Diagnosing schizophrenia is challenging due to the variety of ways patients with it behave and react to treatments. The article outlines research to improve the situation by finding biomarkers that can be used to predict onset of the condition.The objective of the research is to find MRI-based imaging biomarkers for early diagnosis and prediction of treatment response in schizophrenia. Results show that microstructural properties as measured by diffusion MRI can be exploited to distinguish patients with schizophrenia from healthy people, as well as patients with good responses to medication from those without. Specifically, a machine learning approach is used to develop algorithms for individualised prediction of the disease and its prognosis. Currently the accuracy is around 80 per cent. Plans are in place to validate our method by testing the diagnostic performance in people who are in the subclinical phase of the disease, or in drug-naďve patients who are just starting to take antipsychotic medication. If these prospective studies yield positive results, the method can be considered viable to provide imaging biomarkers with clinical values.What is really needed to improve both diagnosis of schizophrenia itself, and assignment of the best treatment, is a more quantitative way of looking at the condition. As it stands there are no biomarkers available that physicians can use when they are trying to establish whether or not someone is showing symptoms of schizophrenia. However, this is something that a pioneering team based in Taiwan is currently trying to change. Professor Wen-Yih Isaac Tseng leads the Advanced Biomedical MRI Lab at the National Taiwan University College of Medicine (NTUCM), and he is leveraging his extensive professional experience using advanced imaging techniques to tackle the problem of individualised schizophrenia diagnosis. Recent work undertaken in the lab demonstrates that it is potentially feasible to discover imaging biomarkers for early diagnosis and prediction of treatment response in schizophrenia.
This work is licensed under a Creative Commons Attribution 4.0 Unported License. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/