Artificial intelligence discovers Parkinson’s years earlier than any doctor

In principle, Parkinson’s disease cannot be cured. However, there are several treatment options that can at least slow the progression of the disease. However, in the last decade, only minimal progress has been made in research, despite high expectations. This can be explained by a special feature of the disease. Because this is usually not diagnosed until the typical motor impairments have already occurred. At the moment, however, the disease is already relatively advanced. In principle, of course, the same applies here: the earlier the disease is detected, the easier the treatment. The problem: Parkinson’s disease can only be detected with complex and stressful diagnostics. The methods available to date are therefore not suitable for routine diagnostics. However, this could now change with an approach developed by MIT researchers.

The change in breathing occurs relatively early

They draw on a discovery by the eponymous researcher Dr. James Parkinson made as early as 1817. It was then that he discovered that there was a connection between the disease and the patient’s breathing. Until now, however, this has had no consequences for diagnosis or treatment. But now the researchers have developed a neural network that specifically analyzes breathing. Over time, the artificial intelligence was then trained to diagnose the disease based on typical patterns and changes in breathing. In a large-scale study, a total of 7,687 people were examined with the new method, of whom 757 were already demonstrably suffering from Parkinson’s disease. In fact, the neural network was able to identify the sick person by analyzing their breathing during sleep. This could be a major breakthrough, as it can identify the disease before more severe symptoms appear.

The success of the treatment can be tracked in detail

It is also very beneficial that no one has to go to a sleep lab to have their breathing analyzed. Instead, the researchers have developed a special device that you can also use in your own four walls. Simply put, radio signals are broadcast. The breathing pattern can be determined on the basis of the reflections. This is then made available to the neural network for analysis. In this way, the disease can not only be diagnosed earlier. Instead, it is also possible to observe the progression of the disease in detail. This is important, for example, to see to what extent a certain therapy is still successful or whether a new approach needs to be tried out. Clinical studies testing new treatment methods can also be monitored more easily and in greater detail this way. Ideally, this will ultimately lead to better treatment options.

Leave a Comment

Your email address will not be published. Required fields are marked *