New research finds that mental health symptoms may be too interconnected for any diagnostic system to separate people into distinct categories.
Scientists analyzed over 10,000 personal stories from people with mental health diagnoses using artificial intelligence (AI) and found that symptoms overlap so extensively that distinguishing between different conditions becomes nearly impossible. The findings suggest the problem isn’t just with how the DSM was created, but with the concept of sorting complex human experiences into neat diagnostic boxes.
“We infer that whether humans attempt to categorise mental illnesses or an AI, the result is that the categories of mental disorders will not be unique enough to be able to distinguish one service seeker from another,” write the researchers, led by Chandril Chandan Ghosh from the School of Psychology at Queen’s University Belfast.
The study strikes at the core of debates over psychiatric diagnosis. For decades, the DSM and ICD have divided mental health into hundreds of categories, even as evidence shows symptoms often overlap and no biological markers clearly separate one disorder from another. Alternatives such as HiTOP and RDoC propose dimensional models that track traits across continua rather than fixed categories. The new analysis suggests the difficulty runs deeper. Whether designed by expert committees or by algorithms, attempts to sort human distress into distinct boxes fail to capture the complexity of lived experience. Drawn from more than 10,000 patient accounts, the findings underscore how people’s actual symptoms rarely match the tidy boundaries of diagnostic systems.
You've landed on a MIA journalism article that is funded by MIA supporters. To read the full article, sign up as a MIA Supporter. All active donors get full access to all MIA content, and free passes to all Mad in America events.
Current MIA supporters can log in below.(If you can't afford to support MIA in this way, email us at [email protected] and we will provide you with access to all donor-supported content.)
Do you have a sense of the metabolism rate of Ai in comparison to human? For the challenge to be healthy and attempt to give your best in the daily experience of “competitive employment” may ignore structural issues in the socio-technologies apart from being human. How to really understand and integrate the complexities with Ai and without? For many may not even have access to the e-world within the culture of being in an economy? We could be undervalued or not even valued in the traditional analytics.
Report comment
This is, of course, completely to be expected and validates what Szaz and others have been saying for decades. Wanna take bets on whether this has the slightest impact on the use of the DSM to “diagnose” people with mental “diseases?”
Report comment