ABSTRACT
INTRODUCTION: This study aims to evaluate the effectiveness of ChatGPT (Chat Generative Pretrained Transformer) in answering patients' questions about colorectal cancer (CRC) screening, with the ultimate goal of enhancing patients' awareness and adherence to national screening programs. METHODS: 15 questions on CRC screening were posed to ChatGPT4. The answers were rated by 20 gastroenterology experts and 20 non-experts in three domains (accuracy, completeness, and comprehensibility), and by 100 patients in three dichotomic domains (completeness, comprehensibility and trustability). RESULTS: According to expert rating, the mean accuracy score was 4.8±1.1 on a scale ranging from 1 to 6. Men completeness score was 2.1±0.7 and mean comprehensibility score was 2.8±0.4 on a scale ranging from 1 to 3. Overall, accuracy (4.8±1.1 vs 5.6±0.7, P<0.001) and completeness (2.1±0.7 vs 2.7±0.4, P<0.001) scores were significantly lower for expert compared to non-expert, while comprehensibility was comparable among the two groups (2.7±0.4 vs 2.8±0.3, P=0.546). Patients rated all questions as complete, comprehensible and trustable in 97 to 100% of cases. CONCLUSIONS: ChatGPT shows good performance with the potential to enhance awareness about CRC and improve screening outcomes. Generative language systems may be further improved after proper training in accordance with scientific evidence and current guidelines.
ABSTRACT
BACKGROUND AND AIMS: Limited evidence exists regarding the impact of glucagon-like peptide-1 receptor agonists (GLP-1RAs) on upper endoscopy. Therefore, a meta-analysis was conducted to comprehensively review the available evidence on this subject. METHODS: A systematic bibliographic search was carried out until May 2024. Pooled estimates were analyzed using a random-effects model, with results presented as odds ratios (OR) and 95% confidence intervals (CI). The primary outcome assessed was the rate of retained gastric content (RGC), while secondary outcomes included rates of aborted and repeated procedures, adverse event (AE) rate, and rates of aspiration. RESULTS: This analysis included 13 studies involving a total of 84,065 patients. Patients receiving GLP-1RA therapy exhibited significantly higher rates of RGC (OR 5.56, 3.35-9.23), a trend that was consistent among patients with diabetes (OR 2.60, 2.23-3.02). Adjusted analysis, accounting for variables such as sex, age, body mass index (BMI), diabetes, and other therapies, confirmed the elevated rates of RGC in the GLP-1RA user group (aOR 4.20, 3.42-5.15). Furthermore, rates of aborted and repeated procedures were higher in the GLP-1RA user group (OR 5.13, 3.01-8.75, and OR 2.19, 1.43-3.35; respectively). However, no significant differences were found in AE and aspiration rates between the two groups (OR 4.04, 0.63-26.03, and OR 1.75, 0.64-4.77; respectively). CONCLUSION: Use of GLP-1RAs is associated with increased retention of gastric contents and more frequent aborted procedures during upper endoscopy. However, the AEs and aspiration rates do not seem different, therefore adjusting fasting time instead of routinely withholding GLP-1RAs could be reasonable in these patients.
ABSTRACT
OBJECTIVE: To develop and externally validate an updated artificial intelligence (AI) prediction system for stratifying the risk of lymph node metastasis (LNM) in T2 colorectal cancer (CRC). SUMMARY BACKGROUND DATA: Recent technical advances allow complete local excision of T2 CRC, traditionally treated with surgical resection. Yet, the widespread adoption of this approach is hampered by the inability to stratify the risk of LNM. METHODS: Data from pT2 CRC patients undergoing surgical resection between April 2000 and May 2022 at one Japanese and one Italian center were analyzed. Primary goal was AI system development for accurate LNM prediction. Predictors encompassed seven variables: age, sex, tumor size and location, lympho-vascular invasion, histological differentiation, and carcinoembryonic antigen level. The tool's discriminating power was assessed via Area Under the Curve (AUC), sensitivity, and specificity. RESULTS: Out of 735 initial patients, 692 were eligible. Training and validation cohorts comprised of 492 and 200 patients, respectively. The AI model displayed an AUC of 0.75 in the combined validation dataset. Sensitivity for LNM prediction was 97.8% and specificity was 15.6%. The Positive and the Negative Predictive Value were 25.7% and 96% respectively. The False Negative (FN) rate was 2.2%, the False Positive was 84.4%. CONCLUSIONS: Our AI model, based on easily accessible clinical and pathological variables, moderately predicts LNM in T2 CRC. However, the risk of FN needs to be considered. The training of the model including more patients across Western and Eastern centers -differentiating between colon and rectal cancers- may improve its performance and accuracy.
