Artificial Intelligence in Geriatric Mental Health: Recent Advances in Clinical Research

Abstract

With the rise of wearable sensors, advancement in comprehensible artificial intelligence (AI) algorithms, and growing acceptance of AI in medicine, AI has great potential to aid in diagnosis, prognosis, and treatment of mental illnesses and improve reliability. The rapidly rising number of older adults worldwide presents a unique challenge for clinicians due to increased mental health needs in the setting of a dwindling clinical workforce. AI has enabled researchers to better understand mental illnesses by taking advantage of ‘big data,’ that often cannot easily be transformed or considered in patient care.This symposium will present an overview of novel research leveraging AI (machine learning, natural language processing, deep learning) to better track, understand, and support mental health and cognitive functioning in older adults.Helmet Karim, PhD will present on machine learning prediction of Alzheimer's disease (AD) 10+ years prior to its onset using electronic health records (EHR) leveraging data from ∼14,000 cases and ∼31,000 controls, and the implications of these models in diagnosis and early detection.Ellen Lee, MD will present on using natural language processing to understand psychosocial functioning and cognitive functioning in older adults.Harmehr Sekhon, PhD will present on virtual reality, telehealth and technology to improve mental health in older adults and discuss how AI can impact treatment delivery and engagementAndrea Iaboni, MD DPhil FRCPC will discuss the potential benefits and limitations of unsupervised deep learning in healthcare, and present work related to its application in video-based detection of agitation in dementia.The proposed session is fully consistent with the theme of the 2022 AAGP meeting “Addressing Disparities in Geriatric Mental Health: Caring for Our Patients and Communities” as AI approaches are well-suited for scalable and remote monitoring and treatment for older adults, especially in underserved communities. AI technology can characterize the heterogeneity observed in aging, mental illnesses, and cognitive decline.