Extracting common processes across psychotherapies through AI

Psychological treatments are effective in addressing mental disorders. However, they often remain confined within two primary limitations: diagnostic categories and theoretical frameworks. This study aimed to investigate how artificial intelligence could help transcend these boundaries.

We recruited 125 psychotherapists. For each psychotherapeutic approach, we asked them to describe: (1) the proposed processes underlying psychopathology; (2) the techniques employed in treatment; and (3) potential factors contributing to resistance.

By applying topic modeling techniques on the psychotherapists' discourse, we identified 30 topics related to psychopathological or psychotherapeutic processes (e.g., emotional regulation). Notably, none of these topics were exclusively associated with a single psychotherapeutic approach, suggesting that psychotherapies share common underlying processes.

Artificial Intelligence techniques can help to identify shared processes across various psychotherapeutic approaches. This proof of concept suggests that further research utilizing these methods is warranted.