As our population ages, our society faces the challenges of how to support and care for our older citizens. For many good reasons — not the least of which are saving on healthcare and other costs — seniors want to live independently in their own homes as long as possible. While it can be fairly easy to identify major illness and physical limitations, it can be far more difficult to monitor an individual’s cognitive abilities. As individuals, we are notoriously bad at self-reporting our own abilities with any accuracy, and those close to us are often inclined to compensate and cover for any mild cognitive impairment.
Screening for cognitive impairment traditionally is a lengthy and expensive process, using techniques ranging from pencil-and-paper tests to MRI scans. Winterlight Labs has developed an alternative approach that is unobtrusive, simple, and potentially inexpensive to administer. The system uses a short one- to five-minute recording of the the subject’s natural speech, which is then analyzed in multiple ways. The spoken words are turned into text transcripts. The audio recording and transcript provide source data for acoustic, lexical, syntax, and semantic analysis. The results of this analysis are then compared with a large dataset based on longitudinal studies. The result is a probability score of whether or not the individual will be diagnosed as healthy or as having dementia (or other cognitive impairment).
This technology could be used as a simple first-level screening for possible Alzheimer’s patients. Not only would this lower the cost of identifying healthy individuals, it could also help find the hundreds of thousands of subjects with as-yet undiagnosed Alzheimer’s. The technology could also speed drug trials while lowering subject screening costs, evaluate the impairment of stroke victims, and even help identify patients suffering from depression. This new approach could save lives and help make healthcare expenditures much more effective.