Investigator: Teo Hock Hai.
In this project, we develop a digitalized screening tool sensitive enough to detect early cognitive impairment. The tool incorporates an artificial intelligence pattern recognition algorithm that has been trained to predict cognitive decline and measure processing speed with a digit symbol substitution test. The algorithm recognizes the patient’s handwriting, at the same time monitors the reaction speed and smoothness of a patient’s handwriting. It then detects outliers among these metrics and predicts the degree of the patient’s cognitive decline by fitting the outliers into our trained model. While paper-and-pen cognitive tests are commonly used in a clinical setting with other cognitive impairment tests, the tool offers a more efficient method for
- Early detection of cognitive problems, and
- Monitoring cognitive functions over time for patients diagnosed with dementia and other related diseases. These will translate to greater cost savings for detection tests in the long term.