Abstract #4385
Brain tumor imaging based, histology trained maps (IBHTMs) of cellularity predict tumor presence in pathologically confirmed regions sampled ex-vivo
Peter S LaViolette 1 , Elizabeth J Cochran 2 , Nikolai Mickevicius 3 , Jennifer Connelly 4 , Kathleen M Schmainda 1,3 , and Scott D Rand 1
1
Radiology, Medical College of Wisconsin,
Milwaukee, WI, United States,
2
Pathology,
Medical College of Wisconsin, Milwaukee, WI, United
States,
3
Biophysics,
Medical College of Wisconsin, Milwaukee, WI, United
States,
4
Neurology,
Medical College of Wisconsin, Milwaukee, WI, United
States
Infiltrative brain cancer is difficult to detect outside
of contrast-enhancing regions. We introduce a new method
that uses co-registered histology and brain tissue from
whole brain donations to generate machine-learning based
maps of cell density. We sampled regions highlighted by
this new method and confirmed tumor in each location in
7 high-grade gliomas. In 5 of 7 patients,
hypercellularity was found outside of contrast
enhancement, and in one case outside of T2/FLAIR
hyperintensity.
This abstract and the presentation materials are available to members only;
a login is required.
Join Here
