ABSTRACT
Psychiatric electronic health records (EHRs) present a distinctive challenge in the domain of ML owing to their unstructured nature, with a high degree of complexity and variability. This study aimed to identify a cohort of patients with diagnoses of a psychotic disorder and posttraumatic stress disorder (PTSD), develop clinically-informed guidelines for annotating these health records for instances of traumatic events to create a gold standard publicly available dataset, and demonstrate that the data gathered using this annotation scheme is suitable for training a machine learning (ML) model to identify these indicators of trauma in unseen health records. We created a representative corpus of 101 EHRs (222,033 tokens) from a centralized database and a detailed annotation scheme for annotating information relevant to traumatic events in the clinical narratives. A team of clinical experts annotated the dataset and updated the annotation guidelines in collaboration with computational linguistic specialists. Inter-annotator agreement was high (0.688 for span tags, 0.589 for relations, and 0.874 for tag attributes). We characterize the major points relating to the annotation process of psychiatric EHRs. Additionally, high-performing baseline span labeling and relation extraction ML models were developed to demonstrate practical viability of the gold standard corpus for ML applications.
ABSTRACT
BACKGROUND: The hippocampus is a heterogeneous structure composed of biologically and functionally distinct subfields. Hippocampal aberrations are proposed to play a fundamental role in the etiology of psychotic symptoms. Bipolar disorder (BPD) has substantial overlap in symptomatology and genetic liability with schizophrenia (SZ), and reduced hippocampal volumes, particularly at the chronic illness stages, are documented in both disorders. Studies of hippocampal subfields in the early stage of BPD are limited and cross-sectional findings to date report no reduction in hippocampal volumes. To our knowledge, there have been no longitudinal studies of BPD evaluating hippocampal volumes in the early phase of illness. We investigated the longitudinal changes in hippocampal regions and subfields in BPD mainly and in early stage of psychosis (ESP) patients more broadly and compared them to those in controls (HC). METHODS: Baseline clinical and structural MRI data were acquired from 88 BPD, from a total of 143 ESP patients, and 74 HCs. Of those, 66 participants (23 HC, 43 patients) completed a 12-month follow-up visit. The hippocampus regions and subfields were segmented using Freesurfer automated pipeline. RESULTS: We found general baseline deficits in hippocampal volumes among BPD and ESP cohorts. Both cohorts displayed significant increases in the anterior hippocampal region and dentate gyrus compared with controls. Additionally, antipsychotic medications were positively correlated with the posterior region at baseline. CONCLUSION: These findings highlight brain plasticity in BPD and in ESP patients providing evidence that deviations in hippocampal volumes are adaptive responses to atypical signaling rather than progressive degeneration.