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).

Using Computer-aided Optical Diagnosis and Expert Review to Evaluate Colorectal Polyps Diagnosed as Normal Mucosa in Pathology.

Pathological assessment of colorectal polyps is considered the current reference standard for histologic diagnosis. About 10% of polyps sent to the pathology lab are returned with the diagnosis of mucosal folds, mucosal prolapse, or normal mucosa.1,2 Two recent publications have indicated that disagreements between endoscopic optical diagnosis and the subsequent pathological diagnoses might be due to misdiagnosis in pathology.3,4 We were therefore interested in re-evaluating pathology-based diagnosis of "mucosal polyps" using expert endoscopists and computer-assisted diagnosis (CADx) evaluation.

Measuring Size of Colorectal Polyps Using a Virtual Scale Endoscope or Visual Assessment: A Randomized Controlled Trial.

INTRODUCTION: This study aimed to compare the accuracy of polyp size measurements using a virtual scale endoscope (VSE) with an integrated laser-based adaptive scale function and visual assessment (VA) during colonoscopies. METHODS: We conducted a single-blinded, prospective randomized controlled trial. Eligible patients (aged 45-80 years) undergoing screening, surveillance, or diagnostic colonoscopies were randomly assigned (1:1) into 2 groups. In the intervention group, all detected polyps were measured for size using VSE; in the control group, all polyps were measured using VA. Size measurements were compared with a reference standard of digital caliper measurement immediately post polypectomy. The primary outcome was the relative accuracy of real-time VSE measurement compared with VA. Secondary outcomes included the mean differences and the correlations between VSE or VA sizes and the reference standard of measurement. RESULTS: Overall, 230 patients were enrolled and randomized. The relative size measurement accuracy of VSE was 84% in 118 polyps, which was significantly higher than that of VA (105 polyps; 68.4%, P < 0.001). VSE resulted in a significantly higher percentage of size measurements within 25% of true size compared with VA (81.4% vs 41%, P < 0.001). VSE had a significantly lower percentage for >5-mm polyps incorrectly sized as 1-5 mm compared with VA (13.5% vs 57.1%; P < 0.001) and a significantly lower percentage for >3-mm polyps incorrectly sized as 1-3 mm compared with VA (11.3% vs 56.5%; P < 0.001). DISCUSSION: VSE significantly improves the size measurement accuracy of colorectal polyps during colonoscopies compared with VA and results in fewer misclassifications at relevant decision-making size thresholds.

Effect of fecal immunochemical test cut-off levels on adenoma detection rate: a systematic review and meta-analysis.

BACKGROUND: Adenoma detection rate (ADR) is higher after a positive fecal immunochemical test (FIT) compared to direct screening colonoscopy. OBJECTIVE: This meta-analysis evaluated how ADR, the rates of advanced adenoma detection (AADR), colorectal cancer detection (CDR), and sessile serrated lesion detection (SSLDR) are affected by different FIT positivity thresholds. METHODS: We searched MEDLINE, EMBASE, CINAHL, and EBM Reviews databases for studies reporting ADR, AADR, CDR, and SSLDR according to different FIT cut-off values in asymptomatic average-risk individuals aged 50-74 years old. Data were stratified according to sex, age, time to colonoscopy, publication year, continent, and FIT kit type. Study quality, heterogeneity, and publication bias were assessed. RESULTS: Overall, 4280 articles were retrieved and fifty-eight studies were included (277,661 FIT-positive colonoscopies; mean cecal intubation 96.3%; mean age 60.8 years; male 52.1%). Mean ADR was 56.1% (95% CI 53.4 - 58.7%), while mean AADR, CDR, and SSLDR were 27.2% (95% CI 24.4 - 30.1%), 5.3% (95% CI 4.7 - 6.0%), and 3.0% (95% CI 1.7 - 4.6%), respectively. For each 20 µg Hb/g increase in FIT cut-off level, ADR increased by 1.54% (95% CI 0.52 - 2.56%, p < 0.01), AADR by 3.90% (95% CI 2.76 - 5.05%, p < 0.01) and CDR by 1.46% (95% CI 0.66 - 2.24%, p < 0.01). Many detection rates were greater amongst males and Europeans. CONCLUSIONS: ADRs in FIT-positive colonoscopies are influenced by the adopted FIT positivity threshold, and identified targets, importantly, proved to be higher than most current societal recommendations.

Expert endoscopist assessment of colorectal polyp size using virtual scale endoscopy, visual or snare-based estimation: a prospective video-based study.

