EGFR Assessment in Lung Cancer CT Images: Analysis of Local and Holistic Regions of Interest Using Deep Unsupervised Transfer Learning

Statistics have demonstrated that one of the main factors responsible for the high mortality rate related to lung cancer is the late diagnosis. Precision medicine practices have shown advances in the individualized treatment according to the genetic profile of each patient, providing better control on cancer response. Medical imaging offers valuable information with an extensive perspective of the cancer, opening opportunities to explore the imaging manifestations associated with the tumor genotype in a non-invasive way. This work aims to study the relevance of physiological features captured from Computed Tomography images, using three different 2D regions of interest to assess the Epidermal growth factor receptor (EGFR) mutation status: nodule, lung containing the main nodule, and both lungs. A Convolutional Autoencoder was developed for the reconstruction of the input image. Thereafter, the encoder block was used as a feature extractor, stacking a classifier on top to assess the EGFR mutation status. Results showed that extending the analysis beyond the local nodule allowed the capture of more relevant information, suggesting the presence of useful biomarkers using the lung with nodule region of interest, which allowed to obtain the best prediction ability. This comparative study represents an innovative approach for gene mutations status assessment, contributing to the discussion on the extent of pathological phenomena associated with cancer development, and its contribution to more accurate Artificial Intelligence-based solutions, and constituting, to the best of our knowledge, the first deep learning approach that explores a comprehensive analysis for the EGFR mutation status classification.

[1]  P. Lambin,et al.  Genomics of non-small cell lung cancer (NSCLC): Association between CT-based imaging features and EGFR and K-RAS mutations in 122 patients-An external validation. , 2019, European journal of radiology.

[2]  S. Kobayashi,et al.  Epidermal growth factor receptor (EGFR) mutations in lung cancer: preclinical and clinical data. , 2014, Brazilian journal of medical and biological research = Revista brasileira de pesquisas medicas e biologicas.

[3]  Xueyan Yu,et al.  EGFR TKI as first-line treatment for patients with advanced EGFR mutation-positive non-small-cell lung cancer , 2017, Oncotarget.

[4]  Richard C. Pais,et al.  The Lung Image Database Consortium (LIDC) and Image Database Resource Initiative (IDRI): a completed reference database of lung nodules on CT scans. , 2011, Medical physics.

[5]  S. Reuzé,et al.  The complexity of tumor shape, spiculatedness, correlates with tumor radiomic shape features , 2019, Scientific Reports.

[6]  Paulo Mazzoncini de Azevedo Marques,et al.  Radiomics-based features for pattern recognition of lung cancer histopathology and metastases , 2018, Comput. Methods Programs Biomed..

[7]  T. Mitsudomi.  Advances in target therapy for lung cancer. , 2010, Japanese journal of clinical oncology.

[8]  Michael S. Bernstein,et al.  ImageNet Large Scale Visual Recognition Challenge , 2014, International Journal of Computer Vision.

[9]  P. Tomasini,et al.  Targeting the KRAS Pathway in Non-Small Cell Lung Cancer. , 2016, The oncologist.

[10]  Ayman El-Baz,et al.  Computer-Aided Diagnosis Systems for Lung Cancer: Challenges and Methodologies , 2013, Int. J. Biomed. Imaging.

[11]  Mohammed A. Fadhel,et al.  Towards a Better Understanding of Transfer Learning for Medical Imaging: A Case Study , 2020, Applied Sciences.

[12]  Sheng Chen,et al.  Pulmonary nodule detection on chest radiographs using balanced convolutional neural network and classic candidate detection , 2020, Artif. Intell. Medicine.

[13]  José João Mendes,et al.  Pre-Training Autoencoder for Lung Nodule Malignancy Assessment Using CT Images , 2020, Applied Sciences.

[14]  Y. Liu,et al.  Radiomic Features Are Associated With EGFR Mutation Status in Lung Adenocarcinomas. , 2016, Clinical lung cancer.

[15]  Ke-qiang Xu,et al.  Radiomics Signature as a Predictive Factor for EGFR Mutations in Advanced Lung Adenocarcinoma , 2020, Frontiers in Oncology.

[16]  Bingbing Ni,et al.  Toward automatic prediction of EGFR mutation status in pulmonary adenocarcinoma with 3D deep learning , 2019, Cancer medicine.

[17]  Kanghan Oh,et al.  Author Correction: Classification and Visualization of Alzheimer’s Disease using Volumetric Convolutional Neural Network and Transfer Learning , 2020, Scientific Reports.

[18]  Khalid M. Hosny,et al.  Skin Lesions Classification Into Eight Classes for ISIC 2019 Using Deep Convolutional Neural Network and Transfer Learning , 2020, IEEE Access.

[19]  Xue Ying,et al.  An Overview of Overfitting and its Solutions , 2019, Journal of Physics: Conference Series.

[20]  Chuong D. Hoang,et al.  Predictive radiogenomics modeling of EGFR mutation status in lung cancer , 2017, Scientific Reports.

[21]  A. Jemal,et al.  Cancer statistics, 2017 , 2017, CA: a cancer journal for clinicians.

[22]  Lawrence H. Schwartz,et al.  Implementation Strategy of a CNN Model Affects the Performance of CT Assessment of EGFR Mutation Status in Lung Cancer Patients , 2019, IEEE Access.

[23]  X. Fu,et al.  Detection of epithelial growth factor receptor (EGFR) mutations on CT images of patients with lung adenocarcinoma using radiomics and/or multi-level residual convolutionary neural networks. , 2018, Journal of thoracic disease.

[24]  Zhehai Wang,et al.  EGFR mutations as a prognostic and predictive marker in non-small-cell lung cancer , 2014, Drug design, development and therapy.

[25]  H. P. Oliveira,et al.  Identifying relationships between imaging phenotypes and lung cancer-related mutation status: EGFR and KRAS , 2019, Scientific Reports.

[26]  Jie Wu,et al.  The emerging treatment landscape of targeted therapy in non-small-cell lung cancer , 2019, Signal Transduction and Targeted Therapy.