NGS opens comment period for brain MRI AI noncoverage policy

Medicare administrative contractor National Government Services (NGS) is soliciting comments on a proposed local coverage determination (LCD) for noncoverage of automated detection and quantification of brain MRIs.

The outcome of this could affect at least a handful of U.S. Food and Drug Administration (FDA)-cleared radiology AI tools. Devices named and explained in the January 22 Centers for Medicare and Medicaid Services (CMS) notice include the following:

1. NeuroQuant Medical Image Processing Software (automatic labeling, visualization, and volumetric quantification of segmentable brain structures) and NeuroQuant 4.0 (uses AI modalities of machine learning and deep learning to aid in identifying complex patterns in imaging data).
    
2. Icobrain (automatic labeling, visualization, and volumetric quantification of segmentable brain structures) and Icobrain aria (computer-assisted detection and diagnosis [CADe/x] software for assisting radiologists with the detection and quantification of amyloid-related imaging abnormalities [ARIA]).
    
3. DeepBrain (automatic labeling, quantification, and visualization software of segmental brain structures).
    
4. Siemens Morphometry Analysis (syngo-based postacquisition imaging processing software for multiple modalities).

All are registered Class II devices under FDA 510(k) clearance. Other FDA-cleared devices may also be available but not listed because they were not found in a search of the literature, according to the CMS notice.

LCDs can vary by region. CGS Administrators, the MAC covering the J15 jurisdiction of Kentucky and Ohio, has already made the noncoverage policy clear with a final LCD (L40224) effective January 19, 2026. The comment period closed on November 8, 2025. No comments were received, according to the CMS.

CGS J15 discussed its stance during an October 28, 2025, meeting, concluding there is not sufficient evidence to support clinical utility or validity, and [automated detection and quantification of brain MRIs] is considered investigational, according to the transcript.

Most tools were found to have been validated using a small number of cases and or a single dataset, according to an analysis published in the notices. The report also highlighted a need to account for intrascanner variability resulting from differences in the scanners and instrument magnetic field and acquisition parameters, according to the report.

The comment period for the NGS proposed LCD (DL40332) opened with the posting on January 22 and will close on March 8, 2026.

Find all details and the opportunity to comment here.