Color paintings useful in diagnosing and prognosticating schizophrenia: BMC study
Figure 1. left panel: painting by healthy control, right panel: by schizophrenia patient
Individuals with psychiatric disorders perceive the world differently. A recent study published in BMC psychiatry by Hui Shen et al., has shown that a color painting-based paradigm can effectively predict clinical symptom severity for chronic schizophrenia patients. These paintings by schizophrenia patients show potential as a tool for clinical diagnosis and prognosis.
It has been observed that color vision is impaired in patients taking antipsychotic medication, potentially due to altered dopaminergic transmission. On the other hand, psychotic patients exhibited aberrant visual aftereffects, differential color priming effect, and weakened color discrimination ability.
Previous studies indicated impaired color vision and weakened color discrimination ability in psychotic patients. Examining the paintings from psychotic patients can measure the visual-motor function.
The present study aimed to explore the differences in color and stroke information between patients from a group of chronic schizophrenia patients and healthy controls (HCs) without psychiatric disorders and predict PANSS scores of schizophrenia patients according to their paintings.
The authors retrospectively analyzed the paintings colored by 281 chronic schizophrenia patients and 35 HCs. The images were scanned and processed using series of computational analyses.
Deep learning, also known as deep structured learning, (a technique in which higher-dimensional features can be learned from raw input features) was used to differentiate between paintings of HCs and Schizophrenia patients. The results showed that schizophrenia patients tend to use less color and exhibit different strokes compared to HCs. (Figure 1)
Using a deep learning residual neural network (ResNet), authors were able to discriminate patients from HCs with over 90% accuracy. Further, they developed a novel convolutional neural network to predict PANSS positive, negative, general psychopathology, and total scores. The Root Mean Square Error (RMSE) of the prediction was low, which indicates higher accuracy of prediction.
To summarize, it was found that the deep learning paradigm based on color painting dataset allowed sensitive prediction of clinical symptom severity for chronic schizophrenia patients.
The color paintings by schizophrenia patients show potential as a tool for clinical diagnosis and prognosis. This approach, therefore, provides novel options in proper evaluation of the rehabilitation stage for patients and has the potential to be adopted in home environment, especially for circumstances when systemic clinical interview is not available.
Source: BMC Psychiatry: https://doi.org/10.1186/s12888-021-03452-3
