Optical diagnosis of colorectal polyps: a randomized controlled trial comparing endoscopic image-enhancing modalities

Autonomous Artificial Intelligence vs Artificial Intelligence-Assisted Human Optical Diagnosis of Colorectal Polyps: A Randomized Controlled Trial

BACKGROUND & AIMS

Artificial intelligence (AI)-based optical diagnosis systems (CADx) have been developed to allow pathology prediction of colorectal polyps during colonoscopies. However, CADx systems have not yet been validated for autonomous performance. Therefore, we conducted a trial comparing autonomous AI to AI-assisted human (AI-H) optical diagnosis.

METHODS

We performed a randomized noninferiority trial of patients undergoing elective colonoscopies at 1 academic institution. Patients were randomized into (1) autonomous AI-based CADx optical diagnosis of diminutive polyps without human input or (2) diagnosis by endoscopists who performed optical diagnosis of diminutive polyps after seeing the real-time CADx diagnosis. The primary outcome was accuracy in optical diagnosis in both arms using pathology as the gold standard. Secondary outcomes included agreement with pathology for surveillance intervals.

RESULTS

A total of 467 patients were randomized (238 patients/158 polyps in the autonomous AI group and 229 patients/179 polyps in the AI-H group). Accuracy for optical diagnosis was 77.2% (95% confidence interval [CI], 69.7-84.7) in the autonomous AI group and 72.1% (95% CI, 65.5-78.6) in the AI-H group (P = .86). For high-confidence diagnoses, accuracy for optical diagnosis was 77.2% (95% CI, 69.7-84.7) in the autonomous AI group and 75.5% (95% CI, 67.9-82.0) in the AI-H group. Autonomous AI had statistically significantly higher agreement with pathology-based surveillance intervals compared to AI-H (91.5% [95% CI, 86.9-96.1] vs 82.1% [95% CI, 76.5-87.7]; P = .016).

CONCLUSIONS

Autonomous AI-based optical diagnosis exhibits noninferior accuracy to endoscopist-based diagnosis. Both autonomous AI and AI-H exhibited relatively low accuracy for optical diagnosis; however, autonomous AI achieved higher agreement with pathology-based surveillance intervals. (ClinicalTrials.gov, Number NCT05236790).

Risk of total metachronous advanced neoplasia at surveillance colonoscopy after detection of serrated lesions: a matched case-cohort study

BACKGROUND : Serrated lesions are potential colorectal cancer precursors. This study evaluated the presence of total metachronous advanced neoplasia (T-MAN) at follow-up in patients with index serrated lesions compared with a matched cohort without serrated lesions. METHODS : Patients aged 45-74 years with serrated lesions were matched 2:1 by sex, age, synchronous polyps, and timing of index colonoscopy, to patients without serrated lesions. The primary outcome was T-MAN (advanced adenoma or high-risk serrated lesion) at follow-up. Secondary outcomes included presence of T-MAN stratified by synchronous polyps and serrated lesion characteristics. RESULTS : 1425 patients were included (475 patients, 642 serrated lesions; 950 controls; median follow-up 2.9 versus 3.6 years). Patients with serrated lesions had greater risk of T-MAN than those without (hazard ratio [HR] 6.1, 95 %CI 3.9-9.6). Patients with serrated lesions and high-risk adenoma (HRA) had higher risk of T-MAN than those with HRA alone (HR 2.6, 95 %CI 1.4-4.7); similarly, patients with serrated lesions plus low-risk adenoma (LRA) had higher risk than those with LRA alone (HR 7.0, 95 %CI 2.8-18.4), as did patients with serrated lesions without adenoma compared with no adenoma (HR 14.9, 95 %CI 6.5-34.0). Presence of proximal sessile serrated lesion (SSL; HR 9.3, 95 %CI 5.4-15.9), large SSL (HR 17.8, 95 %CI 7.4-43.3), and proximal large SSL (HR 25.0, 95 %CI 8.8-71.3), but not distal SSL, were associated with greater risk for T-MAN. CONCLUSION : Patients with serrated lesions had higher risk for T-MAN regardless of synchronous adenomas. Patients with serrated lesions and HRA, and those with large or proximal SSLs, were at greatest risk.

Comparing size measurement of colorectal polyps using a novel virtual scale endoscope, endoscopic ruler or forceps: A preclinical randomized trial

Background and study aims  Accurate polyp size measurement is important for guideline conforming choice of polypectomy techniques and subsequent surveillance interval assignments. Some endoscopic tools (biopsy forceps [BF] or endoscopic rulers [ER]) exist to help with visual size estimation. A virtual scale endoscope (VSE) has been developed that allows superimposing a virtual measurement scale during live endoscopies. Our aim was to evaluate the performance of VSE when compared to ER and BF-based measurement. Methods  We conducted a preclinical randomized trial to evaluate the relative accuracy of size measurement of simulated colorectal polyps when using: VSE, ER, and BF. Six endoscopists performed 60 measurements randomized at a 1:1:1 ratio using each method. Primary outcome was relative accuracy in polyp size measurement. Secondary outcomes included misclassification of sizes at the 5-, 10-, and 20-mm thresholds. Results  A total of 360 measurements were performed. The relative accuracy of BF, ER, and VSE was 78.9 % (95 %CI = 76.2-81.5), 78.4 % (95 %CI = 76.0-80.8), and 82.7 % (95 %CI = 80.8-84.8). VSE had significantly higher accuracy compared to BF ( P  = 0.02) and ER ( P  = 0.006). VSE misclassified a lower percentage of polyps > 5 mm as ≤ 5 mm (9.4 %) compared to BF (15.7 %) and ER (20.9 %). VSE misclassified a lower percentage of ≥ 20 mm polyps as < 20 mm (8.3 %) compared with BF (66.7 %) and ER (75.0 %). Of polyps ≥10mm, 25.6 %, 25.5 %, and 22.5 % were misclassified as <10 mm with ER, BF, and VSE, respectively. Conclusions  VSE had significantly higher relative accuracy in measuring polyps compared to ER or BF assisted measurement. VSE improves correct classification of polyps at clinically important size thresholds.

