Endoscopic Diagnostic Support System for cT1b Colorectal Cancer Using Deep Learning

Abstract

Objective: This study aimed to use convolutional neural network (CNN), a deep learning software, to assist in cT1b diagnosis. Methods: This retrospective study used 190 colon lesion images from 41 cases of colon endoscopies performed between February 2015 and October 2016. Unenhanced colon endoscopy images (520 × 520 pixels) with white light were used. Images included 14 cTis cases with endoscopic resection and 14 cT1a and 13 cT1b cases with surgical resection. Protruding, flat, and recessed lesions were analyzed. AlexNet and Caffe were used for machine learning. Fine tuning of data to increase image numbers was performed. Oversampling for the training images was conducted to avoid impartiality in image numbers, and learning was carried out. The 3-fold cross-validation method was used. Sensitivity, specificity, accuracy, and area under the curve (AUC) values in the receiver operating characteristic curve were calculated for each group. Results: The results were the average of obtained values. With CNN learning, cT1b sensitivity, specificity, and accuracy were 67.5, 89.0, and 81.2%, respectively, and AUC was 0.871. Conclusion: Quantitative diagnosis is possible using an endoscopic diagnostic support system with machine learning, without relying on the skill and experience of endoscopists. Moreover, this system could be used to objectively evaluate endoscopic diagnoses.

Diagnostic performance of artificial intelligence to identify deeply invasive colorectal cancer on non-magnified plain endoscopic images

“…A relatively lower sensitivity value will not influence the therapy for patients with deeply invasive CRC because pathological evaluation will be performed to confirm the invasiveness of the CRC after endoscopic resection. From this viewpoint, the CAD system developed in the present study achieved a better performance, compared with a previous study showing 68 % [17] and 78 % [18] specificity although these direct comparisons may be inappropriate because of using different testing datasets. Future CAD systems should aim to obtain a higher specificity comparable with the results of evaluation by experts.…”

“…The current study has several advantages compared with a previous study [16,17]. A relatively large number (1839) of endoscopic images of CRCs were collected from two hospitals for deep learning.…”

“…Endocytoscopy combined with a CAD system may also play an important role in this new field [16]. Recently, Ito et al proposed a CAD system to identify T1b CRC using plain endoscopic images, but the diagnostic performance did not have excellent results (sensitivity 89 %, specificity 68 %) [17]. Using narrow-band imaging (NBI), more recently, Lui T et al reported an AI image classifier of candidates for endoscopic resection and revealed better diagnostic performance (sensitivity 88.2 %.…”

Lesion-Based Convolutional Neural Network in Diagnosis of Early Gastric Cancer

“…22 Moreover, it is currently being applied for the detection and classification of GI tract cancer, including cancer of the esophagus, stomach, and colon. 5,6,23…”

Supervised Machine Learning in Oncology: A Clinician's Guide

“…Within oncology, most machine learning research has focused on using genetic sequencing, clinical imaging, and images of histopathological specimens to diagnose cancer as early as possible in many types of primary cancers. [5][6][7] Machine learning models have also been used to predict the survival or risk of recurrence for patients already diagnosed with cancer. [8][9][10] More recently, machine learning has been used to help clinicians recommend cancer treatment modalities when evidence-driven guidelines are lacking, inconsistent, or impractical.…”