Neurological Correlates of Treatment-Resistant Schizophrenia and Machine Learning Applications: A Structural Approach to Differentiation
Bibliographic record
Abstract
Treatment-resistant schizophrenia (TRS) is a severe and chronic subtype of schizophreniacharacterized by persistent positive symptoms such as hallucinations and delusions, despite adequate antipsychotic treatment. Although clinically significant, the heterogeneity in schizophrenia’s pathophysiology has hindered the discovery of reliable diagnostic biomarkers for TRS, leaving trial-and-error pharmacotherapy as the primary method of differentiation. This thesis investigates the magnetic resonance imaging (MRI) correlates of TRS and their potential for advancing classification through machine learning algorithms. First, we examined intrinsic cortical curvature (ICC) as a novel measure of gyrification to assess structural differences in TRS. We found ICC to be particularly sensitive in detecting TRS-related abnormalities compared to treatment-responsive (TxR) patients and healthy controls, and more robust than traditional structural metrics. These results were further supported by a multi-site study of local gyrification index (LGI) and surface area, reinforcing the value of gyrification measures for distinguishing TRS. We then explored disease progression using the Subtype and Stage Inference (SuStaIn) model, which identified distinct structural trajectories and stages of illness that correspond with treatment profiles. These morphological findings aligned with our study of neurometabolite differences using proton magnetic resonance spectroscopy (¹H-MRS), which revealed correlations between elevated glutamate levels and cortical thinning, suggesting a possible excitotoxic process underlying TRS-related structural changes. Based on this, we propose a mechanistic model for clozapine efficacy in TRS, involving glutamatergic and astrocytic modulation. Together, these studies offer converging evidence for structural and neurochemical alterations specific to TRS, contributing to the growing body of research supporting biologically informed diagnostic tools in psychiatry. Finally, we leverage these insights to develop a machine learning model for TRS detection and lay the groundwork for the brain imaging automatic classification (BRAINIAC) project, a deep learning architecture designed for automated classification of brain scans.
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.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.004 | 0.016 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.003 | 0.002 |
| Science and technology studies | 0.000 | 0.001 |
| Scholarly communication | 0.002 | 0.001 |
| Open science | 0.000 | 0.001 |
| Research integrity | 0.001 | 0.001 |
| Insufficient payload (model declined to judge) | 0.002 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".