When the mental health of patients is discussed with psychologists or doctors, they are frequently asked to rate their feelings using a rating scale.
Image Credit: Mostphotos/Yuri Arcurs
At present, this is how anxiety and depression have been diagnosed. But a new study performed at Lund University in Sweden has displayed that enabling patients to explain their experience by making use of their words can be highly accurate and favored by the patients. The scientists of Lund have come up with an AI tool that could assist doctors in examining their patients’ answers.
The newly performed study has shown that patients and clinicians vary on their preference, as clinics usually prefer rating scales (for example, little interest in doing things: not at all, sometimes, frequently, daily) whereas patients preferred language models (for example, explaining the mental health of patients).
The study has been reported in the PLOS ONE journal.
The scientists surveyed a team of 150 patients with self-diagnosed anxiety or depression, and posed the same questions to a control team of 150 other participants.
The responses show that making use of free language is particularly significant as it is viewed as the natural method for people to communicate their feelings. This allows the patient to decide what is most important and to be understood.
Especially, patients responded that language allows a highly accurate explanation of their feeling, which cannot be made possible with rating scales. As per the patients, the only benefit of rating scales is that they are simpler and quicker to be answered, but they lack the ability for more individualized care.
AI can be highly precise in quantifying emotions in language compared to rating scales. In the past few years, a huge advance has been made with natural language processing. This is done in such a way that anxiety and depression could be evaluated with high validity depending on open-ended questions that are further examined with AI techniques.
Previously, we were limited to rating scales where patients themselves had to convert their feelings into numbers. Today this is not necessary as this assessment can be done by AI.
Sverker Sikström, Study Lead and Psychology Researcher, Lund University
Sikström has come up with an AI product allowing patients to freely explain their mental health in texts, which are further rendered by an AI engine.
The outcomes are shown in decision support material that could further be utilized by doctors and psychologists. This product is special as it requests direct open-ended questions from patients.
We have received a lot of interest in this from health care providers, given that we now have the technology to support patients individually, yet in an efficient manner.
Sverker Sikström, Study Lead and Psychology Researcher, Lund University
Journal Reference
Sikström, S., et al. (2023) Precise language responses versus easy rating scales—Comparing respondents’ views with clinicians’ belief of the respondent’s views. Plos One. doi.org/10.1371/journal.pone.0267995.