ABSTRACT
Pancreatic fluid collections (PFCs), including pancreatic pseudocysts (PPs) and walled-off pancreatic necrosis (WON), are common complications of pancreatitis and pancreatic surgery. Historically, the treatment of these conditions has relied on surgical and radiological approaches. The treatment of patients with PFCs has already focused toward an endoscopy-based approach, and with the development of dedicated lumen-apposing metal stents (LAMS), it has almost totally shifted towards interventional Endoscopic Ultrasound (EUS)-guided procedures. However, there is still limited consensus on several aspects of PFCs treatment within the multidisciplinary management. The interventional endoscopy and ultrasound (i-EUS) group is an Italian network of clinicians and scientists with special interest in biliopancreatic interventional endoscopy, especially interventional EUS. This manuscript focuses on the second part of the results of a consensus conference organized by i-EUS, with the aim of providing evidence-based guidance on several intra- and post-procedural aspects of PFCs drainage, such as clinical management and follow-up.
ABSTRACT
BACKGROUND AND AIM: Guidelines suggest endoscopic resection for rectal neuroendocrine tumors (rNETs) < 10 mm, but the most appropriate resection technique is unclear. In real-life clinical practice, the endoscopic removal of unrecognized rNETs can take place with "simple" techniques and without preliminary staging. The aim of the current study is to report our own experience at a referral center for both neuroendocrine neoplasms and endoscopy. METHODS: Retrospective analyses of polypectomies were performed at the Humanitas Research Hospital for rNETs (already diagnosed or previously unrecognized). RESULTS: A total of 19 patients were included, with a median lesion size of 5 mm (range 3-10 mm). Only five lesions were suspected as NETs before removal and underwent endoscopic ultrasound (EUS) before resection, being removed with advanced endoscopic techniques. Unsuspected rNETs were removed by cold polypectomy in eleven cases, EMR in two, and biopsy forceps in one. When described, the margins were negative in four cases, positive in four (R1), and indeterminate in one. The median follow-up was 40 months. A 10 mm polypoid lesion removed with cold snare polypectomy (G2 R1) needed subsequent surgery. Eighteen patients underwent EUS after a median time of 6.5 months from resection. The EUS identified local recurrence after 14 months in a 7 mm polypoid lesion removed with cold snare polypectomy (G1 R1); the lesion was treated with cap-assisted EMR. For all the other lesions, the follow-up was negative. CONCLUSIONS: When rNETs are improperly removed without prior staging, caution must be exercised. The data from our cohort suggest that even if inappropriate resection had happened, patients may be safely managed with early EUS evaluation.
ABSTRACT
BACKGROUND AND AIMS: The clinical presentation of celiac disease (CD) has changed over time with more patients presenting with non-classical symptoms, extra-intestinal manifestations (EIM) or no symptoms. We aimed to investigate the main symptoms/signs leading to the diagnosis of CD in adult patients. As secondary end-point, we evaluated the outcome of gastrointestinal (GI) symptoms following gluten-free diet (GFD). METHODS: All consecutive CD adult patients referring to our University Hospital from September 2022 to February 2024 were included. Clinical data were retrospectively evaluated. RESULTS: 134 patients, 104 females/30 males, median age at diagnosis 35 years, were included. 79 patients reported GI symptoms (i.e., diarrhea, abdominal bloating, dyspepsia) as the main symptom leading to CD diagnosis. In 40 patients, the leading symptom/sign was an EIM (i.e., iron deficiency anemia, infertility/miscarriages, dermatitis, osteoporosis, elevated transaminase levels). Fifteen patients were asymptomatic, being diagnosed because of a positive family history or concomitant autoimmune hypothyroidism. Of the 79 patients reporting GI symptoms, 20 did not experience complete resolution with the GFD. Among the 17 patients who reported a strict adherence to GFD (vs 1 patient with low-adherence, 2 non-compliant), lactose intolerance and irritable bowel syndrome overlap were diagnosed in 2 and 15 patients, respectively. CONCLUSION: GI manifestations remain the main symptoms at presentation of CD, however clinicians should be aware of the EIM of CD and the association with other autoimmune disorders. In non-responsive CD patients, an overlap with functional disorders might be considered.
