How Artificial Intelligence Is Unlocking Unstructured Healthcare Data To Help Transform Neurodegenerative Disease Treatment

Verana Health harnesses AI-powered large language models and machine learning to transform unstructured clinical notes into a rich, real-world evidence (RWE) resource for neurological diseases. By partnering with the American Academy of Neurology’s Axon Registry®, Verana accesses a six-year, longitudinal EHR database covering 3 million de-identified patients and 1,000 clinicians. Leveraging cloud computing and de-identified free-text data, its analytics pipeline flags linguistic cues to capture disease milestones often omitted by ICD-10 coding.

For multiple sclerosis, Verana extracts five core variables (subtype, relapse, ambulatory status, gait and EDSS proxies); for Parkinson’s disease, it models the four cardinal signs and their severities; and for Alzheimer’s disease, it decodes cognitive exam results alongside imaging and lab findings. This clinician-guided, performance-validated taxonomy delivers near-real-time insights on progression, treatment response and patient subtypes, enabling more precise intervention, trial design and outcomes research at scale—ushering in a new era of data-driven neurology.