The Stevens Institute of Technology developed an Artificial Intelligence (AI) tool that can diagnose Alzheimer’s disease with 95% accuracy, eliminating the need for expensive scans or in-person testing. Alzheimer’s disease can affect a person’s use of language. When individuals with Alzheimer’s tend to replace nouns with pronouns, and they can express themselves in a very roundabout, awkward way. (1)
This AI identifies hallmark signs of Alzheimer’s and subtle linguistic pattern changes previously unnoticed. Scientists taught the algorithm using texts formed by healthy individuals and known Alzheimer’s subjects, describing a typical scenario, in this case, a kid grabbing cookies from a jar. The team converted each sentence into a unique numerical value or vector. In a way, speech is translated into quantifiable language for the AI to train in recognizing patterns of Alzheimer’s.
Researchers’ next steps will be to collect new data that will feed the algorithm and help recognize and diagnose patients. This approach is extrapolated to other neurological pathologies. The team is also working on uncovering how other neurological conditions, such as aphasia, stroke, traumatic brain injuries, and depression, affect language. (1)
The Stevens Institute of Technology developed an Artificial Intelligence (AI) tool that can diagnose Alzheimer’s disease with 95% accuracy, eliminating the need for expensive scans or in-person testing. Alzheimer’s disease can affect a person’s use of language. When individuals with Alzheimer’s tend to replace nouns with pronouns, and they can express themselves in a very roundabout, awkward way. (1)
This AI identifies hallmark signs of Alzheimer’s and subtle linguistic pattern changes previously unnoticed. Scientists taught the algorithm using texts formed by healthy individuals and known Alzheimer’s subjects, describing a typical scenario, in this case, a kid grabbing cookies from a jar. The team converted each sentence into a unique numerical value or vector. In a way, speech is translated into quantifiable language for the AI to train in recognizing patterns of Alzheimer’s.
Researchers’ next steps will be to collect new data that will feed the algorithm and help recognize and diagnose patients. This approach is extrapolated to other neurological pathologies. The team is also working on uncovering how other neurological conditions, such as aphasia, stroke, traumatic brain injuries, and depression, affect language. (1)
An artificial intelligence (AI) program created to estimate which individuals in the intensive care unit had an increased risk of deteriorating has won FDA clearance. The AI is designed to warn clinicians in a timely manner to prepare and modify management accordingly. This platform analyses the patient’s medical record, vital sign data, and other ICU measurements. (2)
CLEW’S company reports that the AI system can win up to eight hours in advance, enabling early intervention. It also identifies low-risk patients with a lower risk of deteriorating, potentially facilitating better ICU resource management and optimization. The system has improved, as earlier studies would only anticipate these events 3 hours before they occur. The company also plans to explore other healthcare settings where patients deteriorate. (3)