Experts from Research and Development division at IBM have come up with a study, which says that it is possible to predict the onset of Alzheimer’s disease in people by observing their writing patterns. IBM experts have said that the risk of Alzheimer’s disease can be predicted by a writing test years before the symptoms of the disease start showing up. They have said that people who are dealing with a wide range of neurological issues have a distinctive language pattern, which serves as an early sign of their conditions. Scientists have trained an artificial intelligence program (AI), which can trace changes in language ahead of the beginning of neurological issues. In the study, experts have enrolled 80 men and women who have been in their 80s.
Half of the participants have been diagnosed with Alzheimer’s and half of them have been normal. However, seven and a half years earlier, all of them have been cognitively healthy. These participants have been a part of the Framingham Heart Study, a long-running research, which needs physical and cognitive tests. These people have taken a writing test before any of them have started showing signs of Alzheimer’s disease. They have been asked to describe a boy standing on a shaky stool and reaching for a cookie jar, which is kept on the high shelf rack.
Experts have looked at the word usage patterns of the participants with the help of an artificial intelligence program, which find out subtle differences in language pattern. The tool has been able to identify a group of people who have been repetitive in their word usage at a time when all of them have been cognitively normal. These people have made a lot of spelling errors and unsuitably capitalized words. Experts have said that these people have used telegraphic language, which means the language which has simple grammatical formations and it has missing subjects and words such as the, is, and are. People from this group have been the ones who have been diagnosed with Alzheimer’s disease.
The AI program has been able to predict the onset of Alzheimer’s with 75 percent accuracy. Experts have said that it is important to have simple tests, which can identify early signs of such mental illnesses at a time when there are ways to slow down or prevent the disease. They have warned that without early intervention, people will end up developing a progressive brain disease. For many years, scientists have been observing speech and voice changes in people who have been showing signs of neurological ailments such as Alzheimer’s disease, Parkinson’s disease, ALS, bipolar disorder, schizophrenia, and dementia. However, the IBM report has come up with groundbreaking findings.
Dr. Michael Weiner, who has been doing research on Alzheimer’s disease at the University of California, has said that the IBM study is the first one that has included normal people and has been able to identify those people who might develop the disease several years later. Each mental disease leads to some unique changes in speech, which take place long before the diagnosis of the disease, said the experts. People, who are prone to develop such mental issues show apathy, a decline in judgment, self-control, and empathy as well. These parameters are quite hard to objectively measure. However, speech is different as changes in speech can be measured properly. Scientists have noted that at the beginning of the disease, patients have some changes in the pace of speech with pauses distributed randomly. The usage of words as well changes among these patients. They use fewer abstract words.
These changes are directly linked to alterations in the frontotemporal parts of the brain. These changes seem to be universal; they are unique to English only. IBM experts have looked for instances of flight of ideas in 34 other patients, which means when patients are off track while talking and whirling off their ideas in different directions. The team has looked at a group of 96 people. Around 59 of them have been identified with occasional delusions. Some of them have been diagnosed with schizophrenia and others have been healthy. The AI program has been able to identify people who have developed schizophrenia three years later with 85 percent accuracy. The findings of the study have been released in The Lancet Journal EClinicalMedicine.