Speech feature learning is the core and key of speech recognition method for mental illness. Deep feature learning can automatically extract speech features, but it is limited by the problem of small samples. Traditional feature extraction (original features) can avoid the impact of small samples, but it relies heavily on experience and is poorly adaptive. To solve this problem, this paper proposes a deep embedded hybrid feature sparse stack autoencoder manifold ensemble algorithm. Firstly, based on the prior knowledge, the psychotic speech features are extracted, and the original features are constructed. Secondly, the original features are embedded in the sparse stack autoencoder (deep network), and the output of the hidden layer is filtered to enhance the complementarity between the deep features and the original features. Third, the L1 regularization feature selection mechanism is designed to compress the dimensions of the mixed feature set composed of deep features and original features. Finally, a weighted local preserving projection algorithm and an ensemble learning mechanism are designed, and a manifold projection classifier ensemble model is constructed, which further improves the classification stability of feature fusion under small samples. In addition, this paper designs a medium-to-large-scale psychotic speech collection program for the first time, collects and constructs a large-scale Chinese psychotic speech database for the verification of psychotic speech recognition algorithms. The experimental results show that the main innovation of the algorithm is effective, and the classification accuracy is better than other representative algorithms, and the maximum improvement is 3.3%. In conclusion, this paper proposes a new method of psychotic speech recognition based on embedded mixed sparse stack autoencoder and manifold ensemble, which effectively improves the recognition rate of psychotic speech.
Objective To explore factors that affect the assessment of sexual self-defense capacity and to evaluate the effect of social functions on sexual self-defense capacity in patients with mental retardation. Methods A 1∶1 matched case-control study was carried out, which included 174 sexual assault cases, 87 with mental retardation and 87 without mental retardation. A record of forensic psychiatry assessment designed by ourselves was used to collect the general characteristics. Wechsler Adult Intelligence-Rerisedin China (WAIS-RC) was used to determine the intelligence quotient. Rating Scale of Intellectual Disability (RSID) and Global Assessment Function (GAF) were used to assess social functions. Results Besides factors such as culture, occupation, knowledge about sex and payment claim, the scores of GAF and RSID were also related to the assessment of sexual self-defense capacity. Moreover, the correct ratio of discrimination was 73.1% (yes), 66.9% (impaired) and 87.2% (no), respectively, and the general correct ratio of discrimination was 78.1%. Conclusion Not only social and legal factors but also the level of social functions should be considered in the assessment of sexual self-defense capacity. Therefore, it might be concluded that multiple discriminant analysis can be useful when assessing the sexual self-defense capacity of patients with mental retardation.
ObjectiveTo explore the influence factors of therapeutic compliance and emotional expression of first-degree relatives in acute schizophrenic patients with psychotic symptoms.
MethodsThe Brief Psychiatric Rating Scale (BPRS) was used to measure the severity of psychotic symptoms in sixty schizophrenic patients from June to September 2014 in West China Hospital and the Toronto Alexithymia Scale (TAS) was used to survey the emotional expression in their family members. The homemade treatment adherence scale was used to survey the treatment adherence in patients for one week.
ResultsThere was a poor therapeutic compliance in nineteen patients with acute schizophrenia (32%) and the other 41(68%) had good therapeutic compliance; the relatives of schizophrenic patients had high TAS scores (male: 67.61±10.03; female: 69.68±11.46) than the normal models did (P ＜ 0.05) . The differences between the patients with different therapeutic compliance in BPRS total score, reactivator, hostile and suspicion factor (P ＜ 0.05) . The therapeutic compliance was related to the severity of the psychotic symptoms (P ＜ 0.05) . Conclusions There is a bad emotional expression in the relatives of acute schizophrenic patients. The psychotic symptoms can influence the therapeutic compliance. The milder the psychotic symptoms, the better the therapeutic dependence.
Objective To compare the incidence of social function disorders in head injury patients with and without craniocerebral injury and to explore the value for mental identification. Methods SDSS (Social Disability Screening Schedule), GAF (Global Assessment Function) and GAS (Global Assessment Scale) instruments were used to test the social function of 56 patients without craniocerebral injury and 55 patients with craniocerebral injury. Results One hundred and eleven patients with head injury were included and identified as head injury with or without craniocerebral injury by CT or MRI. The incidence of social function deficit, tested by using SDSS instrument, was 33.9% (19/56) in patients without craniocerebral injury and 45.5% (25/55) in patients with craniocerebral injury respectively. There was no statistically significant difference between the two groups (χ2=1.544, P=0.214). This was also no statitical difference in both GAS group (t=0.021, P=0.983) and GAF group (t=0.391, P=0.697). Conclusions The limited evidence showed that the incidence of social function deficit of the head injury patients combined with craniocerebral injury is higher than those who without craniocerebral injury, but the difference between the two groups has no statistically significant difference.We could not detect a difference in the incidence of social deficit between those head injury patients with or without craniocerebral injury ones.
ObjectiveTo systematically evaluate the relationship between the-2548G/A polymorphism in the leptin gene and antipsychotic-induced weight gain (AIWG).
MethodsLiterature for the relationship between the-2548G/A polymorphism in the leptin gene and AIWG was retrieved in electronic databases including PubMed, EMbase, CNKI and WanFang Data from establishment dates to June, 2013. Two reviewers independently screened studies according to the inclusion and exclusion criteria, extracted data and evaluated the methodological quality of the included studies. Then meta-analysis was performed using RevMan 5.2 software.
ResultsA total of 7 case-control studies were included, involving 404 AIWG cases and 508 controls (patients with no significant changes of weight after taking antipsychotic drugs). The results of meta-analysis showed that, regarding the total population, the-2548G/A polymorphism of the leptin gene was not associated with AIWG (OR=1.16, 95%CI 0.70 to 1.93, P=0.57). After stratification analysis, according to Chinese or non-Chinese origin, the results showed that significant association was found between the-2548G/A polymorphism of leptin gene and AIWG for Chinese (OR=2.15, 95%CI 1.41 to 3.26, P=0.000 4) but not for non-Chinese (OR=0.69, 95%CI 0.45 to 1.07, P=0.10).
ConclusionThe current evidence suggests that the-2548G/A polymorphism in the leptin gene is associated with increased risk of AIWG for Chinese. Due to limited quantity of the included studies, the aforementioned conclusion needs to be further validate by more high-quality and large-scale studies.
ObjectiveTo explore the effects of drug management skill training on lightening the family burden of schizophrenic patients in their recovery period.
MethodsBetween December 2011 and December 2013, 101 patients with schizophrenia were randomly divided into experimental group (n=56) and control group (n=45). The experimental group was given drug management skill training, while the control group only received routine follow-up. The course of the research was six months. Both groups were assessed by the positive and negative syndrome scale on patients' psychological symptoms, and family burden scale of diseases was used to assess the burden of the family.
ResultsCompared with the controls, patients in the experimental group improved more in their positive symptoms (t=2.692, P=0.008), negative symptoms (t=2.729, P=0.008), general psychopathology symptoms (t=3.231, P=0.002) and the whole psychiatric symptoms (t=3.870, P<0.001). Moreover, the degree of patients' symptom improvement was positively correlated with the degree of family burden lightening (r=0.44, P<0.001).
ConclusionFor patients with schizophrenia, reasonable drug management skill training can effectively improve patients' medication compliance, promote treatment effect and lighten family burden.
