The green curve shows the distribution of general population, while other curves are abnormal in nature. Mr. Chang's (81y) result (black dot) shows a higher value than normal, yielding an AI predicted brain age of 86.4 years.
In clinical practice, neurologists scrutinize dementia patients with using numerous tools, including medical history, blood tests, cognitive testing, mental assessment, medical imaging and electro-physiological signal testing. Patients usually spend more than an hour to complete complex questionnaires, and health check-up schedules at hospitals for patients, however, are full up and overloaded. To fulfill patient needs and assist physicians to determine severity of dementia more efficiently and accurately, China Medical University Hospital (CMUH, Taiwan) developed ‘BrainHealth’, an AI system powered by AI Brain Age Prediction System and Neural Gene Discrimination System to facilitate complex assessments of dementia severity. The AI model takes only one minute to help physicians diagnose dementia, with an AUC of 87% and a sensitivity of 91.7%.
Mr. Chang (81-year-old) has encountered growing difficulty in recalling what he has been told, become reluctant to interact with family and speaks in a way that is difficult to comprehend. Conventional examinations including blood testing, mental capacity questionnaires and brain MRI were performed at CMUH. Chang was found to have a lower Mini Mental State Examination (MMSE) score than the mean for his age and MRI images showed brain atrophy (but not stroke or tumor). These data, along with the findings of gray matter degeneration and increased cerebrospinal fluid volume, were put into CMUH’s BrainHealth system. The AI model predicted that Mr. Chang’s brain-age is 5 years older than would be expected and indicated that he had early-stage dementia. The routine examination findings were in line with BrainHealth’s prediction of Mr. Chang having early-stage dementia and suggested that further management is warranted.
Dementia is a brain disorder causing diminished mental capacity and memory, and can lead to mood disorder, loss of verbal expression and reduced mobility. Dementia as a neurocognitive disorder presented in a range of severities and subtypes (each with specific pathophysiology). Depending on the area affected and the presentation, it may be termed as Alzheimer’s disease, vascular dementia, dementia with Lewy bodies, frontotemporal dementia or Parkinson’s disease. It is usually seen in people of middle or advanced age and progresses with time. An estimation by Taiwan Ministry of Health and Welfare suggested that there are more than 300,000 people affected by dementia in 2022, with 96% being elderly (65 years or older), but with more younger cases reported recently. The number of dementia patients is projected to exceed 500,000 by 2030.
According to AI Center at CMUH, BrainHealth, the AI Brain Age Prediction System, incorporates a normal distribution curve derived using MRI slides of brain areas (gray/white matter, cerebrospinal fluid, hippocampus) from nearly three thousand healthy individuals and approximately five hundred dementia patients. With this database, CMUH AI Center discovered that dementia patients generally show less gray matter and a noticeable increase of cerebrospinal fluid (due to brain degeneration) compared to the general population. AI Center at CMUH cited a medical article published in 2022 that the brain age of ordinary people is typically within 3 years of physiological age. The Brain Age Prediction System helps clinicians determine severity of dementia in an efficient manner.
Additionally, the hereditary neurological disorder testing system of BrainHealth determines the presence of Alzheimer’s disease based on an AI model that compares the brain wave and genetic characteristics extracted by algorithm with data from the geriatric population both with and without neurological disorders. Dementia patients show significantly lower brain wave intensity and complexity at specific frequency bands, as well as functional connections between brain areas, than dementia-free individuals. As Taiwan enters an aged society, the prevention and treatment of early-stage dementia diseases have become urgent tasks. The AI scrutiny system‘BrainHealth’at CMUH would keep to assist medical professionals in diagnosing dementia.
Related news from AP News: https://apnews.com/press-release/pr-newswire/dementia-7816de947f9fc4367c72f811a3f95dc6