AI model predicts schizophrenia, bipolar progression

Machine learning models trained on routine clinical data from electronic health records (EHRs) predicted diagnostic progression to schizophrenia or bipolar disorder in patients undergoing treatment in psychiatric services for other mental illnesses, according to a new cohort study published in JAMA Psychiatry.
The cohort study, Predicting Diagnostic Progression to Schizophrenia or Bipolar Disorder via Machine Learning, determined that “detecting progression to schizophrenia through machine learning based on routine clinical data is feasible, which may reduce diagnostic delay and duration of untreated illness.”
The cohort included 24, 449 patients aged 15 to 60 years with a total of 398,922 outpatient contacts to the Psychiatric Services of the Central Denmark Region between 2013 and 2016.
Study authors wrote that progression to schizophrenia was predicted with greater accuracy than bipolar disorder, which proved to be a more challenging target.
The authors concluded that it is possible to forecast diagnostic transition to schizophrenia and bipolar disorder from routine clinical data extracted from EHRs, with schizophrenia being easier to predict than bipolar disorder.
“The model predicting schizophrenia performed substantially better than the model predicting bipolar disorder, likely due to heterogenic clinical manifestations of the latter,” the authors wrote.
“These findings suggest that text-based features from clinical notes show great promise for improving prediction of psychiatric outcomes.”
The study did have limitations.
“First, the data were restricted to patients receiving psychiatric treatment and did not contain information from primary care,” the authors wrote.
“Consequently, the prediction models are primarily useful for patients who are progressing from less severe mental disorder to schizophrenia or bipolar disorder. Patients whose initial contact to the psychiatric services is due to clinical suspicion of schizophrenia or bipolar disorder will not see additional benefits from the model.”
THE LARGER TREND
In 2024, Amous Adler, founder and president of MEMOTEXT, discussed in an interview with HIMSS TV, how the company builds and commercializes digital tools, its digital app for schizophrenia and psychosis patients, and digital therapeutics in Canadian versus U.S. markets.
That same year, researchers from South Korea built machine learning-based models that can predict mood episodes using only sleep and circadian rhythm data from wearable devices.
In a study, which findings were published in Nature’s npj Digital Medicine journal, the research team first collected and analysed 429 days worth of sleep-wake data generated from Fitbit of 168 Korean patients with mood disorders, including major depression and bipolar disorders.
In 2022, Pharma company Boehringer Ingelheim and Click Therapeutics expanded their partnership to develop and commercialize another prescription digital therapeutic for patients with schizophrenia. The deal netted Click up to $460 million, plus tiered royalties. The alliance was built upon an earlier collaboration that was announced in 2020 that developed CT-155, an initial therapeutic for schizophrenia.