
FDA Clears MICSI-PET for Automated Neurological PET Analysis
Key Takeaways
- Cleared status may expand neurological PET post-processing options, but the intended clinical role—diagnosis, longitudinal assessment, or treatment monitoring—remains unspecified.
- Missing regulatory details (510(k) number, predicate, classification, clearance date) prevent determining whether MICSI-PET is standalone software, an accessory, or part of a larger analytics suite.
Microstructure Imaging has announced that the FDA cleared its MICSI-PET, an MR-guided PET enhancement platform with automated amyloid Centiloid and tau quantification.
The FDA has granted 510(k) clearance to Microstructure Imaging’s MICSI-PET, a software platform intended to provide MRI enhancement and automated quantification for neurological positron emission tomography (PET).1 The clearance adds a quantitative, MRI-informed layer to PET interpretation at a time when amyloid and tau imaging play an increasingly central role in
“The difference can be dramatic in borderline amyloid cases,” Lawrence N. Tanenbaum, MD, FACR, a neuroradiologist and imaging technology advisor, said in a statement.1 “When you’re trying to determine whether there is true cortical sparing, MICSI-PET can make the pattern substantially easier to see. Adding automated quantification to that improved visualization gives the reader a much more powerful way to evaluate the examination.”
MICSI-PET uses structural MRI data to guide denoising and super-resolution processing of PET images, an approach intended to sharpen image quality while preserving the underlying quantitative PET signal. The platform automatically generates Centiloid values for amyloid PET and CenTauR values for tau PET, alongside regional standardized uptake value ratios (SUVr) and z-scores referenced against normative control data.
It is compatible with amyloid, tau, and FDG PET studies and was validated for use across PET and MRI hardware from multiple vendors and scanner generations, according to the company. The clearance follows a 2024 510(k) clearance for MICSI-RMT, the company’s earlier MRI denoising software built on a random matrix theory–based algorithm.
READ MORE:
Quantitative amyloid and tau PET have taken on greater clinical weight since the approvals of the antiamyloid antibodies lecanemab and donanemab, both of which require biomarker confirmation of amyloid pathology before treatment and rely on serial imaging to track amyloid clearance during therapy.2 The Centiloid scale was developed to harmonize amyloid PET measurements across tracers, scanners, and processing pipelines by anchoring results to a common 0-to-100 scale, addressing longstanding variability in standardized uptake value ratio–based reads.3
The more recently introduced CenTauR scale applies a similar harmonization approach to tau PET, converting tracer-specific SUVr values into a universal unit using a joint propagation model derived from head-to-head imaging data across six commonly used tau tracers.4 Visual interpretation remains the clinical standard for both amyloid and tau PET, but quantitative measures are increasingly used to complement reads, particularly in borderline or equivocal scans.
MICSI is separately anticipating a $2.275 million Phase II Small Business Innovation Research award from the National Cancer Institute to support multicenter validation of an upcoming neuroimaging platform, in collaboration with NYU Langone Health and Memorial Sloan Kettering Cancer Center.1 Independent validation of MICSI-PET's performance across clinical populations and scanner types, along with peer-reviewed reporting of its accuracy relative to established quantification methods, may be needed to define its role alongside existing amyloid and tau PET workflows.

















