Introduction

Key features of psychotic disoders include delusions, hallucinations, disorganized thinking, grossly disorganized or abnormal motor behavior, and negative symptoms (Arciniegas, 2015). Typically, the diagnosis is made according to the 5th edition of the Diagnostic and Statistical Manual of Mental Disoders (American Psychiatric Association, 2013) and interviews, lacking the use of objective biomarkers. A biomarker is defined as a ” characteristic that is objectively measured and evaluated as an indicator of normal biological processes, pathogenic processes, or pharmacologic responses to a therapeutic intervention. In practical terms, a biomarker refers to a broad subcategory of medical signs – that is, objective indications of medical state observed from outside the patient – which can be measured accurately and reproducibly” (Correll et al., 2010). Since the above mentioned symptoms are not unique to the disorder and can be seen across a wide range of mental disorders, plus the disease exhibits a high degree of heterogeneity and individual variability, biomarkers associated with psychotic disorders could provide relief in the diagnostic process.

Machine learning methods (for example, support vector machines, random forest, logistic regression) (Rezaii et al., 2019, Stamate et al., 2019) and magnetic resonance imaging (MRI) (Mikolas et al., 2018, Qi-yong Gong et al., 2018) become apparent as a new approach to support the diagnosis of mental disorders. Machine learning (ML) is a domain of artificial intelligence (AI) that makes computer algorithms feasible to learn patterns by studying data directly without being explicitly programmed (Noble, 2006). Within the realm of medicine, ML has increasingly been applied with regard to diagnosis and outcome prediction (Azimi et al., 2015). Neuroimaging measures assessed with MRI could serve as biomarkers making it feasible to evaluate prognostic accuracy (5). Many studies have employed machine learning (ML) algorhitms to identify or diagnose psychotic disorders or schizophrenia based on magnetic resonance imaging (MRI) data (Dwyer et al.,2018, Lei et al., 2019, Vieira et al., 2020).

Studies of psychotic disorders using structural MRI reveal widespread neuroanatomic abnormalities in cortical thickness and total brain measures (Shenton et al., 2001). For example, the results of the studies by Kuperberg et al. (2003) and Rimol et al. (2010) reveal reduced cortical thicknessn in the frontal and temporal lobes but also to some extent in parietal and occipital regions. Furthermore, Del Casale et al. (2021) compared patients with first episode psychosis and multi-episode schizophrenia with healty controls and found that the multi-episode group showed cortical thickness reduction in left temporal and parietal, and right temporal, parietal, occipital, and hippocampal cortices. Another study by Asmal et al. (2018) examined frontal cortical thickness in first-episode schizophrenia with healthy controls. The results reveal reduced cortical thickness for a number of frontal regions such as left medial orbitofrontal, left superior frontal, left frontal pole, right rostral middle frontal, right lateral orbitofrontal and right superior frontal regions in patients with first-episode schizophrenia. Another study by Jung et al. (2011) compared corical thickness in individuals at ultra-high-risk (UHR) for psychosis, patients with schizophrenia and healthy controls. The authors found out that UHR group showed significant cortical thickness reduction in the prefrontal cortex, anterior cingulate cortex, inferior parietal cortex, parahippocampal cortex, and superior temporal gyrus compared with healthy control subjects. When compared schizophrenia group and UHR group, reductions in the above mentioned regions plus posterior cingulate cortex, insular cortex, and precentral cortex were found in the schizophrenia group. All these studies show that reduced cortical thickness is a common finding in psychotic disorders.

However, functional MRI and diffusion tensor imaging studies of psychotic disorders show aberrant functional and structural connectivity, stressing networks and dysfunctional connectivity rather than brain regions acting in isolation (Meyer- Lindenberg et al., 2005, Bassett et al., 2008, Lynall et al., 2010). For example, resting state functional MRI (rsfMRI) studies reveal reduced connectivity within and between large-scale cortical networks, especially those involving frontal and temporal cortex (Baker et al., 2019).Studies reveal that disprutions in strucutral connectivity related to psychotic disorder as measured by Diffusion Tensor Imaging (DTI) show brain-wide reductions in fractional anisotropy (FA) (Kubicki et al., 2007). Addiotionally, mean diffusivity (MD) has been found to be increased in several brain regions, including the frontal lobe, temporal lobe, and corpus callosum, indicating increased water diffusion and potentially disrupted white matter microstructure.

