Abstract
We have developed a novel medical image processing technology for objective, accurate, and non-invasive diagnosis, prognosis, and monitoring of Alzheimer’s disease using standard clinical brain MRI (magnetic resonance imaging) and basic clinical cognitive assessment. The technology is a robust, highly accurate, fully automated, high-throughput, cloud-based platform, which operates in a fully integrated and controlled environment. In diagnosis mode, our technology performs with 91% balanced accuracy on blind testing to diagnose the current neurocognitive status, and in prognosis mode or early detection at mild cognitive impairment (MCI) stage with 88% balanced accuracy on blind testing to predict progression from MCI to Alzheimer’s dementia (AD) within 5 years. Such prognostic capability is currently non-existent, even in specialty clinics and hospitals, a major factor in Alzheimer’s clinical trial failures. The algorithm’s diagnostic certainty precisely mirrors the diagnostic confidence of an expert cognitive neurologist for both MCI (Spearman’s Rho = 1) and AD (Rho = 1). In addition to widespread clinical applications, this novel technology can enable correct patient selection and therapeutic effect monitoring in clinical trials of Alzheimer’s disease, the crucial elements to finding a cure.
Footnotes
↵† Shared first authors
Abbreviations
- CN
- Cognitively Normal
- MCI
- Mild Cognitive Impairment
- AD
- Alzheimer’s Dementia
- MMSE
- Mini Mental State Exam
- CDR
- Clinical Dementia Rating
- FAQ
- Functional Assessment Questionnaire
- ADAS
- Alzheimer’s Disease Assessment Scale
- MRI
- Magnetic Resonance Imaging
- PET
- Positron Emission Tomography
- CT
- Computed Tomography
- CSF
- Cerebro-Spinal Fluid
- ML
- Machine Learning
- AI
- Artificial Intelligence
- ADNI
- Alzheimer’s Disease Neuroimaging Initiative
- OASIS
- Open Access Series of Imaging Studies
- AIBL
- Australian Imaging, Biomarkers and Lifestyle study