111
-
- Iihara Koji
- Principal Investigator
- 国立研究開発法人国立循環器病研究センター
-
- 東 尚弘
- Co-Investigator
- 国立研究開発法人国立がん研究センター
-
- 森脇 健介
- Co-Investigator
- 立命館大学
-
- 野原 康伸
- Co-Investigator
- 熊本大学
-
- 坂本 哲也
- Co-Investigator
- 帝京大学
-
- 中島 直樹
- Co-Investigator
- 九州大学
-
- 西村 邦宏
- Co-Investigator
- 国立研究開発法人国立循環器病研究センター
-
- 嘉田 晃子
- Co-Investigator
- 独立行政法人国立病院機構(名古屋医療センター臨床研究センター)
-
- 康 東天
- Co-Investigator
- 九州大学
-
- 萩原 明人
- Co-Investigator
- 九州大学
About this project
- Japan Grant Number
- JP18H02914
- Funding Program
- Grants-in-Aid for Scientific Research
- Funding organization
- Japan Society for the Promotion of Science
- Project/Area Number
- 18H02914
- Research Category
- Grant-in-Aid for Scientific Research (B)
- Allocation Type
-
- Single-year Grants
- Review Section / Research Field
-
- Basic Section 56010:Neurosurgery-related
- Research Institution
-
- National Cardiovascular Center Research Institute
- Kyushu University
- Project Period (FY)
- 2018-04-01 〜 2022-03-31
- Project Status
- Completed
- Budget Amount*help
- 17,420,000 Yen (Direct Cost: 13,400,000 Yen Indirect Cost: 4,020,000 Yen)
Research Abstract
The goals of this study were to develop a new prediction model to predict long-term poor prognosis (recurrent stroke within 1/3/5 years) in patients with acute ischemic stroke and to compare the performance of the new prediction model with other classic prediction scales. Machine learning algorithms were applied to train predictive models for estimating the individual risk of recurrent stroke with medical information 105 variables contained in DPC data for 597,134 patients enrolled in the J-ASPECT Study in the 2010-19 period. The prediction accuracy of the recurrence prediction model within 1/3/5 years was 0.62/0.62/0.63, which exceeded the prediction accuracy by classical risk scores. The prediction accuracy was sustained using 16 items including age, gender, medical history, smoking history, rehabilitation and appropriate secondary prevention at discharge, hospital stay, ADL at discharge and discharge location (0.61/0.62/0.62).
Details 詳細情報について
-
- CRID
- 1040282256970207616
-
- Text Lang
- ja
-
- Data Source
-
- KAKEN