| Project/Area Number |
17H06629
|
|---|
| Research Category |
Grant-in-Aid for Research Activity Start-up
|
| Allocation Type | Single-year Grants |
|---|
| Research Field |
Epidemiology and preventive medicine
|
|---|
| Research Institution | The University of Tokyo |
|---|
Principal Investigator |
Ono Sachiko 東京大学, 大学院医学系研究科(医学部), 特任助教 (20797237)
|
|---|
| Project Period (FY) |
2017-08-25 – 2019-03-31
|
| Project Status |
Completed (Fiscal Year 2018)
|
| Budget Amount *help |
¥2,730,000 (Direct Cost: ¥2,100,000、Indirect Cost: ¥630,000)
Fiscal Year 2018: ¥1,300,000 (Direct Cost: ¥1,000,000、Indirect Cost: ¥300,000)
Fiscal Year 2017: ¥1,430,000 (Direct Cost: ¥1,100,000、Indirect Cost: ¥330,000)
|
|---|
| Keywords | 医療ビッグデータ / 予測 / 予測モデル / 機械学習 / 慢性疾患 / 疾患発症予測 / 疾患重症化予測モデル |
|---|
| Outline of Final Research Achievements |
In this study, we investigated the onset of hyperthyroidism and renal dysfunction in diabetic patients by using multiple machine learning methods and conventional epidemiologic methods. The latter study was presented at an international conference. In addition, in order to examine the usefulness of laboratory data in the construction of prediction models, we used data obtained from cooperating institutions to predict surgical site infection after abdominal surgery, resumption of antibiotics, and postoperative length of stay. Although the number of cases of surgical site infection and resumption of antibiotics were small and unpredictable, the difference between the predicted value and the observed value of postoperative length of stay using Random Forest was within 2 days in each deciles.
|
| Academic Significance and Societal Importance of the Research Achievements |
現状の研究利用可能な医療レセプトデータは、サンプル数、説明変数、追跡期間が限られており、機械学習の手法で重症化や疾患の発症予測は困難であった。一方、術後入院期間は一定の精度で予測することが確認されたため、既存の医療レセプトデータは機械学習を用いた短期的アウトカムの予測に利用できる可能性が示唆された。
|
