Comprehensive Diagnostic Performance of Real-Time Characterization of Colorectal Lesions Using an Artificial Intelligence–Assisted System: A Prospective Study

Minegishi, Yosuke; Kudo, Shin‐ei; Miyata, Yuki; Nemoto, Tetsuo; Mori, Kensaku; Misawa, Masashi

doi:10.1053/j.gastro.2022.03.053

Cited by 22 publications

(34 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, such a system requires a large training dataset, as evidenced in a recent single-center study where sensitivity and specificity for differentiating SSLs from hyperplastic polyps were 80.8% and 62.1%, respectively. 19 Our CNN achieved favorable results without using a confidence threshold despite a smaller sample size of polyps for training than other models. A plausible explanation is that annotating high-and moderate-quality video frames compensated for the modest number of polyps.…”

Section: Discussionmentioning

confidence: 87%

See 1 more Smart Citation

Polyp characterization using deep learning and a publicly accessible polyp video database

Kader

Cid‐Mejias²,

Brandão³

et al. 2023

Digestive Endoscopy

View full text Add to dashboard Cite

Objectives: Convolutional neural networks (CNN) for computer-aided diagnosis of polyps are often trained using high-quality still images in a single chromoendoscopy imaging modality with sessile serrated lesions (SSLs) often excluded. This study developed a CNN from videos to classify polyps as adenomatous or nonadenomatous using standard narrow-band imaging (NBI) and NBI-near focus (NBI-NF) and created a publicly accessible polyp video database. Methods:We trained a CNN with 16,832 high and moderate quality frames from 229 polyp videos (56 SSLs). It was evaluated with 222 polyp videos (36 SSLs) across two test-sets. Test-set I consists of 14,320 frames (157 polyps, 111 diminutive). Test-set II, which is publicly accessible, 3317 video frames (65 polyps, 41 diminutive), which was benchmarked with three expert and three nonexpert endoscopists.Results: Sensitivity for adenoma characterization was 91.6% in test-set I and 89.7% in test-set II. Specificity was 91.9% and 88.5%.Sensitivity for diminutive polyps was 89.9% and 87.5%; specificity 90.5% and 88.2%. In NBI-NF, sensitivity was 89.4% and 89.5%, with a specificity of 94.7% and 83.3%. In NBI, sensitivity was 85.3% and 91.7%, with a specificity of 87.5% and 90.0%, respectively. The CNN achieved preservation and incorporation of valuable endoscopic innovations (PIVI)-1 and PIVI-2 thresholds for each test-set. In the benchmarking of test-set II, the CNN was significantly more accurate than nonexperts (13.8% difference [95% confidence interval 3.2-23.6], P = 0.01) with no significant difference with experts.Conclusions: A single CNN can differentiate adenomas from SSLs and hyperplastic polyps in both NBI and NBI-NF. A publicly accessible NBI polyp video database was created and benchmarked.

show abstract

Section: Discussionmentioning

confidence: 87%

“…A multitier classification that separates each polyp category would be of greater clinical benefit. However, such a system requires a large training dataset, as evidenced in a recent single‐center study where sensitivity and specificity for differentiating SSLs from hyperplastic polyps were 80.8% and 62.1%, respectively 19 …”

Section: Discussionmentioning

confidence: 99%

Polyp characterization using deep learning and a publicly accessible polyp video database

Kader

Cid‐Mejias²,

Brandão³

et al. 2023

Digestive Endoscopy

View full text Add to dashboard Cite

show abstract

“…Minegishi et al reported a single-arm prospective study using CADx for narrow-band imaging (NBI-CADx) classifying 3-classes: hyperplastic polyp, sessile serrated lesion, and neoplastic/adenomatous. Using still NBI images, the NBI-CAD achieved an overall 93.3% sensitivity and 61.5% specificity for diagnosing diminutive (< 5 mm) neoplasms [ 30 ]. Similar results were reported by Ozawa et al using about 17,000 images from about 4800 colorectal polyps and about 4000 normal colorectum images [ 31 ].…”

