Kevin Chikai Ma BS, Presenter: Nothing to Disclose

Joohyung Suh, Abstract Co-Author: Nothing to Disclose

Kathleen A. Garrison PhD, Abstract Co-Author: Nothing to Disclose

James Reza F. Fernandez MD, MS, Abstract Co-Author: Nothing to Disclose

Alexander Lerner MD, Abstract Co-Author: Nothing to Disclose

Mark S. Shiroishi MD, Abstract Co-Author: Nothing to Disclose

Lilyana Amezcua MD, Abstract Co-Author: Consultant, Biogen Idec Inc Consultant, Bayer AG Consultant, Teva Pharmaceutical Industries Ltd Consultant, Merck KGaA (Merck Serono International SA) Consultant, Pfizer Inc Advisory Board, Biogen Idec Inc Advisory Board, Bayer AG Advisory Board, Teva Pharmaceutical Industries Ltd Advisory Board, Pfizer Inc Advisory Board, Merck KGaA (Merck Serono International SA) Speaker, Biogen Idec Inc Speaker, Bayer AG Speaker, Teva Pharmaceutical Industries Ltd Speaker, Merck KGaA (Merck Serono International SA) Speaker, Pfizer Inc

Brent Julius Liu PhD, Abstract Co-Author: Nothing to Disclose

Automatic lesion detection and quantification algorithm in the MS eFolder system aims to provide disease progress visual monitoring on MRI studies. A brain and lesion registration and tracking tool has been added to the existing CAD algorithm to both identify lesions’ anatomical locations and track a lesion’s activity and size in a longitudinal study. Preliminary results have shown success in lesion registration, matching in longitudinal studies, and progression viewing in an eFolder’s web-based GUI. 

We have presented an automatic lesion detection and quantification algorithm for multiple sclerosis patients the previous year at RSNA. The computer-aided detection (CAD) tool has been integrated into an imaging informatics-based system, called eFolder, to provide a complete solution to patient tracking for both clinical and research purposes. In order to achieve longitudinal tracking, the CAD algorithm is modified to include a lesion registration and tracking tool. The goal is to register MS lesions on a patient’s MRI and track lesion changes and volumes in subsequent longitudinal studies.

Brain matter has been segmented and extracted from axial slices. Brain matter is reconstructed in 3D space, and spatially morphed in accordance to a normal brain template. MS lesions, detected by the CAD algorithm, are numbered for each case and registered for the patients’ profiles. Locations of each lesion are characterized by coordinates as well as anatomical locations relative to the brain template. Data collected for the system evaluation include 5 patients with 3 longitudinal studies each, and results of lesion registration is stored and displayed by the MS eFolder system. Two neuroradiologists are asked to review the registration results and to ensure accuracy of lesion detection and registration.

The lesion registering and tracking algorithm has been designed and developed. For preliminary results, five patients with three longitudinal studies have been analyzed. Lesions present in brain MRI of all five patients have been successfully registered, and the lesions were subsequently identified in their longitudinal scans. 

Ma, K, Suh, J, Garrison, K, Fernandez, J, Lerner, A, Shiroishi, M, Amezcua, L, Liu, B, An Automatic Multiple Sclerosis Lesion Tracking Tool for Longitudinal MRI Studies.  Radiological Society of North America 2011 Scientific Assembly and Annual Meeting, November 26 - December 2, 2011 ,Chicago IL. http://archive.rsna.org/2011/11008154.html