ABSTRACT
Pancreatic fluid collections (PFCs), including pancreatic pseudocysts (PPs) and walled-off pancreatic necrosis (WON), are common complications of pancreatitis and pancreatic surgery. Historically, the treatment of these conditions has relied on surgical and radiological approaches; however, it has later shifted toward an endoscopy-based approach. With the development of dedicated lumen-apposing metal stents (LAMS), interventional Endoscopic Ultrasound (EUS)-guided procedures have become the standard approach for PFC drainage. However, there is still limited consensus on several aspects of the multidisciplinary management of PFCs. The interventional endoscopy and ultrasound (i-EUS) group is an Italian network of clinicians and scientists with special interest in biliopancreatic interventional endoscopy, especially interventional EUS. This manuscript describes the first part of the results of a consensus conference organized by i-EUS with the aim of providing evidence-based guidance on aspects such as indications for treating PFCs, the timing of intervention, and different technical strategies for managing patients with PFCs.
ABSTRACT
Introduction The role of endoscopic submucosal dissection (ESD) in the treatment of Barrett's associated neoplasia (BEN) has been evolving. We examined the efficacy and safety of ESD and EMR for BEN. Methods A database search was performed for studies reporting efficacy and safety outcomes of ESD and EMR for BEN. Pooled proportional and comparative meta-analyses were performed. Results 47 studies (23 ESD, 19 EMR, and 5 comparative) were included. Mean lesion size for ESD and EMR were 22.5 mm and 15.8 mm respectively. Majority of lesions were Paris type IIa. Pooled analysis for ESD showed en-bloc resection, R0 resection, curative resection, and local recurrence rates of 98%, 78%, 65%, and 2%, respectively. Complete eradication of dysplasia (CE-D) and complete eradication of intestinal metaplasia (CE-IM) were achieved in 94% and 59% of cases. Pooled rates of perforation, intraprocedural bleeding (IPB), delayed bleeding (DB), and stricture were 1%, 1%, 2%, and 10%, respectively. Pooled analysis for EMR showed en-bloc resection, R0 resection, curative resection, and local recurrence rates of 37%, 67%, 62%, and 6%, respectively. CE-D and CE-IM were achieved in 94% and 76% of cases. Pooled rates of perforation, IPB, DB, and stricture were 0.1%, 1%, 0.4%, and 7.7%, respectively. The mean procedure time for ESD and EMR were 111.3 and 22.3 mins respectively. Comparative analysis showed higher en-bloc and R0 resection rates with ESD compared to EMR, with comparable adverse events. Conclusion ESD and EMR both can be employed to treat BEN depending on the lesion type, size, and expertise.
ABSTRACT
BACKGROUND: Artificial intelligence (AI)-based chatbots have shown promise in providing counseling to patients with metabolic dysfunction-associated steatotic liver disease (MASLD). While ChatGPT3.5 has demonstrated the ability to comprehensively answer MASLD-related questions in English, its accuracy remains suboptimal. Whether language influences these results is unclear. This study aims to assess ChatGPT's performance as a counseling tool for Italian MASLD patients. METHODS: Thirteen Italian experts rated the accuracy, completeness and comprehensibility of ChatGPT3.5 in answering 15 MASLD-related questions in Italian using a six-point accuracy, three-point completeness and three-point comprehensibility Likert's scale. RESULTS: Mean scores for accuracy, completeness and comprehensibility were 4.57 ± 0.42, 2.14 ± 0.31 and 2.91 ± 0.07, respectively. The physical activity domain achieved the highest mean scores for accuracy and completeness, whereas the specialist referral domain achieved the lowest. Overall, Fleiss's coefficient of concordance for accuracy, completeness and comprehensibility across all 15 questions was 0.016, 0.075 and -0.010, respectively. Age and academic role of the evaluators did not influence the scores. The results were not significantly different from our previous study focusing on English. CONCLUSION: Language does not appear to affect ChatGPT's ability to provide comprehensible and complete counseling to MASLD patients, but accuracy remains suboptimal in certain domains.
