Noninvasive KRAS mutation estimation in colorectal cancer using a deep learning method based on CT imaging

Background The detection of Kirsten rat sarcoma viral oncogene homolog (KRAS) gene mutations in colorectal cancer (CRC) is key to the optimal design of individualized therapeutic strategies. The noninvasive prediction of the KRAS status in CRC is challenging. Deep learning (DL) in medical imaging has shown its high performance in diagnosis, classification, and prediction in recent years. In this paper, we investigated predictive performance by using a DL method with a residual neural network (ResNet) to estimate the KRAS mutation status in CRC patients based on pre-treatment contrast-enhanced CT imaging. Methods We have collected a dataset consisting of 157 patients with pathology-confirmed CRC who were divided into a training cohort (n = 117) and a testing cohort (n = 40). We developed an ResNet model that used portal venous phase CT images to estimate KRAS mutations in the axial, coronal, and sagittal directions of the training cohort and evaluated the model in the testing cohort. Several groups of expended region of interest (ROI) patches were generated for the ResNet model, to explore whether tissues around the tumor can contribute to cancer assessment. We also explored a radiomics model with the random forest classifier (RFC) to predict KRAS mutations and compared it with the DL model. Results The ResNet model in the axial direction achieved the higher area under the curve (AUC) value (0.90) in the testing cohort and peaked at 0.93 with an input of ’ROI and 20-pixel’ surrounding area. AUC of radiomics model in testing cohorts were 0.818. In comparison, the ResNet model showed better predictive ability. Conclusions Our experiments reveal that the computerized assessment of the pre-treatment CT images of CRC patients using a DL model has the potential to precisely predict KRAS mutations. This new model has the potential to assist in noninvasive KRAS mutation estimation.

[1]  A. Burgun,et al.  Author Correction: Deep Learning and Radiomics predict complete response after neo-adjuvant chemoradiation for locally advanced rectal cancer , 2018, Scientific Reports.

[2]  D. Dong,et al.  Non-invasive radiomics approach potentially predicts non-functioning pituitary adenomas subtypes before surgery , 2018, European Radiology.

[3]  Zaiyi Liu,et al.  Effects of contrast-enhancement, reconstruction slice thickness and convolution kernel on the diagnostic performance of radiomics signature in solitary pulmonary nodule , 2016, Scientific Reports.

[4]  Gene Kitamura,et al.  Ankle Fracture Detection Utilizing a Convolutional Neural Network Ensemble Implemented with a Small Sample, De Novo Training, and Multiview Incorporation , 2019, Journal of Digital Imaging.

[5]  Yoshua Bengio,et al.  Gradient-based learning applied to document recognition , 1998, Proc. IEEE.

[6]  Bradley J. Erickson,et al.  Residual Deep Convolutional Neural Network Predicts MGMT Methylation Status , 2017, Journal of Digital Imaging.

[7]  Peter Schmid,et al.  Abstract 104: miRNAs in the 14q32 cluster are involved in lapatinib resistance , 2017 .

[8]  Sabine Tejpar,et al.  Fluorouracil, leucovorin, and irinotecan plus cetuximab treatment and RAS mutations in colorectal cancer. , 2015, Journal of clinical oncology : official journal of the American Society of Clinical Oncology.

[9]  N. Paragios,et al.  A radiomics approach to assess tumour-infiltrating CD8 cells and response to anti-PD-1 or anti-PD-L1 immunotherapy: an imaging biomarker, retrospective multicohort study. , 2018, The Lancet. Oncology.

[10]  Xinzhong Zhu,et al.  Radiomic signature as a diagnostic factor for histologic subtype classification of non-small cell lung cancer , 2018, European Radiology.

[11]  Lubomir M. Hadjiiski,et al.  Bladder Cancer Treatment Response Assessment in CT using Radiomics with Deep-Learning , 2017, Scientific Reports.

[12]  Jirí Sedlár,et al.  Predicting Deletion of Chromosomal Arms 1p/19q in Low-Grade Gliomas from MR Images Using Machine Intelligence , 2017, Journal of Digital Imaging.

