Meeting Banner
Abstract #1939

Machine learning and computer vision based quantification of cell number in MRI-based cell tracking

Muhammed Jamal Afridi 1 , Matt Latourette 2 , Margaret F Bennewitz 3 , Arun Ross 1 , Xiaoming Liu 1 , and Erik M Shapiro 2

1 Department of Computer Science and Engineering, Michigan State University, East Lansing, MI, United States, 2 Department of Radiology, Michigan State University, East Lansing, MI, United States, 3 Vascular Medicine Institute, University of Pittsburgh, Pittsburgh, PA, United States

MRI detection of single cells is an underutilized advancement in MRI-based cell tracking. One reason for its underutilization has been a lack of methods for quantifying information in these images. For example, single cell detection enables quantification of cell numbers and accurate cell localization. To achieve single cell detection by MRI, cells are labeled with superparamagnetic iron oxide particles allowing their detection as punctate hypointensities in T2*-weighted MRI. We have developed a machine learning and computer vision based strategy for the generalizable detection and quantification of MRI-based single cell detection. Our approach can detect spots with an accuracy of 99.8%.

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

Join Here