Identifying Non-enhancing Infiltrative Tumor in High Grade Glioma

Purpose	To accurately delineate and quantify non-enhancing tumor in high-grade  glioma.
Tag(s)
Panel	Neuroradiology
Define-AI ID	21020023
Originator	Mark Lum
Lead	Mark Lum
Panel Chair	Alex Norbash
Panel Reviewers	Neuoradiology Panel
License	Creative Commons 4.0
Status	Public Comment
RadElement Set	RDES188

Value Proposition

Differentiating vasogenic edema from non-enhancing T2-hyperintense infiltrative tumor in high grade gliomas has prognostic and therapeutic implications but remains challenging to radiologists on conventional MR. While some features of non-enhancing tumor have been described such as gray matter involvement and degree of T2 signal, changes can be subtle and there is considerable overlap of features (Lasocki 2019). Texture analysis appears promising in identifying non-enhancing tumor (Artzi 2018).

Narrative(s)

62-year-old woman with a history of a left frontal glioblastoma that was resected 3 months ago and received adjuvant chemoradiation and bevacizumab presents for a routine follow-up MRI. Her nearest imaging center does not have the capability of performing perfusion MRI.

Workflow Description

A texture analysis algorithm identifies areas suspicious for non-enhancing tumor based on multiple conventional sequences. A heat map indicating the likelihood of non-enhancing tumor is fused to T2-weighted FLAIR images. Volume of abnormal T2 signal meeting a predetermined likelihood threshold for non-enhancing tumor may be reported by the radiologist.

Inputs

 

DICOM Study

Primary Outputs

Volume of infiltrative, non-enhancing tumor

Lasocki A, Gaillard F. Non-Contrast-Enhancing Tumor: A New Frontier in Glioblastoma Research. AJNR Am J Neuroradiol. 2019;40(5):758-765.

Artzi M, Liberman G, Blumenthal DT, Aizenstein O, Bokstein F, Ben bashat D. Differentiation between vasogenic edema and infiltrative tumor in patients with high-grade gliomas using texture patch-based analysis. J Magn Reson Imaging. 2018