Section: Discussionmentioning

confidence: 99%

Robust automated prediction of the revised Vienna Classification in colonoscopy using deep learning: development and initial external validation

et al. 2022

View full text Add to dashboard Cite

Background Improved optical diagnostic technology is needed that can be used by also outside expert centers. Hence, we developed an artificial intelligence (AI) system that automatically and robustly predicts the pathological diagnosis based on the revised Vienna Classification using standard colonoscopy images. Methods We prepared deep learning algorithms and colonoscopy images containing pathologically proven lesions (56,872 images, 6775 lesions). Four classifications were adopted: revised Vienna Classification category 1, 3, and 4/5 and normal images. The best algorithm—ResNet152—in the independent internal validation (14,048 images, 1718 lesions) was used for external validation (255 images, 128 lesions) based on neoplastic and non-neoplastic classification. Diagnostic performance of endoscopists was compared using a computer-assisted interpreting test. Results In the internal validation, the sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and accuracy for adenoma (category 3) of 84.6% (95% CI 83.5–85.6%), 99.7% (99.5–99.8%), 90.8% (89.9–91.7%), 89.2% (88.5–99.0%), and 89.8% (89.3–90.4%), respectively. In the external validation, ResNet152’s sensitivity, specificity, PPV, NPV, and accuracy for neoplastic lesions were 88.3% (82.6–94.1%), 90.3% (83.0–97.7%), 94.6% (90.5–98.8%), 80.0% (70.6–89.4%), and 89.0% (84.5–93.6%), respectively. This diagnostic performance was superior to that of expert endoscopists. Area under the receiver-operating characteristic curve was 0.903 (0.860–0.946). Conclusions The developed AI system can help non-expert endoscopists make differential diagnoses of colorectal neoplasia on par with expert endoscopists during colonoscopy. (229/250 words).

show abstract

“…These advances have already led to the introduction of the first commercially available, regulatory-approved CADx systems [6]. Although several of these CADx systems were able to reach the PIVI and SODA thresholds within prospective validation settings, none of the systems have yet been able to reach all performance thresholds [7][8][9][10][11]. In addition, there remain several limitations that might hamper widespread implementation of these systems in clinical practice [12].…”

Section: Introductionmentioning

confidence: 99%

Computer-aided diagnosis for optical diagnosis of diminutive colorectal polyps including sessile serrated lesions: a real-time comparison with screening endoscopists

Houwen

Hazewinkel

Giotis³

et al. 2023

Endoscopy

View full text Add to dashboard Cite

Background: In this study, we aimed to compare the accuracy for optical diagnosis of diminutive colorectal polyps between a CADx system (POLAR system) and endoscopists during real-time colonoscopy, including sessile serrated lesions (SSLs). Patients and methods: We developed the POLAR system, capable of performing real-time characterization of diminutive colorectal polyps. For pre-training, the Microsoft-COCO dataset with non-polyp object images (300k) was used. For training, 8 hospitals prospectively collected 2.637 annotated images from 1.339 polyps (i.e. new publicly available POLAR database). For clinical validation, POLAR was tested during colonoscopy in fecal immunochemical test (FIT)-positive individuals, and compared with the performance of 20 endoscopists from 8 hospitals. Endoscopists were blinded to the POLAR outcome. Primary outcome was the comparison of accuracy in the optical diagnosis of diminutive colorectal polyps between POLAR and endoscopists (neoplastic [adenomas and SSLs] vs non-neoplastic [hyperplastic polyps]). Histopathology served as the reference standard. Results: During clinical validation, a total of 423 diminutive polyps detected in 194 FIT-positive individuals were included for analysis (300 adenomas, 41 SSLs, 82 hyperplastic polyps). POLAR distinguished neoplastic from non-neoplastic lesions with 79% accuracy, 89% sensitivity and 38% specificity. The endoscopists achieved 83% accuracy, 92% sensitivity, and 44% specificity. The optical diagnosis accuracy between POLAR and endoscopists was not significantly different (P=.10). The success rate for acquiring a histological prediction by POLAR was 98%. Conclusions: We have developed a CADx system that differentiated neoplastic from non-neoplastic diminutive polyps during endoscopy with an accuracy comparable to the accuracy of screening endoscopists, with near-perfect technical efficacy.

show abstract

Comprehensive Diagnostic Performance of Real-Time Characterization of Colorectal Lesions Using an Artificial Intelligence–Assisted System: A Prospective Study

Cited by 22 publications

References 15 publications

Polyp characterization using deep learning and a publicly accessible polyp video database

Polyp characterization using deep learning and a publicly accessible polyp video database

Robust automated prediction of the revised Vienna Classification in colonoscopy using deep learning: development and initial external validation

Computer-aided diagnosis for optical diagnosis of diminutive colorectal polyps including sessile serrated lesions: a real-time comparison with screening endoscopists

Contact Info

Product

Resources

About