Data collaboration analysis in predicting diabetes from a small amount of health checkup data.

Recent studies showed that machine learning models such as gradient-boosting decision tree (GBDT) can predict diabetes with high accuracy from big data. In this study, we asked whether highly accurate prediction of diabetes is possible even from small data by expanding the amount of data through data collaboration (DC) analysis, a modern framework for integrating and analyzing data accumulated at multiple institutions while ensuring confidentiality. To this end, we focused on data from two institutions: health checkup data of 1502 citizens accumulated in Tsukuba City and health history data of 1399 patients collected at the University of Tsukuba Hospital. When using only the health checkup data, the ROC-AUC and Recall for logistic regression (LR) were 0.858 ± 0.014 and 0.970 ± 0.019, respectively, while those for GBDT were 0.856 ± 0.014 and 0.983 ± 0.016, respectively. When using also the health history data through DC analysis, these values for LR improved to 0.875 ± 0.013 and 0.993 ± 0.009, respectively, while those for GBDT deteriorated because of the low compatibility with a method used for confidential data sharing (although DC analysis brought improvements). Even in a situation where health checkup data of only 324 citizens are available, the ROC-AUC and Recall for LR were 0.767 ± 0.025 and 0.867 ± 0.04, respectively, thanks to DC analysis, indicating an 11% and 12% improvement. Thus, we concluded that the answer to the above question was "Yes" for LR but "No" for GBDT for the data set tested in this study.

DC-COX: Data collaboration Cox proportional hazards model for privacy-preserving survival analysis on multiple parties.

The demand for the privacy-preserving survival analysis of medical data integrated from multiple institutions or countries has been increased. However, sharing the original medical data is difficult because of privacy concerns, and even if it could be achieved, we have to pay huge costs for cross-institutional or cross-border communications. To tackle these difficulties of privacy-preserving survival analysis on multiple parties, this study proposes a novel data collaboration Cox proportional hazards (DC-COX) model based on a data collaboration framework for horizontally and vertically partitioned data. By integrating dimensionality-reduced intermediate representations instead of the original data, DC-COX obtains a privacy-preserving survival analysis without iterative cross-institutional communications or huge computational costs. DC-COX enables each local party to obtain an approximation of the maximum likelihood model parameter, the corresponding statistic, such as the p-value, and survival curves for subgroups. Based on a bootstrap technique, we introduce a dimensionality reduction method to improve the efficiency of DC-COX. Numerical experiments demonstrate that DC-COX can compute a model parameter and the corresponding statistics with higher performance than the local party analysis. Particularly, DC-COX demonstrates outstanding performance in essential feature selection based on the p-value compared with the existing methods including the federated learning-based method.

Sequential reinforcement active feature learning for gene signature identification in renal cell carcinoma.

Renal cell carcinoma (RCC) is one of the deadliest cancers and mainly consists of three subtypes: kidney clear cell carcinoma (KIRC), kidney papillary cell carcinoma (KIRP), and kidney chromophobe (KICH). Gene signature identification plays an important role in the precise classification of RCC subtypes and personalized treatment. However, most of the existing gene selection methods focus on statically selecting the same informative genes for each subtype, and fail to consider the heterogeneity of patients which causes pattern differences in each subtype. In this work, to explore different informative gene subsets for each subtype, we propose a novel gene selection method, named sequential reinforcement active feature learning (SRAFL), which dynamically acquire the different genes in each sample to identify the different gene signatures for each subtype. The proposed SRAFL method combines the cancer subtype classifier with the reinforcement learning (RL) agent, which sequentially select the active genes in each sample from three mixed RCC subtypes in a cost-sensitive manner. Moreover, the module-based gene filtering is run before gene selection to filter the redundant genes. We mainly evaluate the proposed SRAFL method based on mRNA and long non-coding RNA (lncRNA) expression profiles of RCC datasets from The Cancer Genome Atlas (TCGA). The experimental results demonstrate that the proposed method can automatically identify different gene signatures for different subtypes to accurately classify RCC subtypes. More importantly, we here for the first time show the proposed SRAFL method can consider the heterogeneity of samples to select different gene signatures for different RCC subtypes, which shows more potential for the precision-based RCC care in the future.

Schizophrenia Quality of Life Scale: validation of the Japanese version.

This study was performed to test the validity and value of the Schizophrenia Quality of Life Scale as an assessment tool in a Japanese-language version (JSQLS). The subjects for the present study were 55 inpatients with a diagnosis of schizophrenia as defined by DSM-IV. The JSQLS was administered together with two other self-report measures, the Medical Outcomes Study 36-item short-form health survey questionnaire (SF-36) and the WHO QOL-26, to assess validity. Psychotic symptoms and extrapyramidal symptoms were assessed using the BPRS and the Drug-Induced Extrapyramidal Symptoms Scale (DIEPSS), respectively. All the scales (psychosocial, motivation/energy and symptoms/side effects) showed good internal consistency reliability (Cronbach's alpha=0.93, 0.73 and 0.80, respectively). The correlations of items with their scale total revealed that almost all items were significantly correlated with their own scale score. There were associations with relevant SF-36, WHO QOL-26, and DIEPSS scores. From the results of the testing of the reliability and validity of the JSQLS, it is concluded that the JSQLS is a simple and reliable scale.