Internet Searches Could Reveal Neurodegenerative Disorders, According To Study

A study carried out by Duke University (United States), in collaboration with Microsoft, has provided clues to discover neurodegenerative disorders at a distance. The data has been collected by capturing and analyzing certain signals that appear when users search the Internet.

In the study, the authors intend to provide evidence to detect some neurodegenerative disorders, such as Parkinson’s disease or Alzheimer’s. So far, the results seem promising, but this has not avoided controversy over the way in which the information has been collected.

This is because users do not know that they are collecting information when they search on the Internet. Although the study is motivated by a laudable cause, the truth is that people should be better informed about this mode of data capture.

Neurodegenerative disorders

Neurodegenerative disorders are common in today’s world. One of the great problems to deal with these evils correctly is the fact that they are usually detected when it is too late. Usually the symptoms  initials are very subtle and it is usual that they are confused with other conditions.

When neurodegenerative disorders are detected early, the prognosis is much more favorable. Treatment can be started early and this goes a long way in slowing down illnesses and making them less severe.

Thinking about it, scientists at Duke University concluded that Internet searches could provide valuable data. Several of the actions that are carried out during such searches can be classified as “digital phenotypes”. That is, visible characteristics that appear when an individual interacts with his environment.

The study in Internet searches

To carry out this study, the researchers partnered with Microsoft. The objective was to analyze certain actions of users when they search on the Internet. Specifically, users of the Bing platform were analyzed , but none of them were notified that they were participating in a study.

In total, data from 31.3 million Bing users was analyzed . The research was carried out over 18 months, between September 2015 and February 2017. The study focused on the follow-up of four variables:

• Searches for information on the symptoms of neurodegenerative disorders.
• Shaky and awkward movements with the cursor.
• Repetitive clicking.
• Searches associated with risk factors such as exposure to toxins and advanced age.

Follow-up was carried out routinely during the 18 months of the study. In the end, the information of all the variables was crossed and 700 cases of potential users who had received a recent diagnosis of Parkinson’s disease were found.

The conclusions of the study

To draw definitive conclusions, the researchers developed an algorithm that made it possible to cross the four variables studied and establish the probability that the user had Parkinson’s disease. This pathology was chosen because it is one of the most easily detectable degenerative disorders with the technique used.

In the end, these results were contrasted with the 700 previously selected cases. The method was effective in detecting Parkinson’s disease in 94.2%, 83.1%, 42% and 34.6%, for each variable, respectively. Likewise, it was found that the number of false positives was very low compared to traditional diagnostic methods.

The researchers think that in the near future it will be possible to use similar techniques for the early diagnosis of other neurodegenerative disorders, such as Alzheimer’s disease.

The controversy

The Duke University and Microsoft study sparked a lot of controversy among scientists. Some think that no ethical principle was violated, since at no time were the personal data of those who had participated in the study unknowingly disclosed.

Others believe that this is not enough, but that it is necessary to have the consent of the users so that they can participate in any research. They think that if this is tolerated, in a short time similar studies could appear that point to more compromising issues, which would constitute a violation of privacy. The debate is open.