester, MI, United States, 3 National Institute of Neurological Disorders and Stroke, MD, United States, 4 Neurology, Henry Ford Hospital, Detroit, MI, United States, 5 CIPCE, ECE Dept., University of Tehran, Tehran, Iran

This pilot study introduces four different adaptive-models (the inelastic-collision (IC) model, the Kohonen-Multi-Parametric-Self-Organizing-Map (KMP-SOM), the Generalized-Linear-Model (GLM) and an Artificial-Neural-Network) for multi-parametric analysis. These models are applied on acute MR information of eleven treatment-nave patients to predict tissue recovery in chronic stroke. All patients presenting with acute neurological deficit consistent with stroke, and had MRI studies done within 24h of onset. Results imply that adaptive models are capable of identifying the ischemic growth (in pattern and size), and may describe tissue viability. Thus adaptive models can play important role in the assessment of acute and sub-acute therapeutic interventions of stroke.

This abstract and the presentation materials are available to members only; a login is required.

Join Here