BACKGROUND AND AIMS: Accurate polyp size estimation during colonoscopy has an impact on clinical decision-making. A laser-based virtual scale endoscope (VSE) is available to allow measuring polyp size using a virtual adaptive scale. This study evaluates video-based polyp size measurement accuracy among expert endoscopists using either VSE or visual assessment (VA) with either snare as reference size or without any reference size information. METHODS: A prospective, video-based study was conducted with 10 expert endoscopists. Video sequences from 90 polyps with known reference size (fresh specimen measured using calipers) were distributed on three different slide sets so that each slide set showed the same polyp only once with either VSE, VA or snare-based information. A slide set was randomly assigned to each endoscopist. Endoscopists were asked to provide size estimation based on video review. RESULTS: Relative accuracies for VSE, VA, and snare-based estimation were 75.1% (95% CI [71.6-78.5]), 65.0% (95% CI [59.5-70.4]) and 62.0% (95% CI [54.8-69.0]), respectively. VSE yielded significantly higher relative accuracy compared to VA (p = 0.002) and to snare (p = 0.001). A significantly lower percentage of polyps 1-5 mm were misclassified as >5 mm using VSE versus VA and snare (6.52% vs. 19.6% and 17.5%, p = 0.004) and a significantly lower percentage of polyps >5 mm were misclassified as 1-5 mm using VSE versus VA and snare (11.4% vs. 31.9% and 14.9%, p = 0.038). CONCLUSIONS: Endoscopists estimate polyp size with the highest accuracy when virtual adaptive scale information is displayed. Using a snare to assist sizing did not improve measurement accuracy compared to displaying visual information alone.

Comparing size measurements of simulated colorectal polyp size and morphology groups when using a virtual scale endoscope or visual size estimation: Blinded randomized controlled trial.

OBJECTIVES: The virtual scale endoscope (VSE) allows projection of a virtual scale onto colorectal polyps allowing real-time size measurements. We studied the relative accuracy of VSE compared to visual assessment (VA) for the measuring simulated polyps of different size and morphology groups. METHODS: We conducted a blinded randomized controlled trial using simulated polyps within a colon model. Sixty simulated polyps were evenly distributed across four size groups (1-5, >5-9.9, 10-19.9, and ≥20 mm) and three Paris morphology groups (flat, sessile, and pedunculated). Six endoscopists performed polyp size measurements using random allocation of either VA or VSE. RESULTS: A total of 359 measurements were completed. The relative accuracy of VSE was significantly higher when compared to VA for all size groups >5 mm (P = 0.004, P < 0.001, P < 0.001). For polyps ≤5 mm, the relative accuracy of VSE compared to VA was not significantly higher (P = 0.186). The relative accuracy of VSE was significantly higher when compared to VA for all morphology groups. VSE misclassified a lower percentage of >5 mm polyps as ≤5 mm (2.9%), ≥10 mm polyps as <10 mm (5.5%), and ≥20 mm polyps as <20 mm (21.7%) compared to VA (11.2%, 24.7%, and 52.3% respectively; P = 0.008, P < 0.001, and P = 0.003). CONCLUSION: Virtual scale endoscope had significantly higher relative accuracies for every polyp size group or morphology type aside from diminutive. VSE enables the endoscopist to better classify polyps into correct size categories at clinically relevant size thresholds of 5, 10, and 20 mm.

EMR combined with hybrid argon plasma coagulation to prevent recurrence of large nonpedunculated colorectal polyps (with videos).

BACKGROUND AND AIMS: EMR is the mainstay of therapy for large colorectal polyps. Local recurrence after EMR is common and can be reduced using margin ablation. Our aim was to evaluate recurrence rates when using hybrid argon plasma coagulation (h-APC) ablation after EMR. METHODS: Adult patients (aged 18-89 years) undergoing EMR of nonpedunculated colorectal polyps ≥20 mm were enrolled in a prospective multicenter study. h-APC was used to ablate all defect margins and also the resection surface in selected cases. The primary study outcome was recurrence rates found during the first follow-up colonoscopy. Secondary outcomes were technical success and adverse event rates. RESULTS: EMR with h-APC ablation was used in 101 polyps (84 patients, 46.4% women). EMR with h-APC ablation was technically successful in all cases (median EMR time, 15 minutes; median h-APC ablation time, 4 minutes). Median polyp size was 30 mm (range, 20-60). Resected polyps were either adenomas (68/101 [67.3%]), sessile serrated lesions (27/101 [27%]), or adenocarcinomas (6/101 [6%]). The post-EMR recurrence rate was 2.2% (2/91) (95% confidence interval, .27-7.71). All 6 patients with cancer (intramucosal cancer, 4; T1sm cancer, 2) were found to have complete eradication of the primary tumor after EMR with h-APC, and none had lymph node metastasis. Four serious adverse events occurred in 3 patients (2 delayed bleeding [2.4%], 1 abdominal pain [1.2%], and 1 microperforation [1.2%]. All serious adverse events resolved with either endoscopic or antibiotic treatment only. CONCLUSIONS: EMR with h-APC showed a high technical success rate, low adverse event rate, and very low post-EMR recurrence rates. (Clinical trial registration number: NCT04015765.).