Analysis of the Influence of Comprehensive Nursing Intervention on Vital Signs and Negative Emotions of Patients with Gastrointestinal Polyps Treated by Digestive Endoscopy

This research sets out to elucidate the influence of comprehensive nursing intervention (CNI) on vital signs (VSs) and negative emotions (NEs) of patients with gastrointestinal polyps treated by digestive endoscopy. To address this, from January 2020 to February 2021, 92 cases of gastrointestinal polyps treated by digestive endoscopy in the Yuyao People's Hospital of Zhejiang Province were partitioned into two different groups: the control group (CG; n = 46) intervened by routine nursing intervention, and the research group (RG; n = 46) treated by CNI based on routine care. The VSs, NEs, time of first postoperative exhaust and defecation, early food intake, length of stay (LOS), quality of life (QOL), nursing satisfaction, and the incidence of postoperative complications were compared between groups. When entering the operating room, elevated systolic/diastolic (SBP/DBP) was observed in CG compared with RG (P < 0.05) while there was no statistical difference in SBP/DBP in RG before intervention and when entering the operating room (P > 0.05). Statistically, less time of first exhaust and defecation as well as LOS and early food intake were observed in RG (P < 0.05). The HAMA and HAMD scores declined in both arms postintervention, and the decrease was more obvious in RG (P < 0.05). After intervention, the scores of various domains of GQOLI-74 elevated in both arms, and the increase was more obvious in RG (P < 0.05). RG also showed a lower total complication rate and a higher degree of nursing satisfaction than CG (P < 0.05). CNI for patients with gastrointestinal polyps treated by digestive endoscopy can obviously relieve their anxiety and depression, with stable VSs, short LOS, and low incidence of complications.

Comparison of blue laser imaging and narrow band imaging for the differentiation of diminutive colorectal polyps: A randomized controlled trial

BACKGROUND

To compare the diagnostic efficacy of blue laser imaging (BLI)- bright and narrow band imaging (NBI) modes of image enhanced endoscopy (IEE) in differentiating neoplastic and non-neoplastic lesions of diminutive colorectal polyps.

METHODS

We conducted a prospective randomized controlled trial from September 2015 to July 2016. The participants were randomly assigned (1:1) for colonoscopy with polyp classification under NBI or BLI-bright mode without magnification. Histopathologic diagnosis was used as the gold standard.

RESULTS

Three hundred and twenty-four diminutive polyps in 164 patients were included for analysis (BLI: 162 polyps in 73 patients, NBI: 162 polyps in 91 patients). These polyps were located at colon proximal to sigmoid (61.1 and 58.0%) and rectosigmoid colon (38.9 and 42.0%) in the BLI and NBI groups, respectively. Most polyps (71.9%) were adenomatous with one malignant polyp (0.3%). BLI achieved 86.4% accuracy, 98.3% sensitivity, 55.6% specificity, 85.2% positive predictive value (PPV), and 92.6% negative predictive value (NPV), similar to NBI which exhibited 90.1% accuracy, 99.1% sensitivity, 67.4% specificity, 88.5% PPV, and 96.9% NPV in the diagnosis of adenomatous polyps. Based on the location of the polyp, both modes of IEE provided ≥ 95% NPV for diagnosis of adenomatous polyps at the rectosigmoid colon.

CONCLUSIONS

BLI-bright and NBI modes of IEE have similar accuracy in differentiation between neoplastic and non-neoplastic lesions of diminutive polyps. Both modes provided ≥ 90% NPV which allows for the adaptation of the American Society of Gastrointestinal Endoscopy "diagnose-and-leave" recommended strategy for diminutive polyps at the rectosigmoid colon.

Can image-enhanced endoscopy improve adenoma detection rate?

An accumulating body of evidence has shown that detection and resection of pre-cancerous adenoma by colonoscopy could effectively prevent colorectal cancer (CRC) and its related mortality. Among various colonoscopy quality indicators, such as cecal intubation rate, withdrawal time, and adenoma detection rate (ADR); ADR is the most important and most closely associated with the subsequent risk of CRC. Image-enhanced endoscopy (IEE), including digital and dye-based IEE, was originally developed to discriminate neoplastic from non-neoplastic lesions but later studies have demonstrated that it can also enhance lesion detection by enhancing the contrast between the lesion and background colonic mucosa. Nevertheless, using IEE in colonoscopy for lesion detection is still not the standard way of practice in the real world. For a better understanding of current IEE modalities, this review introduces and compares the currently available IEE modalities and their efficacy in detecting adenoma from the results of randomized controlled trials or meta-analyses.