Personalized Colorectal Cancer Survivability Prediction with Machine Learning Methods*

Abstract INTRODUCTION New non-invasive technologies for improving early diagnosis of colorectal cancer (CRC) are demanded by clinicians. Optical Coherence Tomography (OCT) provides sub-surface structural information and offers diagnosis capabilities of colon polyps, further improved by machine learning methods. Databases of OCT images are necessary to facilitate algorithms development and testing. MATERIALS AND METHODS A database has been acquired from rat colonic samples with a Thorlabs OCT system with 930nm centre wavelength that provides 1.2KHz A-scan rate, 7μm axial resolution in air, 4μm lateral resolution, 1.7mm imaging depth in air, 6mm x 6mm FOV, and 107dB sensitivity. The colon from anaesthetised animals has been excised and samples have been extracted and preserved for ex-vivo analysis with the OCT equipment. RESULTS This database consists of OCT 3D volumes (C-scans) and 2D images (B-scans) of murine samples from: 1) healthy tissue, for ground-truth comparison (18 samples; 66 C-scans; 17,478 B-scans); 2) hyperplastic polyps, obtained from an induced colorectal hyperplastic murine model (47 samples; 153 C-scans; 42,450 B-scans); 3) neoplastic polyps (adenomatous and adenocarcinomatous), obtained from clinically validated Pirc F344/NTac-Apcam1137 rat model (232 samples; 564 C-scans; 158,557 B-scans); and 4) unknown tissue (polyp adjacent, presumably healthy) (98 samples; 157 C-scans; 42,070 B-scans). CONCLUSIONS A novel extensive ex-vivo OCT database of murine CRC model has been obtained and will be openly published for the research community. It can be used for classification/segmentation machine learning methods, for correlation between OCT features and histopathological structures, and for developing new non-invasive in-situ methods of diagnosis of colorectal cancer.

Download Full-text

Machine learning for CT texture analysis in chemosensitivity of colorectal liver metastases: Initial results.

Journal of Clinical Oncology ◽

10.1200/jco.2019.37.15_suppl.e15148 ◽

2019 ◽

Vol 37 (15_suppl) ◽

pp. e15148-e15148

Author(s):

Yuji Miyamoto ◽

Takeshi Nakaura ◽

Yukiharu Hiyoshi ◽

Hideo Baba

Keyword(s):

Colorectal Cancer ◽

Machine Learning ◽

Prediction Model ◽

Liver Metastases ◽

Texture Analysis ◽

Gray Level ◽

Learning Methods ◽

Machine Learning Methods ◽

Chemotherapeutic Response ◽

Ct Texture Analysis

e15148 Background: Despite advances in cancer treatment over the last decades, more efficacious biomarkers are needed in patients with metastatic colorectal cancer. Several studies have reported that CT texture analysis is a useful prognostic biomarker for patients with colorectal cancer liver metastases (CRLM), however, little study has been done to explore those efficacies using machine learning methods. The present study aimed to evaluate the clinical efficacy of CT texture analysis using machine learning methods as a predictive marker of systemic chemotherapy in patients with CRLM. Methods: Sixty-four patients with CRLM who received first-line chemotherapy were included. Texture analysis was performed on 92 features (First Order Statistics, Gray Level Cooccurrence Matrix, Gray Level Run Length Matrix, Gray Level Size Zone Matrix, Neighbouring Gray Tone Difference Matrix and Gray Level Dependence Matrix) using CT within 1 month before treatment. We evaluated the association between those features and chemotherapeutic response by RECIST (CR+PR vs. SD+PD+NE). We performed eXtreme gradient boost (XGBoost) as a machine learning method to predict the chemotherapeutic response and used the receiver operating characteristic curves to evaluate this prediction model. Results: Main characteristics were the following: male/female = 36/28; median age = 63.5. Patients were treated with oxaliplatin-based chemotherapy (80% of patients), bevacizumab (77%) and anti-EGFR antibody (23%). Thirty-nine patients had confirmed responders, for an overall response rate of 61%, whereas 25 patients (39%) were classified as non-responders (CR: PR: SD: PD: NE = 0: 39: 20: 4: 1). The area under curve of this prediction model was 0.771. Conclusions: We confirmed that CT texture analysis using machine learning for CRLM was feasible. Further analyses are ongoing.

Download Full-text