Doctors agree that early diagnosis can have a positive impact on the course of almost every disease, and Alzheimer’s is no different.

There’s a new study that points to the ability of artificial intelligence (AI) to predict whether a person will go on to develop the disease.

This method involves using AI to measure simple language skills and is shown to have a 70 percent accuracy rate. Here’s the story…

Over the last decade, researchers have searched high and low looking for the key to unlocking accurate early diagnosis of Alzheimer’s disease.

We’ve reported on their many findings in this newsletter, ranging from the use of biomarkers and brain imaging or neuroimaging to examining changes in other body parts—such as the eyes—as well as genetic risk profiling.1

Past research has shown that language is a notable component of age-related cognitive decline. Researchers have recognized that even simple linguistic skills, such as object naming, involve extensive brain networks.

Linguistic “Mistakes” Predict Alzheimer’s Disease

Now, a new study led by Dr. Guillermo Cecchi of IBM Research examined linguistic patterns in word usage revealed by participants in the famous Framingham Heart Study, which followed 14,000 participants for three decades.2

Dr. Cecchi’s team used digital transcriptions of handwritten responses from Framingham Heart Study participants who were asked to describe a picture of a woman who is preoccupied with washing dishes while two kids raid a cookie jar behind her back.

While these transcriptions did not preserve the handwriting from the original responses, the team’s artificial intelligence (AI) model was able to detect linguistic “mistakes” that are sometimes related to early signs of cognitive impairment. For example, the AI model found certain misspellings, repeated words or the use of simplified phrases rather than grammatically complex sentences.

Armed with these variables, researchers developed computer models to predict whether a person would develop mild cognitive impairment (MCI) leading to Alzheimer’s disease.

The researchers first looked at the responses of 40 patients who eventually developed Alzheimer’s. They compared these to a control group of 40 who remained dementia-free until at least age 85. The computer model analysis revealed that future onset of Alzheimer’s disease is closely linked to repetitiveness, misspellings, and lack of basic grammatical structure in the handwritten responses.

Worked Better than Neuropsychological Evaluations

When Dr. Cecchi compared predictive models that used linguistic variables to predictive models that incorporated neuropsychological variables associated with Alzheimer’s risk he found performance was better with the linguistic models.

“For disease predictions based on a combination of these traditional variables, the accuracy was 59 percent,” Dr. Cecchi notes. On the other hand, the linguistic test had a 70 percent accuracy rate in predicting Alzheimer’s disease onset years before cognitive decline begins.3

The researchers hope that the new results will lead to the use of simple, inexpensive speech probes that not only detect early dementia, but also monitor its progression.

“The value of this [type of test] is that it can be done quickly; it’s not intrusive and can be done at any time,” Dr. Cecchi says.

And hypothetically, doctors could use the tool to identify patients who might benefit from enrollment in clinical trials of potential preventive therapies, he adds.

What’s more, patients who are identified as high risk could potentially make lifestyle changes to delay the decline.

“These could include following a healthy diet, becoming physically active, and enhancing social and cognitive engagement,” Dr. Cecchi explains.

The study was published in the online medical journal, EClinicalMedicine, a publication of The Lancet.4

My Takeaway

While this is intriguing research, it’s still in its early stages. There’s certainly more work to be done before this type of test can be used as a diagnostic tool.

Until that time, I’ll keep an eye on future research and advise that we all continue to take a proactive approach. I encourage you to follow healthy lifestyle strategies that can help slow the progression of dementia and may even prevent it altogether. Healthy diet and regular exercise, social interaction and stress management are great first steps.


  1. https://www.alz.org/alzheimers-dementia/research_progress/earlier-diagnosis
  2. https://framinghamheartstudy.org/
  3. https://www.scientificamerican.com/article/ai-assesses-alzheimers-risk-by-analyzing-word-usage/
  4. https://www.thelancet.com/journals/eclinm/article/PIIS2589-5370(20)30327-8/fulltext

Comments

comments