Journal Article

Deep learning for lung cancer prognostication: A retrospective multi-cohort radiomics study

Coroller, Thibaud P., Aerts, Hugo J. W. L., Grossmann, Patrick, Kumar, Avnish, Mak, Raymond H., Parmar, Chintan, Bussink, Johan, Gillies, Robert J., Hosny, Ahmed, Zeleznik, Roman

2018

Overview

Non-small-cell lung cancer (NSCLC) patients often demonstrate varying clinical courses and outcomes, even within the same tumor stage. This study explores deep learning applications in medical imaging allowing for the automated quantification of radiographic characteristics and potentially improving patient stratification. We performed an integrative analysis on 7 independent datasets across 5 institutions totaling 1,194 NSCLC patients (age median = 68.3 years [range 32.5-93.3], survival median = 1.7 years [range 0.0-11.7]). Using external validation in computed tomography (CT) data, we identified prognostic signatures using a 3D convolutional neural network (CNN) for patients treated with radiotherapy (n = 771, age median = 68.0 years [range 32.5-93.3], survival median = 1.3 years [range 0.0-11.7]). We then employed a transfer learning approach to achieve the same for surgery patients (n = 391, age median = 69.1 years [range 37.2-88.0], survival median = 3.1 years [range 0.0-8.8]). We found that the CNN predictions were significantly associated with 2-year overall survival from the start of respective treatment for radiotherapy (area under the receiver operating characteristic curve [AUC] = 0.70 [95% CI 0.63-0.78], p < 0.001) and surgery (AUC = 0.71 [95% CI 0.60-0.82], p < 0.001) patients. The CNN was also able to significantly stratify patients into low and high mortality risk groups in both the radiotherapy (p < 0.001) and surgery (p = 0.03) datasets. Additionally, the CNN was found to significantly outperform random forest models built on clinical parameters-including age, sex, and tumor node metastasis stage-as well as demonstrate high robustness against test-retest (intraclass correlation coefficient = 0.91) and inter-reader (Spearman's rank-order correlation = 0.88) variations. To gain a better understanding of the characteristics captured by the CNN, we identified regions with the most contribution towards predictions and highlighted the importance of tumor-surrounding tissue in patient stratification. We also present preliminary findings on the biological basis of the captured phenotypes as being linked to cell cycle and transcriptional processes. Limitations include the retrospective nature of this study as well as the opaque black box nature of deep learning networks. Our results provide evidence that deep learning networks may be used for mortality risk stratification based on standard-of-care CT images from NSCLC patients. This evidence motivates future research into better deciphering the clinical and biological basis of deep learning networks as well as validation in prospective data.

• Publisher: Public Library of Science,Public Library of Science (PLoS)
• Subject: Adult / Age / Aged / Aged, 80 and over / Algorithms / Artificial intelligence / Artificial neural networks / Biology and Life Sciences / Biomarkers / Black boxes / Cancer metastasis / Cancer patients / Cancer research / Cancer therapies / Carcinoma, Non-Small-Cell Lung - diagnostic imaging / Carcinoma, Non-Small-Cell Lung - mortality / Carcinoma, Non-Small-Cell Lung - pathology / Carcinoma, Non-Small-Cell Lung - therapy / CAT scans / Cell cycle / Chemotherapy / Clinical Decision-Making / Computed tomography / Computer and Information Sciences / Correlation coefficient / Correlation coefficients / Deep Learning / Diagnosis, Computer-Assisted - methods / Diagnostic imaging / Evidence-based medicine / Female / Gene expression / Genomics / Hospitals / Humans / Learning / Lung cancer / Lung Neoplasms - diagnostic imaging / Lung Neoplasms - mortality / Lung Neoplasms - pathology / Lung Neoplasms - therapy / Machine learning / Male / Medical imaging / Medical imaging equipment / Medical research / Medical schools / Medicine / Medicine and Health Sciences / Metastases / Methods / Middle Aged / Mortality / Neoplasm Staging / Neural networks / Non-small cell lung cancer / Non-small cell lung carcinoma / Oncology / Patients / People and Places / Phenotypes / Predictive Value of Tests / Preliminary Data / Prognosis / Radiation therapy / Radiographic Image Interpretation, Computer-Assisted - methods / Radiomics / Radiotherapy / Reproducibility of Results / Research and Analysis Methods / Retrospective Studies / Risk Assessment / Risk Factors / Risk groups / Science Policy / Semantics / Software / Stratification / Surgery / Survival / Tomography / Tomography, X-Ray Computed - methods / Transcription / Transfer learning / Tumors / Women