14.2 INDIVIDUALIZED IDENTIFICATION AND TREATMENT RESPONSE PREDICTION OF FIRST-EPISODE DRUG-NAÏVE SCHIZOPHRENIA USING BRAIN FUNCTIONAL CONNECTIVITY: FROM SMALL TO BIG DATA

Identifying biomarkers in schizophrenia during the first episode without the confounding effects of treatment has been challenging. Leveraging these biomarkers to establish diagnosis and make individualized predictions of future treatment responses to antipsychotics would be of great value, but there has been limited progress. Applying machine learning algorithms, such as LASSO (least absolute shrinkage and selection operator), support vector machines (SVM) and deep neural networks, to brain imaging data is a promising approach to provide individualized identification of schizophrenia and prediction of treatment outcome, because the algorithms may capture the complex patterns within the imaging data as objective biomarkers for the disorder. In this study, we aim to investigate biomarkers in first-episode drug-naïve (FEDN) schizophrenia, identify the FEDN schizophrenia patients and predict their responses to antipsychotic treatment at an individual level using machine learning. In this prospective cohort study, FEDN schizophrenia patients and healthy controls were recruited at a baseline time point, and the patients were treated with risperidone for 10 weeks. The patients with FEDN schizophrenia were inpatients (N=43; Age 28.3±9.9 years; 24 females). Healthy controls (HC) were recruited from the community (N=29; Age 27.7±7.8 years; 16 females). Patients subjects were only included if they were diagnosed with schizophrenia during their first psychotic episode using the Structured Clinical Interview for DSM-IV (SCID). Subjects were excluded if they had a history of taking psychiatric medication, head trauma with residual effects, neurological disorders, and uncontrolled major medical conditions. HC were excluded if they had a history of any Axis I disorder according to SCID, had a first-degree relative with any Axis I disorder, or used psychoactive medication less than two-weeks before the study. The functional connections (FC) were derived using the mutual information and the correlations between the blood-oxygen-level dependent signals of the superior temporal cortex and other cortical regions acquired with the resting-state functional magnetic resonance imaging. The individualized identification and treatment response prediction were performed only using the data from the baseline. Cross-validated balanced accuracy of FEDN identification and correlation of predicted and actual symptom alleviation were considered as the main outcomes. We successfully identified the first-episode drug-naïve (FEDN) schizophrenia patients (balanced accuracy: 78.6%) and predicted their responses to antipsychotic treatment (r=0.69; p<0.0001; balanced accuracy for predicting responders and non-responders 82.5%). We also found that the mutual information and correlation FC was informative in identifying individual FEDN schizophrenia and prediction of treatment response, respectively. The methods and findings in this study could provide a critical step towards individualized identification and treatment response prediction in first-episode drug-naïve schizophrenia, which could complement other biomarkers in the development of precision medicine approaches for this severe mental disorder. We will discuss why further validation of such a preliminary study on large samples will be necessary and how we could link small data to big data. The potential benefits and challenges of machine learning studies in psychiatry based on the “big data” will be also discussed.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.