The test, developed by neurologist Jane Alty and computer scientist Quan Bai, emerged from a 2019 observation in a dementia clinic waiting room, where Alty noticed patients exhibiting movement patterns similar to those of Parkinson’s Disease. This insight, which challenged conventional views of dementia as primarily a cognitive disorder, sparked a seven-year collaboration involving interdisciplinary researchers.
Alty, a UK-born neurologist specializing in movement disorders, made the pivotal observation in 2019 while working at the University of Tasmania. I was thinking, ‘But they don’t have Parkinson’s, they’ve got dementia,’ then the penny dropped and I thought, ‘but what if there are more similarities here?’
she recalled. This insight proposed that movement changes might precede cognitive decline, a theory that contradicted the traditional understanding of dementia as a memory-related condition. It might be that right at the beginning, movement is what’s changing before any typical memory symptoms,
Alty explained.
Scientist's brainstorm transforms dementia-risk testing
The idea gained momentum through interdisciplinary collaboration. Physiotherapist Associate Professor Kate Lawler and dementia researcher Professor James Vickers helped transform the concept into a practical test. I’m not a computer scientist, I didn’t have the technical know-how to actually translate that into a test,
Alty said. She found a partner in Quan Bai, a computer scientist at the University of Tasmania, who had no medical background but possessed the AI expertise needed to develop the system. We had no funding, no anything, so we started it as a PhD student project,
Bai said, describing the project’s humble beginnings.
The team created a prototype web application that analyzed finger-tapping movements to detect subtle changes indicative of dementia risk. Surprisingly, we got some brilliant results,
Bai said. This early success led to pilot testing and data collection from 111 older adults who were comprehensively evaluated and administered the Loewenstein-Acevedo Scales for Semantic Interference and Learning, Brief Computerized Version (LASSI-BC). The study, published in an abstract, focused on expanding previous work to determine whether the LASSI-BC could differentiate between cognitively unimpaired (CU) and amnestic mild cognitive impairment (aMCI) groups. Results indicated that failure to recover from proactive semantic interference (frPSI) on the LASSI-BC differentiated between CU and aMCI groups, with an area under the ROC curve of 0.876 (76.1% sensitivity, 82.7% specificity). FrPSI was also associated with volumetric reductions in the hippocampus, amygdala, inferior temporal lobes, precuneus, and posterior cingulate.
Brain scientist warns we're heading for AI-fueled 'dementia crisis
The TAS Test leverages computer vision algorithms to analyze hand movements during a simple finger-tap task. We can measure movement because we’ve got these computer vision algorithms that just require a smartphone or a digital camera,
Alty explained.
The collaboration involved students who joined the effort, many drawn by personal connections to dementia. Many later said they became involved because they had family members with dementia,
Alty noted. The next phase involves refining the algorithm based on ongoing data collection and expanding its use globally.
A Novel Computerized Cognitive Test for the Detection of
The study published in the abstract highlights the need for cognitive assessments sensitive enough to measure early brain-behavior manifestations of Alzheimer’s disease. Weak sensitivity to early cognitive change has been a major limitation of traditional cognitive assessments,
the study states. The LASSI-BC, which correlates with biomarkers of neurodegeneration, represents a step forward in addressing this gap.
The collaboration between Alty, Bai, Lawler, and Vickers exemplifies how interdisciplinary research can address complex health challenges, offering new hope for early dementia detection.