Due to those finding, these MRI metrics has been discussed multiple times in the literature as potential biomarkers for psychotic disorders. However, the results of the studies using machine learning to predict psychotic disorder based on different MRI metrics reveal a great variability in the accuracies (Arbabshirani et al., 2017) and ML has been described as a “black box”, indicating that is is not clear which of the MRI metrics are most informative for achieving accurate machine classification.

While the majority of studies have focused on functional MRI (fMRI),diffusion tensor imaging (DTI) measures have been underrepresented in the literature (Ardekani et al., 2017) in predicting psychotic disorder. This research project aims to gain more insights in the predicting psychotic disorder specifically using macrostructural MRI data (cortical thickness (CT)) and microstructural MRI data (mean diffusivity (MD) and fractional anisotropy (FA)) with respect to their accuracies. So the main research question that follows is: Which of the MRI metrics (CT, MD, FA) are most informative for achieving accurate machine classification of psychotic disorder?.

References

Arbabshirani, M. R., Plis, S. M., Sui, J., & Calhoun, V. D. (2017). Single subject prediction of brain disorders in neuroimaging: Promises and pitfalls. NeuroImage, 145, 137–165. https://doi.org/10.1016/j.neuroimage.2016.02.079

Arciniegas, D. B. (2015). Psychosis. Continuum, 21, 715–736. https://doi.org/10.1212/01.con.0000466662.89908.e7

Ardekani, B. A., Tabesh, A., Sevy, S., Robinson, D., Bilder, R. M., & Szeszko, P. R. (2011). Diffusion tensor imaging reliably differentiates patients with schizophrenia from healthy volunteers. Human Brain Mapping, 32(1), 1–9. https://doi.org/10.1002/hbm.20995

Asmal, L., Du Plessis, S. S., Vink, M., Chiliza, B., Kilian, S., & Emsley, R. (2018). Symptom attribution and frontal cortical thickness in first-episode schizophrenia. Early Intervention in Psychiatry, 12(4), 652–659. https://doi.org/10.1111/eip.12358

American Association (2013). Diagnostic and Statistical Manual of Mental Disorders. https://doi.org/10.1176/appi.books.9780890425596

Azimi, P., Mohammadi, H. R., Benzel, E. C., Shahzadi, S., Azhari, S., & Montazeri, A. (2015). Artificial neural networks in neurosurgery. Journal of Neurology, Neurosurgery, and Psychiatry, 86(3), 251–256. https://doi.org/10.1136/jnnp-2014-307807

Baker, J. N., Dillon, D. G., Patrick, L. M., Roffman, J. L., Brady, R. O., Pizzagalli, D. A., Öngür, D., & Holmes, A. J. (2019). Functional connectomics of affective and psychotic pathology. Proceedings of the National Academy of Sciences of the United States of America, 116(18), 9050–9059. https://doi.org/10.1073/pnas.1820780116

Bassett, D. S., Bullmore, E. T., Verchinski, B. A., Mattay, V. S., Weinberger, D. R., & Meyer-Lindenberg, A. (2008). Hierarchical Organization of Human Cortical Networks in Health and Schizophrenia. The Journal of Neuroscience, 28(37), 9239–9248. https://doi.org/10.1523/jneurosci.1929-08.2008

Correll, C. U., Hauser, M., Auther, A. M., & Cornblatt, B. A. (2010). Research in people with psychosis risk syndrome: a review of the current evidence and future directions. Journal of Child Psychology and Psychiatry, 51(4), 390–431. https://doi.org/10.1111/j.1469-7610.2010.02235.x

Del Casale, A., Espagnet, M. C. R., Napolitano, A., Lucignani, M., Bonanni, L., Kotzalidis, G. D., Buscajoni, A., Manelfi, L., Perrone, V., Gualtieri, I., Brugnoli, R., De Pisa, E., Girardi, P., Romano, A., Ferracuti, S., Bozzao, A., & Pompili, M. (2021). Cerebral cortical thickness and gyrification changes in first-episode psychoses and multi-episode schizophrenia. Archives Italiennes De Biologie, 1, 3–20. https://doi.org/10.12871/00039829202111

Dwyer, D. E., Cabral, C., Kambeitz-Ilankovic, L., Sanfelici, R., Kambeitz, J., Calhoun, V. D., Falkai, P., Pantelis, C., Meisenzahl, E. M., & Koutsouleris, N. (2018). Brain Subtyping Enhances The Neuroanatomical Discrimination of Schizophrenia. Schizophrenia Bulletin, 44(5), 1060–1069. https://doi.org/10.1093/schbul/sby008

