Researchers in Sweden have created an innovative AI tool capable of detecting signs of attention deficit hyperactivity disorder (ADHD) in adults by examining their medical records. This advanced model was developed using the medical histories of nearly 8,000 individuals, which included those with confirmed ADHD as well as patients diagnosed with various other psychiatric disorders.
The AI system identifies subtle trends within a patient's medical background. For instance, it may highlight complications during birth or diagnoses related to substance use, as these can sometimes indicate efforts to manage undiagnosed ADHD symptoms. Notably, the tool proved most effective in predicting ADHD up to six months prior to a formal diagnosis. In a study involving 800 participants, equally divided between those with and without ADHD, the model successfully identified 80% of ADHD cases while accurately ruling out 77% of cases without the disorder.
However, there are significant limitations to consider. The model faces challenges in differentiating ADHD from conditions such as depression and anxiety, where symptoms often overlap. This indicates that additional refinement is necessary before the tool can be implemented in clinical environments, where patients frequently present with multiple overlapping diagnoses.
Informational material. 18+.