What size cutoff level should be used to implement optical polyp diagnosis?

BACKGROUND : The risk of advanced pathology increases with polyp size, as does the potential for mismanagement when optical diagnosis is used. This study aimed to evaluate the proportion of patients who would be assigned inadequate surveillance intervals when different size cutoffs are adopted for use of optical diagnosis. METHODS : In a post hoc analysis of three prospective studies, the use of optical diagnosis was evaluated for three polyp size groups: 1-3, 1-5, and 1-10 mm. The primary outcome was the proportion of patients in whom advanced adenomas were found and optical diagnosis resulted in delayed surveillance. Secondary outcomes included agreements between surveillance intervals based on high confidence optical diagnosis and pathology outcomes, reduction in histopathological examinations, and proportion of patients who could receive an immediate surveillance recommendation. RESULTS : We included 3374 patients (7291 polyps ≤ 10 mm) undergoing complete colonoscopies (median age 66.0 years, 75.2 % male, 29.6 % for screening). The percentage of patients with advanced adenomas and either 2- or 7-year delayed surveillance intervals (n = 79) was 3.8 %, 15.2 %, and 25.3 % for size cutoffs of 1-3, 1-5, and 1-10 mm polyps, respectively (P < 0.001). Surveillance interval agreements between pathology and optical diagnosis for the three groups were 97.2 %, 95.5 %, and 94.2 %, respectively. Total reductions in pathology examinations for the three groups were 33.5 %, 62.3 %, and 78.2 %, respectively. CONCLUSION : A 3-mm cutoff for clinical implementation of optical diagnosis resulted in a very low risk of delayed management of advanced neoplasia while showing high surveillance interval agreement with pathology and a one-third reduction in overall requirement for pathology examinations.

The location-based resect and discard strategy for diminutive colorectal polyps: a prospective clinical study.

BACKGROUND: Clinical implementation of the resect-and-discard strategy has been difficult because optical diagnosis is highly operator dependent. This prospective study aimed to evaluate a resect-and-discard strategy that is not operator dependent. METHODS: The study evaluated a resect-and-discard strategy that uses the anatomical polyp location to classify colonic polyps into non-neoplastic or low risk neoplastic. All rectosigmoid diminutive polyps were considered hyperplastic and all polyps located proximally to the sigmoid colon were considered neoplastic. Surveillance interval assignments based on these a priori assumptions were compared with those based on actual pathology results and on optical diagnosis. The primary outcome was ≥ 90 % agreement with pathology in surveillance interval assignment. RESULTS: 1117 patients undergoing complete colonoscopy were included and 482 (43.1 %) had at least one diminutive polyp. Surveillance interval agreement between the location-based strategy and pathological findings using the 2020 US Multi-Society Task Force guideline was 97.0 % (95 % confidence interval [CI] 0.96-0.98), surpassing the ≥ 90 % benchmark. Optical diagnoses using the NICE and Sano classifications reached 89.1 % and 90.01 % agreement, respectively (P < 0.001), and were inferior to the location-based strategy. The location-based resect-and-discard strategy allowed a 69.7 % (95 %CI 0.67-0.72) reduction in pathology examinations compared with 55.3 % (95 %CI 0.52-0.58; NICE and Sano) and 41.9 % (95 %CI 0.39-0.45; WASP) with optical diagnosis. CONCLUSION: The location-based resect-and-discard strategy achieved very high surveillance interval agreement with pathology-based surveillance interval assignment, surpassing the ≥ 90 % benchmark and outperforming optical diagnosis in surveillance interval agreement and the number of pathology examinations avoided.

Novel clinical scores for occupational asthma due to exposure to high-molecular-weight agents.

