Editor’s Note: We believe that patients are a key part of developing and leading the conversation in disease communities. Patient Worthy sometimes partners with reputable agencies that wish to speak with patients about opportunities related to their diagnosed conditions. These opportunities can include activities such as sharing stories with other patients or health professionals about their diagnosis journey or recording video testimonials. To learn more about how to get involved with an opportunity for Alzheimer’s Disease patients or caretakers, click here.

BBC NEWS recently reported on the versatility of AI in predicting dementia. Alzheimer’s destroys connections between the brain and nerve cells, thereby damaging cognitive abilities and memory. It is estimated that approximately 5.5 million individuals in the United States are living with the disorder.

IBM’s V.P. of Healthcare Research, Ajay Royyuru, told BBC NEWS that admittedly there is no cure for Alzheimer’s. However, the knowledge of who is at risk of developing the disease would facilitate earlier treatment. Alzheimer’s disease affects changes in the brain up to fifteen years prior to symptoms occurring. Preclinical tests have diagnosed dementia even when no signs of injury can be seen on a brain scan.

About Patterns

In addition, after one brain scan AI can determine whether someone’s condition will stabilize for several years, deteriorate or require immediate attention. Currently, several tests and brain scans must be used to determine the stages of dementia.

Professor Zoe Kourtzi of the Turing Institute commented that symptoms may occur later in a person’s life or perhaps not at all. Earlier intervention allows treatment to slow disease progression and prevent more damage.

Professor Kourtzi’s system analyzes the brain scans of people concerned about getting Alzheimer’s against thousands of patients with dementia together with their applicable medical records.

The system’s algorithm identifies patterns that even expert neurologists are unable to see. It then matches them to patient outcomes.

The Cookie Test

A line drawing of two children taking cookies from a cookie jar has been used by an AI system to detect early Alzheimer’s disease for years. Voice samples describing the scene predict whether healthy people are at risk for the disease. The AI model was developed by IBM Research together with Pfizer using natural language processing. The system analyzes speech responses to the Cookie Theft cognitive test.

The AI program can predict Alzheimer’s with seventy percent accuracy and fifteen years prior to predictions by doctors. The AI model was developed by IBM Research together with Pfizer using natural language processing. The system analyzes speech responses to the Cookie cognitive test asking people to describe the drawing.

AI was able to detect grammatical errors and changes in sentence structure indicating a cognitive decline. IBM’s samples used for the test were provided by the Framingham Heart Study research project. The study has existed since 1948 and is ongoing. Five thousand people, including their families, have participated to date.

About five hundred patients have been selected to participate in the current study. The results of the study will be sent to the patients’ doctors who will suggest appropriate treatment.

Dr. Royyuru acknowledged that many patients may not want to know about Alzheimer’s being part of their future. He noted that these individuals must give their consent to know, and their status will not be divulged without that consent.

The Lancet eClinicalMedicine journal published the paper.