Computerized cognitive remediation therapy (CCRT) is generally effective for the cognitive deficits of schizophrenia. However, there is much uncertainty about what factors mediate or moderate effectiveness and are therefore important to personalize treatment and boost its effects.

In total, 311 Chinese inpatients with Diagnostic and Statistical Manual of Mental Disorders-IV schizophrenia were randomized to receive CCRT or Active control for 12 weeks with four to five sessions per week. All participants were assessed at baseline, post-treatment and 3-month follow-up. The outcomes were cognition, clinical symptoms and functional outcomes.

There was a significant benefit in the MATRICS Consensus Cognitive Battery (MCCB) total score for CCRT (F1,258 = 5.62; p = 0.02; effect size was 0.27, 95% confidence interval 0.04–0.49). There were no specific moderators of CCRT improvements. However, across both groups, Wisconsin Card Sort Test improvement mediated a positive effect on functional capacity and Digit Span benefit mediated decreases in positive symptoms. In exploratory analyses younger and older participants showed cognitive improvements but on different tests (younger on Symbol Coding Test, while older on the Spatial Span Test). Only the older age group showed MSCEIT benefits at post-treatment. In addition, cognition at baseline negatively correlated with cognitive improvement and those whose MCCB baseline total score was around 31 seem to derive the most benefit.

CCRT can improve the cognitive function of patients with schizophrenia. Changes in cognitive outcomes also contributed to improvements in functional outcomes either directly or solely in the context of CCRT. Age and the basic cognitive level of the participants seem to affect the cognitive benefits from CCRT.

Artificial Neural Network (ANN), as a potential powerful classifier, was explored to assist psychiatric diagnosis of the Composite International Diagnostic Interview (CIDI).

Both Back-Propagation (BP) and Kohonen networks were developed to fit psychiatric diagnosis and programmed (using 60 cases) to classify neurosis, schizophrenia and normal people. The programmed networks were cross-tested using another 222 cases. All subjects were randomly selected from two mental hospitals in Beijing.

Compared to ICD-10 diagnosis by psychiatrists, the overall kappa of BP network was 0.94 and that of Kohonen was 0.88 (both P < 0.01). In classifying patients who were difficult to diagnose, the kappa of BP was 0.69 (P < 0.01). ANN-assisted CIDI was compared with expert system assisted CIDI (kappa=0.72–0.76); ANN was more powerful than a traditional expert system.

ANN might be used to improve psychiatric diagnosis.