Background and Aims
Magnifying narrow-band imaging (M-NBI) is important in the diagnosis of early gastric cancers (EGCs) but requires expertise to master. We developed a computer-aided diagnosis (CADx) system to assist endoscopists in identifying and delineating EGCs.
We retrospectively collected and randomly selected 66 EGC M-NBI images and 60 non-cancer M-NBI images into a training set and 61 EGC M-NBI images and 20 non-cancer M-NBI images into a test set. After preprocessing and partition, we determined 8 gray-level co-occurrence matrix (GLCM) features for each partitioned 40 × 40 pixel block and calculated a coefficient of variation of 8 GLCM feature vectors. We then trained a support vector machine (SVMLv1) based on variation vectors from the training set and examined in the test set. Furthermore, we collected 2 determined P and Q GLCM feature vectors from cancerous image blocks containing irregular microvessels from the training set, and we trained another SVM (SVMLv2) to delineate cancerous blocks, which were compared with expert-delineated areas for area concordance.
The diagnostic performance revealed accuracy of 96.3%, precision (positive predictive value [PPV]) of 98.3%, recall (sensitivity) of 96.7%, and specificity of 95%, at a rate of 0.41 ± 0.01 seconds per image. The performance of area concordance, on a block basis, demonstrated accuracy of 73.8% ± 10.9%, precision (PPV) of 75.3% ± 20.9%, recall (sensitivity) of 65.5% ± 19.9%, and specificity of 80.8% ± 17.1%, at a rate of 0.49 ± 0.04 seconds per image.
This pilot study demonstrates that our CADx system has great potential in real-time diagnosis and delineation of EGCs in M-NBI images.
Background and Aims
Endoscopic real-time imaging of Barrett’s esophagus (BE) with advanced imaging technologies enables targeted biopsies and may eliminate the need for random biopsies to detect dysplasia during endoscopic surveillance of BE. This systematic review and meta-analysis was performed by the American Society for Gastrointestinal Endoscopy (ASGE) Technology Committee to specifically assess whether acceptable performance thresholds outlined by the ASGE Preservation and Incorporation of Valuable Endoscopic Innovations (PIVI) document for clinical adoption of these technologies have been met.
We conducted meta-analyses calculating the pooled sensitivity, negative predictive value (NPV), and specificity for chromoendoscopy by using acetic acid and methylene blue, electronic chromoendoscopy by using narrow-band imaging, and confocal laser endomicroscopy (CLE) for the detection of dysplasia. Random effects meta-analysis models were used. Statistical heterogeneity was evaluated by means of I2 statistics.
The pooled sensitivity, NPV, and specificity for acetic acid chromoendoscopy were 96.6% (95% confidence interval [CI], 95-98), 98.3% (95% CI, 94.8-99.4), and 84.6% (95% CI, 68.5-93.2), respectively. The pooled sensitivity, NPV, and specificity for electronic chromoendoscopy by using narrow-band imaging were 94.2% (95% CI, 82.6-98.2), 97.5% (95% CI, 95.1-98.7), and 94.4% (95% CI, 80.5-98.6), respectively. The pooled sensitivity, NPV, and specificity for endoscope-based CLE were 90.4% (95% CI, 71.9-97.2), 98.3% (95% CI, 94.2-99.5), and 92.7% (95% CI, 87-96), respectively.
Our meta-analysis indicates that targeted biopsies with acetic acid chromoendoscopy, electronic chromoendoscopy by using narrow-band imaging, and endoscope-based CLE meet the thresholds set by the ASGE PIVI, at least when performed by endoscopists with expertise in advanced imaging techniques. The ASGE Technology Committee therefore endorses using these advanced imaging modalities to guide targeted biopsies for the detection of dysplasia during surveillance of patients with previously nondysplastic BE, thereby replacing the currently used random biopsy protocols.