Program Information
The Meaningful Radiomics Features for Magnetic Resonance Images
P Hu, J Wang , y sun , J Gan , R Luo , W Hu , Z Zhang* , Fudan University Shanghai Cancer Center, Shanghai, shanghai
Presentations
WE-RAM1-GePD-JT-3 (Wednesday, August 2, 2017) 9:30 AM - 10:00 AM Room: Joint Imaging-Therapy ePoster Theater
Purpose: To evaluate the meaningful radiomics features by analyzing double-scanned magnetic resonance (MR) images before radiotherapy and MR images before and after radiotherapy in rectal cancer.
Methods: 11 rectal cancer patients with stage Ⅱ were enrolled in this study, each of whom underwent two MR scans within average 18.6 days before radiotherapy. 256 radiomics features were defined in our study by using QIAT (Quantitative Image Analysis Toolbox, in-house developed radiomics feature extraction software), to calculate radiomics features for analysis. These features were divided into eight groups: (1) gray level co-occurrence matrix (GLCM), (2) gray level run-length matrix (GLRLM), (3)Histogram, (4)wavelet GLCM, (5)wavelet GLRLM, (6)wavelet histogram, (7)sharp and (8)fractal dimension. CCC of two MR scans was calculated to assess the reproducibility. 9 rectal cancer patients with stage Ⅱ enrolled in this study, MR scans were obtained before and after radiotherapy, and used to determine the concordance correlation coefficient (CCC) for 256 different radiomics features.
Results: 23 radiomics features showed high reproducibility (CCC≥0.8) for double-scanned MR images before radiotherapy, 83 radiomics features showed low reproducibility (CCC＜0.4) for MR images before and after radiotherapy. 8 radiomics features showed high reproducibility (CCC≥0.8) for double-scanned MR images before radiotherapy and showed low reproducibility (CCC＜0.4) for MR images before and after radiotherapy.
Conclusion: 8 radiomics features showed good reproducibility for double-scanned MR images before radiotherapy and showed low reproducibility for MR images before and after radiotherapy in rectal cancer. These 8 radiomics are maybe the meaningful features for radiomics study in rectal cancer.
Contact Email: