Brian Zaboski
Has grant
Assistant Professor
Research Interests
Make your next move.
Prepare your application
Review drafts before you send them.
Explore related searches
Recent Grants
Grant: Close
Using Convolutional Neural Networks and a Parcel-wise Approach to Predict Treatment Response to Selective Serotonin Reuptake Inhibitors in Obsessive-Compulsive Disorder
Open Date: 2023-01-01
Close Date: 2026-01-01
Articles (27)
Objective: Identifying obsessive-compulsive disorder (OCD) using brain data remains challenging. Resting-state electroencephalography (EEG) offers an affordable and noninvasive approach, but identifying predictive signals in EEG data has met with little success, even with the application of traditional machine learning methods. We explored whether convolutional neural networks (CNNs) applied to EEG time-frequency representations can distinguish individuals with OCD from healthy controls. Method: We collected resting-state EEG data from 20 unmedicated participants (10 with OCD, 10 healthy controls). Four-second EEG segments were transformed into time-frequency representations. We then trained a 2D CNN using a leave-one-subject-out cross-validation framework to perform subject-level classification and compared its performance to a more traditional support vector machine (SVM) approach. Next, using multimodal fusion, we examined whether adding clinical and demographic information improved classification. Results: The CNN classifier achieved high subject-level performance, distinguishing individuals with an accuracy of 85.0% and an area under the curve (AUC) of 0.88. This significantly outperformed the SVM baseline, which performed no better than chance (45.0% accuracy, AUC: 0.47). A subsequent multimodal analysis revealed that clinical and demographic variables did not contribute any additional independent information. Conclusion: CNNs applied to resting-state EEG show promise for identifying OCD, outperforming traditional machine learning methods. These findings highlight the potential of deep learning to uncover complex, diagnostically relevant patterns in neural data. While limited by sample size, this work supports further investigation into multimodal models for psychiatric classification, warranting replication in larger, more diverse samples.
Year:
2026
Year:
2026
Year:
2026
Year:
2026
Year:
2026
Background Existing large language model (LLM) evaluations rely on accuracy benchmarks that fail to capture whether models reason well while making diagnoses. Studies that do analyse reasoning focus on post hoc explanations accompanying model outputs rather than distinct, clinician-visible artifacts such as detailed reasoning traces. This creates a translational gap in domains such as psychiatry, where diagnosis relies on narrative interpretation, diagnostic reasoning, and clinical judgment under uncertainty. Methods We conducted a mixed-methods evaluation of four state-of-the-art LLMs using a clinician-curated dataset of 196 psychiatric case vignettes, including 135 published cases and 61 novel clinician-authored vignettes. Diagnostic accuracy was assessed using multiple metrics (top-1 accuracy, top-5 accuracy, recall@5, and mean reciprocal rank) based on ranked lists of five differential diagnoses per vignette. Clinical reasoning quality was evaluated on a randomly selected subset of 30 vignettes through clinician assessment of model-generated diagnostic reasoning traces alongside qualitative commentary from board-certified psychiatrists. We examined the association between clinician-rated reasoning quality and diagnostic correctness and included an illustrative comparison with psychiatry residents. Findings Clinician-rated diagnostic reasoning quality was strongly associated with diagnostic correctness in mixed-effects logistic regression analyses (β = 1·80; p < 0·001), whereas data extraction quality alone was not. Across the full vignette set, models demonstrated moderate to high diagnostic accuracy. The highest-performing model achieved a top-5 accuracy of 0·801 and also received the highest clinician-rated reasoning scores. In an illustrative comparison, model diagnostic accuracy fell within the range observed for psychiatry residents. Interpretation Diagnostic reasoning quality captures clinically meaningful variation in LLM performance beyond accuracy metrics. Psychiatry may represent a stringent testbed for evaluating reasoning in narrative-driven clinical domains. Evaluation frameworks for LLM-based clinical decision support should incorporate structured assessment of reasoning processes, not accuracy alone. Research in Context Evidence before this study We searched PubMed and Scopus for studies evaluating large language models for psychiatric diagnosis and/or differential diagnosis from text vignettes. Searches were run from database inception to February 6, 2026, using terms including (“large language model” OR LLM “artificial intelligence” OR “generative AI” OR AI OR ChatGPT OR GPT OR Claude OR Gemini OR DeepSeek OR Llama) AND (psychiatr* OR mental OR DSM OR “differential diagnosis” OR diagnos*) AND (vignette OR case OR “case report”). We included empirical studies that evaluated model diagnostic outputs using psychiatric cases/vignettes and excluded editorials/commentaries and