[13]  Andriy Fedorov,et al.  Computational Radiomics System to Decode the Radiographic Phenotype. , 2017, Cancer research.

[14]  Jian Sun,et al.  Deep Residual Learning for Image Recognition , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[15]  Lyle J. Palmer,et al.  Precision Radiology: Predicting longevity using feature engineering and deep learning methods in a radiomics framework , 2017, Scientific Reports.

[16]  Raymond Y Huang,et al.  Residual Convolutional Neural Network for the Determination of IDH Status in Low- and High-Grade Gliomas from MR Imaging , 2017, Clinical Cancer Research.

[17]  S. Horvath,et al.  Unsupervised Learning With Random Forest Predictors , 2006 .

[18]  Senén Barro,et al.  Do we need hundreds of classifiers to solve real world classification problems? , 2014, J. Mach. Learn. Res..

[19]  S. Chiang,et al.  Correlation Between PET/CT Parameters and KRAS Expression in Colorectal Cancer , 2014, Clinical nuclear medicine.

[20]  Brian B. Avants,et al.  N4ITK: Improved N3 Bias Correction , 2010, IEEE Transactions on Medical Imaging.

[21]  Olivier Gevaert,et al.  Non–Small Cell Lung Cancer Radiogenomics Map Identifies Relationships between Molecular and Imaging Phenotypes with Prognostic Implications , 2017, Radiology.

[22]  Yali Yang,et al.  Developed and validated a prognostic nomogram for recurrence-free survival after complete surgical resection of local primary gastrointestinal stromal tumors based on deep learning , 2018, EBioMedicine.

[23]  Sabine Tejpar,et al.  Prognostic role of KRAS and BRAF in stage II and III resected colon cancer: results of the translational study on the PETACC-3, EORTC 40993, SAKK 60-00 trial. , 2010, Journal of clinical oncology : official journal of the American Society of Clinical Oncology.

[24]  Kaori Togashi,et al.  Relationship between 18F-Fluorodeoxyglucose Accumulation and KRAS/BRAF Mutations in Colorectal Cancer , 2012, Clinical Cancer Research.

[25]  M. Hatt,et al.  (18)F-FDG PET/CT imaging in rectal cancer: relationship with the RAS mutational status. , 2016, The British journal of radiology.

[26]  Hayaru Shouno,et al.  Analysis of function of rectified linear unit used in deep learning , 2015, 2015 International Joint Conference on Neural Networks (IJCNN).

[27]  J. E. van Timmeren,et al.  Influence of gray level discretization on radiomic feature stability for different CT scanners, tube currents and slice thicknesses: a comprehensive phantom study , 2017, Acta oncologica.

[28]  C. Mathers,et al.  Cancer incidence and mortality worldwide: Sources, methods and major patterns in GLOBOCAN 2012 , 2015, International journal of cancer.

[29]  Marc Peeters,et al.  Analysis of KRAS/NRAS Mutations in a Phase III Study of Panitumumab with FOLFIRI Compared with FOLFIRI Alone as Second-line Treatment for Metastatic Colorectal Cancer , 2015, Clinical Cancer Research.

[30]  Di Dong,et al.  Can CT-based radiomics signature predict KRAS/NRAS/BRAF mutations in colorectal cancer? , 2018, European Radiology.

[31]  H. Blum,et al.  Transcriptome analysis of dominant-negative Brd4 mutants identifies Brd4-specific target genes of small molecule inhibitor JQ1 , 2017, Scientific Reports.

[32]  S. Thrun,et al.  Corrigendum: Dermatologist-level classification of skin cancer with deep neural networks , 2017, Nature.

[33]  Lubomir M. Hadjiiski,et al.  Urinary bladder segmentation in CT urography using deep-learning convolutional neural network and level sets. , 2016, Medical physics.

[34]  Ying Cheng,et al.  Comparison of the effect of mean-based method and z-score for field normalization of citations at the level of Web of Science subject categories , 2014, Scientometrics.