Anchoring Data-Driven Staging in Clinical Staging: From Hippocampus to Functioning in Psychosis
Poster No:
510
Submission Type:
Abstract Submission
Authors:
Jana Totzek1, Matthew Danyluik1, Mallar Chakravarty1, Jai Shah1, Ridha Joober1, Ashok Malla1, Delphine Raucher-Chéné1, Alexandra Young2, Martin Lepage1, Katie Lavigne1
Institutions:
1McGill University, Montreal, Quebec, 2University College London, London, United Kingdom
First Author:
Co-Author(s):
Introduction:
Clinical staging models of mental illness are primarily anchored in illness history and suggest declining cognition and functioning and increasing symptoms from familial high-risk to clinical high-risk, first-episode psychosis (FEP) and multi-episode psychosis (MEP) (McGorry et al, 2007). While these models hold immense clinical utility, longitudinal validation across stages is scarce (Martinez-Cao et al, 2022). Recent advances in machine learning can address this gap by inferring the longitudinal development of markers based on cross-sectional data. Through this approach, we previously identified distinct progression patterns from brain to behaviour in psychosis, with the hippocampus leading the progression in one patient subtype (Totzek et al, 2024). As those data-driven approaches base their inferences purely on momentary performance, the clinical utility of these findings over the course of the illness remains to be determined. We aimed to enhance the clinical utility of our multiscale model of psychosis (from hippocampal volume to cognition, symptoms and functioning) by anchoring the data-driven inference of machine learning in traditional clinical staging models while also addressing the heterogeneity of impairments in psychosis (Totzek et al, 2024).
Methods:
We sampled data from two cross-sectional datasets spanning early and multi-episode psychosis. The first dataset included 29 individuals at familial high-risk, 38 individuals at clinical high-risk, 54 individuals with FEP, and 53 matched non-clinical controls. The second dataset included 108 individuals with MEP, and 66 matched non-clinical controls. Global cognition was derived by averaging the domains of the Cogstate Schizophrenia Battery (Pietrzak et al, 2009). Negative and positive symptoms were assessed through the Scales for the Assessment of Negative and Positive Symptoms (Andreasen, 1983,1984), and functional outcomes through the Social and Occupational Functional Assessment Scale (Goldman et al, 1992). We then implemented the MAGeT-algorithm (Chakravarty et al, 2013), resulting in 18 hippocampal subfields and adjacent white matter volumes, which we averaged for one measure of hippocampal volume. The patient data was z-scored relative to control values (and relative to familial high-risk for positive symptoms) and used as input markers for the Subtype and Stage Inference (SuStaIn) analysis (Young et al, 2018). SuStaIn is a machine learning algorithm which merges clustering and disease progression modeling, allowing us to infer the progression of these multiscale markers across heterogeneous subtypes of psychosis. Results were 10-fold cross-validated.
Results:
SuStaIn identified Subtype 1 (n=112) which progressed from reduced cognition towards reduced hippocampal volume, increased negative symptoms and poorer functioning, and higher positive symptoms (Fig. 1&2a). Individuals in data-driven stage 0 were grouped into a normal appearing subtype (n=117). Subtype 1 scored poorer than the normal appearing subtype on all markers (p's<.001). When comparing clinical and data-driven stages, we observed a tendency for individuals at familial and clinical high-risk to be sorted into earlier data-driven stages, while individuals with a FEP and MEP were predominantly sorted into later data-driven stages (Fig. 2b), although not statistically significant (p=.521).
·Progression of Subtype 1
·Means and Clinical Stages per Data-Driven Stages of Subtype 1
Conclusions:
These findings allow for a first merging of clinical staging models and novel machine learning approaches by anchoring data-driven predictions in clinical history. Our data-driven results provide support for clinical staging models progressing from cognition toward negative symptoms and functioning (McGorry et al, 2007) and underline the importance of hippocampal volume in the early stages of psychosis. Combining clinical and data-driven stages thus provides our multiscale model with enhanced clinical utility and paves the way toward precision psychiatry while establishing the importance of the hippocampus in subclinical predictions.
Disorders of the Nervous System:
Psychiatric (eg. Depression, Anxiety, Schizophrenia) 1
Modeling and Analysis Methods:
Classification and Predictive Modeling 2
Neuroinformatics and Data Sharing:
Informatics Other
Keywords:
ADULTS
Cognition
Computational Neuroscience
Machine Learning
MRI
Psychiatric Disorders
Schizophrenia
1|2Indicates the priority used for review
Abstract Information
By submitting your proposal, you grant permission for the Organization for Human Brain Mapping (OHBM) to distribute your work in any format, including video, audio print and electronic text through OHBM OnDemand, social media channels, the OHBM website, or other electronic publications and media.
The Open Science Special Interest Group (OSSIG) is introducing a reproducibility challenge for OHBM 2025. This new initiative aims to enhance the reproducibility of scientific results and foster collaborations between labs. Teams will consist of a “source” party and a “reproducing” party, and will be evaluated on the success of their replication, the openness of the source work, and additional deliverables. Click here for more information. Propose your OHBM abstract(s) as source work for future OHBM meetings by selecting one of the following options:
Please indicate below if your study was a "resting state" or "task-activation” study.
Healthy subjects only or patients (note that patient studies may also involve healthy subjects):
Was this research conducted in the United States?
Were any human subjects research approved by the relevant Institutional Review Board or ethics panel? NOTE: Any human subjects studies without IRB approval will be automatically rejected.
Were any animal research approved by the relevant IACUC or other animal research panel? NOTE: Any animal studies without IACUC approval will be automatically rejected.
Please indicate which methods were used in your research:
For human MRI, what field strength scanner do you use?
Which processing packages did you use for your study?
Provide references using APA citation style.
Andreasen, N. C. (1984). The Scale for the Assessment of Positive Symptoms (SAPS). The British journal of psychiatry, 155(S7), 49-52.
Chakravarty, M. M. et al. (2013). Performing label‐fusion‐based segmentation using multiple automatically generated templates. Human brain mapping, 34(10), 2635-2654.
Goldman, H. H. et al. (1992). Revising axis V for DSM-IV: a review of measures of social functioning. Am J Psychiatry, 149, 9. doi:https://ssrn.com/abstract=2143992
Martinez-Cao, C. et al. (2022). Is it possible to stage schizophrenia? A systematic review. Translational Psychiatry, 12(1), 197.
McGorry, P. D. et al. (2007). Clinical staging: a heuristic model for psychiatry and youth mental health. Medical Journal of Australia, 187(S7), S40-S42.
Pietrzak, R. H. et al. (2009). A comparison of the CogState Schizophrenia Battery and the Measurement and Treatment Research to Improve Cognition in Schizophrenia (MATRICS) Battery in assessing cognitive impairment in chronic schizophrenia. Journal of clinical and experimental neuropsychology, 31(7), 848-859.
Totzek, J. F. et al. (2024). Longitudinal inference of multiscale markers in psychosis: from hippocampal centrality to functional outcome. Molecular psychiatry, 1-10. doi:https://doi.org/10.1038/s41380-024-02549-x
Young, A. L. et al. (2018). Uncovering the heterogeneity and temporal complexity of neurodegenerative diseases with Subtype and Stage Inference. Nature communications, 9(1), 1-16. doi:https://doi.org/10.1038/s41467-018-05892-0