Jung, W. H., Kim, J. M., Jang, J. H., Choi, J., Jung, M. H., Park, J., Han, J. W., Choi, C. K., Kang, D. H., Chung, C. K., & Kwon, J. S. (2011). Cortical Thickness Reduction in Individuals at Ultra-High-Risk for Psychosis. Schizophrenia Bulletin, 37(4), 839–849. https://doi.org/10.1093/schbul/sbp151

Kubicki, M., McCarley, R. W., Westin, C., Park, H., Maier, S. E., Kikinis, R., Jolesz, F. A., & Shenton, M. E. (2007). A review of diffusion tensor imaging studies in schizophrenia. Journal of Psychiatric Research, 41(1–2), 15–30. https://doi.org/10.1016/j.jpsychires.2005.05.005

Kuperberg, G. R., Broome, M. R., McGuire, P., David, A. S., Eddy, M. D., Ozawa, F., Goff, D. C., West, W. C., Williams, S., Van Der Kouwe, A., Salat, D. H., Dale, A. M., & Fischl, B. (2003). Regionally Localized Thinning of the Cerebral Cortex in Schizophrenia. Archives of General Psychiatry, 60(9), 878. https://doi.org/10.1001/archpsyc.60.9.878

Lei, D., Pinaya, W. H. L., Van Amelsvoort, T., Marcelis, M., Donohoe, G., Mothersill, D., Corvin, A., Gill, M., Vieira, S. M. G., Huang, X., Lui, S., Scarpazza, C., Young, J. A., Arango, C., Bullmore, E. T., Qiyong, G., McGuire, P., & Mechelli, A. (2020). Detecting schizophrenia at the level of the individual: relative diagnostic value of whole-brain images, connectome-wide functional connectivity and graph-based metrics. Psychological Medicine, 50(11), 1852–1861. https://doi.org/10.1017/s0033291719001934

Lynall, M., Bassett, D. S., Kerwin, R., McKenna, P. J., Kitzbichler, M. G., Müller, U., & Bullmore, E. T. (2010). Functional Connectivity and Brain Networks in Schizophrenia. The Journal of Neuroscience, 30(28), 9477–9487. https://doi.org/10.1523/jneurosci.0333-10.2010

Meyer-Lindenberg, A., Olsen, R. K., Kohn, P., Brown, T. M., Egan, M. F., Weinberger, D. R., & Berman, K. F. (2005). Regionally Specific Disturbance of Dorsolateral Prefrontal–Hippocampal Functional Connectivity in Schizophrenia. Archives of General Psychiatry, 62(4), 379. https://doi.org/10.1001/archpsyc.62.4.379

Noble, W. S. (2006). What is a support vector machine? Nature Biotechnology, 24(12), 1565–1567. https://doi.org/10.1038/nbt1206-1565

Rezaii, N., Walker, E. F., & Wolff, P. (2019). A machine learning approach to predicting psychosis using semantic density and latent content analysis. Npj Schizophrenia, 5(1). https://doi.org/10.1038/s41537-019-0077-9

Rimol, L. M., Hartberg, C. B., Nesvåg, R., Fennema-Notestine, C., Hagler, D. J., Pung, C. J., Jennings, R. G., Haukvik, U. K., Lange, E., Nakstad, P. H., Melle, I., Andreassen, O. A., Dale, A. M., & Agartz, I. (2010). Cortical Thickness and Subcortical Volumes in Schizophrenia and Bipolar Disorder. Biological Psychiatry, 68(1), 41–50. https://doi.org/10.1016/j.biopsych.2010.03.036

Stamate, D., Katrinecz, A., Stahl, D., Verhagen, S. J. W., Delespaul, P., Van Os, J., & Guloksuz, S. (2019). Identifying psychosis spectrum disorder from experience sampling data using machine learning approaches. Schizophrenia Research, 209, 156–163. https://doi.org/10.1016/j.schres.2019.04.028

Vieira, S. M. G., Gong, Q., Pinaya, W. H. L., Scarpazza, C., Tognin, S., Crespo-Facorro, B., Tordesillas-Gutiérrez, D., Ortiz-García, V., Setién-Suero, E., Scheepers, F. E., Van Haren, N. E., Marques, T. R., Murray, R. M., David, A. S., Dazzan, P., McGuire, P., & Mechelli, A. (2020). Using Machine Learning and Structural Neuroimaging to Detect First Episode Psychosis: Reconsidering the Evidence. Schizophrenia Bulletin, 46(1), 17–26. https://doi.org/10.1093/schbul/sby189