Machine learning prediction of hematoma growth in acute intracerebral hemorrhage

2021 Fiscal Year Final Research Report

• Project/Area Number: 20K17947
• Research Category: Grant-in-Aid for Early-Career Scientists
• Allocation Type: Multi-year Fund
• Review Section: Basic Section 56010:Neurosurgery-related
• Research Institution: Department of Clinical Research, Nationai Hospital Organization Mie Chuo Medical Center
• Principal Investigator: Tanioka Satoru 独立行政法人国立病院機構三重中央医療センター(臨床研究部), その他部局等, 脳神経外科医師 (80838003)
• Project Period (FY): 2020-04-01 – 2022-03-31
• Keywords: 機械学習 / 脳出血 / 増大 / 予測

Outline of Final Research Achievements

Hematoma expansion occasionally occurs in patients with acute intracerebral hemorrhage (ICH), associating with poor outcome. Machine learning (ML) approaches perform well in outcome prediction. Patients with acute ICH from three hospitals (n=351) and those from another hospital (n=71) were retrospectively assigned to the development and validation cohorts, respectively. Machine learning (ML) models were evaluated for their performance on the patient data in the validation cohort, which was compared with previous scoring methods, the BAT, BRAIN, and 9-point scores.
The k-NN algorithm achieved the highest area under the receiver operating characteristic curve (AUC) of 0.790 among all ML models, and the sensitivity, specificity, and accuracy were 0.846, 0.733, and 0.775, respectively. The BRAIN score achieved the highest AUC of 0.676 among all previous scoring methods, which was lower than the best ML model by k-NN algorithm (p=0.016).

Free Research Field

脳卒中

Academic Significance and Societal Importance of the Research Achievements

機械学習は人工知能の根幹となる技術で，近年様々な分野で応用されているが，データの分類やデータから導き出される結果の予測に秀でている．本研究では，脳出血患者の入院時の年齢や性別，採血データ，既往歴等の臨床情報と，血腫量や血腫吸収値の特徴等のCT所見を，機械学習を用いて解析し，血腫増大の予測モデルを作成した．
使用したアルゴリズム，データについては，第三者による使用や検証が可能な状態とすることが重要である．アルゴリズム，匿名化したデータはインターネット上にアップロードした．