studies that did not report case-level diagnostic performance. We did not formally assess risk of bias/quality; studies were heterogeneous in vignette sources, model access, and outcome definitions, precluding quantitative pooling. Prior studies have typically examined small vignette sets, focused on narrow diagnostic domains, evaluated single models, or relied primarily on outcome-based accuracy metrics. When diagnostic reasoning has been assessed, it has usually been inferred from post hoc explanations accompanying model outputs rather than evaluated as a distinct, clinician-visible artifact. Clinician-grounded evaluations of diagnostic reasoning across multiple contemporary models remain limited. Added value of this study This study provides a large-scale, clinician-grounded evaluation of diagnostic accuracy and diagnostic reasoning quality across four contemporary large language models using a diverse dataset of psychiatric case vignettes. Rather than relying solely on outcome-based explanations, we directly evaluated model-generated d
Year:
2026
Year:
2025
Year:
2025
Misinformation regarding the symptoms and treatment of obsessive-compulsive disorder (OCD) has persisted for decades, influencing public perceptions and propagating misconceptions. Content posted on social media regarding mental health conditions, like OCD, impacts the health information ecosystem, as consumers can find it difficult to distinguish factual information from misleading generalizations. This study examines the quality of OCD-related content on TikTok, with a focus on the prevalence of misinformation and its potential impact on viewers. We analyzed the top videos tagged with #OCD, categorizing them into two groups: those directly discussing OCD and those portraying stereotypical behaviors associated with OCD without dialogue. Engagement metrics were analyzed using negative binomial regression to compare engagement levels within and between groups. Among videos discussing OCD, only 18 percent contained accurate information, while 29 percent were misleading, indicating a significant lack of reliable material on the platform. Additionally, health care providers were underrepresented among content creators, with most videos coming from nonprofessional sources. Analyses of stereotype-driven content revealed these videos were more frequently viewed, liked, and shared than videos that provided accurate information about OCD. This suggests that users are more engaged with content that reinforces stereotypes rather than educational content about the disorder. The study highlights the need for more evidence-based information on social media to counteract the stereotypes and misconceptions about OCD, leading to stigmatization and misunderstanding of the condition.
Year:
2025
The Subjective Units of Distress Scale (SUDS) is a widely used self-report measure clinicians rely on during exposure and response prevention (ERP) to monitor progress, guide exposure pacing, and assess intervention efficacy. However, despite its ubiquity in clinical and research settings, foundational investigations of its psychometrics are often atheoretical, fail to evaluate its longitudinal properties, and lack a rigorous construct validation framework. This paper addresses these shortcomings by evaluating the SUDS as a measure of state negative affective intensity using the Strong Program of Construct Validation. Our evaluation demonstrates that the SUDS suffers from significant psychometric weaknesses, including construct underrepresentation, construct irrelevance, poorly defined measurement occasions, and structural limitations, challenging its validity as a precise measure of subjective distress. These limitations have crucial implications for clinical practice, potentially leading to misinterpretations of patient distress and compromising treatment decisions. We discuss these clinical implications, highlight them with a brief clinical vignette, outline a research roadmap for potential improvement using modern psychometric methods, and provide practical recommendations for clinicians currently using the SUDS. Given these validity concerns, caution is warranted when interpreting SUDS scores in both clinical and research contexts until its psychometric properties are more robustly established and understood.
Year:
2025
Year:
2025
Year:
2025
Year:
2025
Year:
2025
Year:
2025
Year:
2025
Year:
2025
Year:
2025
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Year:
2024
Year:
2023
Year:
2023
Year:
2023
Year:
2023
Collaborators (12)
Walter Mathis
Yale University
Matteo Malgaroli
Assistant Professor
NYU School of Medicine
Sarah Fineberg
Assistant Professor
Yale University School of Medicine
Salih Selek
Professor
University of Texas System
Helen Pushkarskaya
Assistant Professor
Yale University
Qingyu Chen
Assistant Professor
National Institutes of Health
Hua Xu
Yale University
Jiyeong Kim
Assistant Professor
University of California, San Francisco
Christopher Pittenger
Professor (secondary)
Yale University
Gregory Muller
Assistant Professor
University of Texas at Austin
Carolyn Rodriguez
Professor and Associate Dean
Stanford University
Patrick D. Skosnik
Yale University School of Medicine
Make your next move.
Prepare your application
Review drafts before you send them.