Ewbank MP, Cummins R, Tablan V, et al. - The authors carried out this quality improvement study using a deep learning model applied to a large-scale clinical data set of cognitive behavioral therapy (CBT) session transcripts to create a quantifiable measure of treatment delivered and to explore the link between the quantity of each aspect of therapy delivered and clinical results. All data were collected in England between June 2012 and March 2018 from patients receiving internet-enabled CBT for the treatment of a mental health disorder. In the final analysis, treatment sessions were involved from a total of 14,899 individuals aged between 18 and 94 years (median age, 34.8 years). A deep learning model was trained to categorize therapist pronouncements automatically from about 90,000 hours of internet-enabled CBT. There was a connection between clinical results in psychotherapy and the content of therapist utterances. Such outcomes support the concept that CBT methods of change can improve the presenting symptoms of patients. Applying deep learning to large clinical data sets can provide valuable insights into psychotherapy, inform new treatment development, and help standardize clinical practice.