TL;DR:
- FDA approves BrainSee, an AI software by Darmiyan, for predicting Alzheimer’s disease (AD) progression.
- BrainSee uses MRI data and cognitive test scores to estimate the likelihood of AD within five years.
- Initial studies indicate high accuracy in predicting AD for individuals with amnestic mild cognitive impairment (aMCI).
- Experts emphasize the need for multiple information sources in dementia prediction, including blood-based biomarkers.
- BrainSee offers a same-day AD dementia probability score and is available to various healthcare professionals.
- Pricing for BrainSee is $1,500, with a reduced rate of $300 and potential Medicare coverage.
Main AI News:
In a groundbreaking development for the medical field, the FDA has granted its approval to BrainSee, an AI-driven software platform created by San Francisco-based Darmiyan. This innovative software harnesses the power of artificial intelligence to analyze MRI data and cognitive test scores, offering predictions regarding the likelihood of an individual with mild cognitive impairment progressing to Alzheimer’s disease (AD) within a span of five years.
Traditionally, amyloid-PET scans and fluid biomarkers have been considered the gold standard for assessing AD risk. However, BrainSee introduces a novel approach by detecting subtle brain atrophy patterns within routine MRI scans, shedding light on potential AD progression. The FDA’s stamp of approval for BrainSee follows its recognition with a breakthrough designation in 2021.
One question that has piqued the curiosity of experts like Nick Fox from University College London is the extent to which BrainSee relies on hippocampal atrophy as a predictor. He suggests that the FDA approval may be primarily based on estimating the risk of progression in individuals with amnestic mild cognitive impairment (aMCI), given the well-established link between hippocampal atrophy and AD dementia. Further peer-reviewed studies are necessary to fully comprehend the sensitivity and specificity of BrainSee’s approach. However, preliminary findings suggest that the software can predict impending AD dementia with an impressive accuracy rate ranging from 88 to 91 percent among those with aMCI.
Kate Papp, a representative from Brigham and Women’s Hospital in Boston, acknowledges the promise of this technology in helping patients understand the trajectory of their disease. However, she emphasizes the importance of gathering information from multiple sources, such as blood-based biomarkers, cognitive assessments, genetics, and current cognitive function, as the ultimate predictor of dementia is unlikely to rely on a single modality or test. Blood biomarkers, boasting an accuracy of 90 percent or more in detecting AD pathology, are also awaiting regulatory approval, further enhancing the diagnostic landscape.
Long before Alzheimer’s-related brain atrophy becomes evident on an MRI scan, neurons begin to deteriorate in specific brain regions. Researchers have long hypothesized that these subtle atrophy patterns could serve as early indicators of AD and other neurodegenerative diseases. However, the challenge lies in detecting these minute changes and accurately assessing their predictive value—a task ideally suited for the computational prowess of AI, which has become an invaluable tool in dementia research.
BrainSee’s training process involved analyzing brain scans and cognitive scores from thousands of patients. The algorithm incorporates scores from the mini-mental state examination (MMSE), the Clinical Dementia Rating scale sum of boxes (CDR-SB), and standard whole-brain MRI data. It then generates an AD dementia probability score on a scale of 0 to 100, with higher scores indicating a greater likelihood of dementia within five years.
Structural MRI scans are already a routine part of the clinical assessment for individuals with subjective cognitive complaints. BrainSee is readily available to healthcare professionals across various specialties, including neurologists, geriatricians, psychiatrists, family physicians, and internal medicine physicians. Physicians can upload their patients’ MRI scans and cognitive assessment scores to BrainSee’s secure web portal, receiving a same-day AD dementia probability score. The software also includes a comprehensive tutorial and interpretation guide for both healthcare providers and patients and caregivers.
While BrainSee carries a price tag of $1,500, Darmiyan offers it at a reduced rate of $300, with potential Medicare coverage. Considering that Medicare already covers the cost of MRI scans (averaging around $1,000 for out-of-pocket payments), BrainSee is positioned as an initial screening tool. Individuals with a low probability of dementia can explore alternative explanations for their cognitive concerns, while those with a high likelihood of AD can pursue further diagnostic testing and treatment options.
Conclusion:
The FDA’s approval of BrainSee signifies a significant advancement in Alzheimer’s disease prediction, leveraging AI and MRI data. This innovation will likely impact the healthcare market by enhancing early diagnosis and guiding patient care decisions. The combination of BrainSee with emerging blood biomarkers is poised to revolutionize dementia assessment, potentially reducing healthcare costs and improving patient outcomes.
