| Project/Area Number |
17K16466
|
|---|
| Research Category |
Grant-in-Aid for Young Scientists (B)
|
| Allocation Type | Multi-year Fund |
|---|
| Research Field |
Radiation science
|
|---|
| Research Institution | Sapporo Medical University |
|---|
Principal Investigator |
|
|---|
| Project Period (FY) |
2017-04-01 – 2019-03-31
|
| Project Status |
Completed (Fiscal Year 2018)
|
| Budget Amount *help |
¥4,030,000 (Direct Cost: ¥3,100,000、Indirect Cost: ¥930,000)
Fiscal Year 2018: ¥1,820,000 (Direct Cost: ¥1,400,000、Indirect Cost: ¥420,000)
Fiscal Year 2017: ¥2,210,000 (Direct Cost: ¥1,700,000、Indirect Cost: ¥510,000)
|
|---|
| Keywords | 放射線治療 / 前立腺癌 / 機械学習 / 人工ニューラルネットワーク / 放射線治療生物学 |
|---|
| Outline of Final Research Achievements |
We examined the application of an artificial neural network (ANN) model to predict the outcome of radiation therapy using immunohistochemical staining and clinical factors of Ku70 for prostate and hypopharyngeal cancer. Age, Gleason score, biopsy positive rate, pre-treatment PSA value, risk classification, prostate volume were used as clinical factors in analysis of prostate cancer. Similarly, in hypopharyngeal cancer, age, gender, performance status, clinical T staging and subsite were used as clinical factors.
The treatment result prediction by ANN was a result that the sensitivity, the specificity, and the area under curve (AUC) of the ROC curve based on the prediction result were all superior in prediction ability as compared with the conventional method.
|
| Academic Significance and Societal Importance of the Research Achievements |
従来の方法としては、log-rank法やcox比例ハザードモデルを用いた多変量解析により、危険度の高い因子の解析結果から予後予測が試みられている場合が多い。機械学習のアルゴリズムのひとつである人工ニューラルネットワーク（ANN）では、コンピュータ上に神経細胞組織を模した構造を作成し、擬似的に神経活動を行わせることによって、線形分離し難い情報処理を行うことが可能となる。この方法は、問題となる入力信号と、その答えとなる出力信号を与え、学習させることにより、多数ある患者背景や臨床因子の中から、必要な予測因子を適切に選択することでより精度の高い予測モデルの構築を目指す。
|