OBJECTIVE: Specific inhalation challenge (SIC) as the reference diagnostic test for occupational asthma (OA) is not widely available worldwide. We aimed to develop non-SIC-based models for OA. METHODS: Of 427 workers who were exposed to high-molecular-weight agents and referred to OA clinic at Montréal Sacré-Cœur Hospital between 1983 and 2016, we analysed 160 workers who completed non-specific bronchial hyper-responsiveness (NSBHR) tests and still worked 1 month before SIC. OA was defined as positive SIC. Logistic regression models were developed. The accuracy of the models was quantified using calibration and discrimination measures. Their internal validity was evaluated with bootstrapping procedures. The final models were translated into clinical scores and stratified into probability groups. RESULTS: The final model, which included age ≤40 years, rhinoconjunctivitis, inhaled corticosteroid use, agent type, NSBHR, and work-specific sensitisation had a reasonable internal validity. The area under the receiver operating characteristics curve (AUC) was 0.91 (95% CI 0.86 to 0.95), statistically significantly higher than the combination of positive NSBHR and work-specific sensitisation (AUC=0.84). The top 70% of the clinical scores (ie, the high probability group) showed a significantly higher sensitivity (96.4%vs86.9%) and negative predictive value (93.6%vs84.1%) than the combination of positive NSBHR and work-specific sensitisation (p value <0.001). CONCLUSIONS: We developed novel scores for OA induced by high-molecular-weight agents with excellent discrimination. It could be helpful for secondary-care physicians who have access to pulmonary function test and allergy testing in identifying subjects at a high risk of having OA and in deciding on appropriate referral to a tertiary centre.

A Preclinical Blinded Randomized-Controlled Trial Evaluating the Clinical Relevance of Polyp Size Measurement Using a Virtual Scale Endoscope.

Background: The virtual scale endoscope (VSE) helps endoscopists measure colorectal polyp size more accurately compared to visual assessment (VA). However, previous studies were not adequately powered to evaluate the sizing of polyps at clinically relevant size thresholds and relative accuracy for size subgroups. Methods: We created 64 artificial polyps of varied sizes and Paris class morphology, randomly assigned 1:1 to be measured (383 total measurement datapoints with VSE and VA by 6 endoscopists blinded to true size) in a colon model. We added data from two previous trials (480 measurement datapoints). We evaluated for correct classification of polyps into size groups at 3 mm, 5 mm, 10 mm, and 20 mm size thresholds and the relative size measurement accuracy for diminutive polyps (≤5 mm), small polyps (5-9 mm), large polyps at 10-19 mm, and polyps (≥20). Results: VSE had significantly less size group misclassifications at the 5 mm, and 10 mm thresholds (28 percent vs. 45 percent, P = 0.0159 and 26 percent vs. 44 percent, P = 0.0135, respectively). For the 3 mm and 20 mm thresholds, VSE had lower misclassifications; however, this was not statistically significant (36 percent vs. 46 percent, P = 0.3853 and 38 percent vs. 41 percent, P = 0.2705, respectively). The relative size measurement accuracy was significantly higher for VSE compared to VA for all size subgroups (diminutive (P < 0.01), small polyps (P < 0.01), 10-19 mm (P < 0.01), and ≥20 mm (P < 0.01)). Conclusion: VSE outperforms VA in categorizing polyps into size groups at the clinically relevant size thresholds of 5 mm and 10 mm. Using VSE resulted in significantly higher relative measurement accuracy for all size subgroups.

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.

Large serrated polyps indicate a greater risk of advanced metachronous colorectal neoplasia than high-grade adenomas.

Background and study aims The risk of developing total metachronous advanced neoplasia (TMAN) in patients with index serrated lesions (SL) or adenoma with high-grade dysplasia (HGD) is unknown. We evaluated this risk in patients with either HGD, SL < 10 mm or SL ≥ 10 mm at index colonoscopy, who underwent surveillance colonoscopies. Patients and methods This retrospective cohort study evaluated all consecutive patients (n = 2477) diagnosed between 2010 and 2019 with colorectal HGD, SLs < 10 mm or SLs ≥ 10 mm. We excluded patients aged < 45 or > 75 years or those who had inflammatory bowel disease, hereditary colorectal cancer (CRC) syndromes, previous or synchronous CRC, or no follow-up colonoscopy. Descriptive variables were compared using analysis of variance or Pearson chi-squared tests. Multivariate Cox regressions were used to compare the risk of TMAN between the HGD, SL < 10 mm and SL ≥ 10 mm groups. Results Overall, 585 patients (mean age 63 years; 55% male; mean follow-up 3.67 years) were included (226 with SLs < 10 mm, 204 with SLs ≥ 10 mm, 155 with HGD). Compared with SLs < 10 mm, patients with HGD did not have a significantly different rate of TMAN (HR=0.75 [0.39-1.44]) and patients with SLs ≥ 10 mm had a higher rate of TMAN (HR=2.08 [1.38-3.15]). Compared with HGD, patients with SLs ≥ 10 mm had a higher rate of TMAN (HR=1.87 [1.04-3.36]). Conclusions The risk for TMAN was higher for patients with SLs ≥ 10 mm than with HGD or SLs < 10 mm. This risk should be considered when planning surveillance intervals for patients diagnosed with large SLs.