2019 Fiscal Year Final Research Report
Super hybrid analysis for the grade of breast cancer that combined diffusion MRI, contrast-enhanced MRI, and deep learning
| Project/Area Number |
17K10394
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Research Field |
Radiation science
|
|---|
| Research Institution | Kanazawa University |
|---|
Principal Investigator |
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
宮地 利明 金沢大学, 保健学系, 教授 (80324086)
|
|---|
| Project Period (FY) |
2017-04-01 – 2020-03-31
|
| Keywords | 拡散MRI / 造影MRI / ディープラーニング / 乳癌 / 悪性度 |
|---|
| Outline of Final Research Achievements |
This study was intended to stratify grade of breast cancer using a large number of information to be obtained from diffusion and contrast-enhanced MRI. For triple negative breast cancer and HER2 breast cancer, ADC was significantly low in pathological high grade group. About three kinds of diffusion coefficients of the triexponential analysis, the significant difference was not found between high and low grade group. For luminal breast cancer, ADC and D of the triexponential analysis were significantly low in a high-grade group. Also, SER(signal enhancement ratio) from contrast-enhanced MRI was significantly high in a high-grade group. ADC and SER were the factors which were useful in the extraction of the high-grade group by the multivariate analysis in luminal breast cancer.
|
| Free Research Field |
放射線科学
|
| Academic Significance and Societal Importance of the Research Achievements |
乳癌はそのサブタイプによって大筋の治療方針が決定されるが，１つのサブタイプの中でも生物学的悪性度には幅がある．手術後の病理検体の詳細な解析によって初めて癌の悪性度が判定されるという現状がある．術前化学療法が多くの患者に行われるようになってきている今，手術前に癌の悪性度を把握し，迅速な治療に結びつけられる手法の開発が待たれている．MRIは乳癌患者の術前検査として広く定着しており，多数の情報を得ることができる．我々は本邦でもほとんど行われていない拡散MRIのtriexponential解析の情報も含め，拡散MRIおよび造影MRIの多数の因子の中から，乳癌悪性度に強く関連する項目を抽出した．
|
