臨床研究等提出・公開システム

Development of a diagnostic medical device program for mental disorders

Development of a diagnostic medical device program for mental disorders

79

There were 79 subjects enrolled, 76 in the safety analysis population and 70 in the efficacy analysis population (FAS and PPS). The mean (standard deviation) age distribution of the FAS group, by diagnosis, was 42.4 (10.9) years for the healthy group and 35.8 (12.2) years for the schizophrenia group . The proportion of women was 52.2% in the healthy group and 55.2% in the schizophrenic group.

The recruitment of the participants were started in October 2021 and discontinued in January 2023, due to delays in the development of a cognitive function test to be mounted on the tablet.Of the 9 participating medical facilities, 1 medical facility registered participants.

No adverse events and device failure occurred.

In the FAS population, eye movement tests with program version 1.0 or higher were performed in 90 cases(70 participants), cognitive function tests were performed 74 cases (63 participants), and both tests were performed in 74 cases (63subjects). Of these, data from 62cases (62 participants) with a diagnosis of healthy or schizophrenia (for those who were tested more than once, the latest dated information) was used for the analysis of the primary endpoints. No feature and model structure selection was based on machine learning and deep learning. The area under the curve of the ROC curve (ROC-AUC), the primary endpoint, was calculated by analysis using a logistic model that uses the eye movement test (Free Viewing), the word choice task evaluation score of the cognitive function test, and the symbol choice task evaluation score. A five-fold cross-validation method for evaluation of ROC-AUC was performed as the primary analysis, and ROC-AUC mean was 0.861, variance 0.009, and min-max values were 0.714 - 0.976. For reference, the ROC-AUC for all training data (N=62) was calculated 0.873 (95% confidence interval: 0.781-0.965).

Although the sample size is small and the performance evaluation is limited to internal validation, the linear sum of the three indicators evaluated in this study indicates that developed program may have predictive performance regarding the diagnosis of mental disorders.

Jan. 12, 2024

No

Not applicable

https://jrct.mhlw.go.jp/latest-detail/jRCTs032210197

Hashimoto Ryota

National Center of Neurology and Psychiatry

4-1-1 Ogawa-Higashi, Kodaira, Tokyo 187-8551, Japan

ryotahashimoto55@ncnp.go.jp

Ishizuka Takami

National Center of Neurology and Psychiatry

4-1-1 Ogawa-Higashi, Kodaira, Tokyo 187-8551, Japan

+81-42-341-2711

ishizukatakami@ncnp.go.jp

Complete

July. 14, 2021

Interventional

non-randomized controlled trial

open(masking not used)

no treatment control/standard of care control

factorial assignment

diagnostic purpose

Patients diagnosed as healthy, schizophrenia and other mental disorders

Patients deemed inappropriate by Principal investigator

Both

Schizophrenia

Use a diagnostic medical device for healthy, schizophrenia and other mental disorders patients

Schizophrenia

1. Area under ROC Curve in discriminating between healthy volunteers and schizophrenias

1. Sensitivity in discriminating patients with schizophrenia
2. Specificity at healthy volunteers
3. Specificity at other mental disorders patients
4. Percentage of correct classifications in discriminating between healthy volunteers and schizophrenias
5. Percentage of correct classifications in discriminating between other mental disorders and schizophrenias
6. Positive predictive value in discriminating between healthy volunteers and schizophrenias
7. Positive predictive value in discriminating between other mental disorders and schizophrenias
8. Correlation between the results of conventional eye movement tests and the results of eye movement tests using clinical trial equipment
9. Correlation between the results of conventional cognitive function tests and the results of cognitive function tests using clinical trial equipment
10. Correlation between clinical evaluation and eye movement test using clinical trial equipment
11. Correlation between clinical evaluation and cognitive function test using clinical trial equipment

+81-42-341-2712-7828

July. 02, 2021

none