Yeah, I think we’re also trying to build several assistive tools that can aid in practice. So right now we are working on trying to create this kind of a multi-modal AI framework that combines all this clinical neurology workup data to make highly accurate predictions of differential diagnosis of dementia. So what we did was we collected data across many different cohorts around the world...
Yeah, I think we’re also trying to build several assistive tools that can aid in practice. So right now we are working on trying to create this kind of a multi-modal AI framework that combines all this clinical neurology workup data to make highly accurate predictions of differential diagnosis of dementia. So what we did was we collected data across many different cohorts around the world. Many obviously are based in the U.S., but also outside the U.S., and we’re trying to combine all this information that actually includes patient-level history, lab information, laboratory values, medications, cognitive test results, neuroimaging data, and then we’re trying to see if AI can be used to sort of combine all this information to perform differential diagnosis of dementia. So what it does is it basically does two kinds of predictions. One is to sort of really take all this information and then see if the person has healthy cognition or mild cognitive impairment or dementia. And once the model identifies the person to have some sort of cognitive impairment, it sort of goes one step further to see if the model can look at the primary causes that are contributing to that cognitive impairment. So for example, a person may have dementia due to Alzheimer’s disease. Some other person may have dementia due to Alzheimer’s as well as depression. So the model is able to now capture all those nuances and not only identify the status of cognitive impairment, but also the root causes of cognitive impairment.
This transcript is AI-generated. While we strive for accuracy, please verify this copy with the video.