ABSTRACT
BACKGROUND AND AIM: Randomised trials show improved polyp detection with computer-aided detection (CADe), mostly of small lesions. However, operator and selection bias may affect CADe's true benefit. Clinical outcomes of increased detection have not yet been fully elucidated. METHODS: In this multicentre trial, CADe combining convolutional and recurrent neural networks was used for polyp detection. Blinded endoscopists were monitored in real time by a second observer with CADe access. CADe detections prompted reinspection. Adenoma detection rates (ADR) and polyp detection rates were measured prestudy and poststudy. Histological assessments were done by independent histopathologists. The primary outcome compared polyp detection between endoscopists and CADe. RESULTS: In 946 patients (51.9% male, mean age 64), a total of 2141 polyps were identified, including 989 adenomas. CADe was not superior to human polyp detection (sensitivity 94.6% vs 96.0%) but outperformed them when restricted to adenomas. Unblinding led to an additional yield of 86 true positive polyp detections (1.1% ADR increase per patient; 73.8% were <5 mm). CADe also increased non-neoplastic polyp detection by an absolute value of 4.9% of the cases (1.8% increase of entire polyp load). Procedure time increased with 6.6±6.5 min (+42.6%). In 22/946 patients, the additional detection of adenomas changed surveillance intervals (2.3%), mostly by increasing the number of small adenomas beyond the cut-off. CONCLUSION: Even if CADe appears to be slightly more sensitive than human endoscopists, the additional gain in ADR was minimal and follow-up intervals rarely changed. Additional inspection of non-neoplastic lesions was increased, adding to the inspection and/or polypectomy workload.
ABSTRACT
INTRODUCTION: Endoscopic submucosal dissection (ESD) is a widely used technique to remove early neoplastic lesions. It was primarily used in the initial days to treat gastric lesions, but recently, the horizon of this endoscopic procedure has expanded, which has allowed us to manage other technically more complex locations, such as the colorectum. AREAS COVERED: There has been an exponential growth regarding the wide range of devices available in the market for performing colorectal ESD. As a result, the aim of this review is to highlight the indication of this endoscopic technique, which device is best suited for which indication, as well as future trajectories in this field. EXPERT OPINION: Although some devices have proven to be more advantageous than others in this area, very often the choice is still subjective, which is commonly attributed to individual preferences and experience. However, an accurate knowledge of the available tools and their functioning, with their pros and cons, is fundamental for any endoscopist venturing into the field of third space endoscopy. In this way, one can choose which device best suits a particular situation, along with simultaneously having the wealth of knowledge related to therapeutic armamentarium at our disposal in the endoscopy suite.
ABSTRACT
BACKGROUND AND AIMS: Serrated polyps (SPs) are precursors to 15% to 20% of colorectal cancers (CRCs). However, there are uncertainties regarding which SPs require surveillance and at what intervals, with recommendations adapted from those for adenomas in the absence of solid evidence. Our aim was to assess which SP risk characteristics relate to a higher risk of metachronous CRC or advanced polyps. METHODS: We systematically searched PubMed, Embase, and Cochrane for cohort studies, case-control studies, and clinical trials from inception to December 31, 2023, of CRC or advanced polyps (advanced adenoma [AA] or advanced SP) incidence at surveillance stratified by baseline SP size, dysplasia, location, and multiplicity. We defined advanced SPs as those ≥10 mm or with dysplasia. CRC and advanced polyp incidence per 1000 person-years were estimated. We performed a meta-analysis by calculating pooled relative risks (RRs) using a random-effects model. RESULTS: A total of 5903 studies were reviewed, and 14 were included with 493,949 patients (mean age, 59.5 years; 55% men). The mean follow-up was 4.9 years. CRC incidence per 1000 person-years was 2.09 (95% confidence interval [CI], 1.29-2.90) for advanced SPs, 1.52 (95% CI, 0.78-2.25) for SPs of ≥10 mm, 5.86 (95% CI, 2.16-9.56) for SPs with dysplasia, 1.18 (95% CI, 0.77-1.60) for proximal SPs, 0.52 (95% CI, 0.08-1.12) for ≥3 SPs, 0.50 (95% CI, 0.35-0.66) for nonadvanced SPs, and 0.44 (95% CI, 0.41-0.46) for normal colonoscopy findings. Metachronous CRC risk was higher in advanced SPs versus nonadvanced SPs (RR, 1.84; 95% CI, 1.11-3.04) and versus normal colonoscopy findings (RR, 2.92; 95% CI, 2.26-3.77), in SPs of ≥10 mm versus <10 mm (RR, 2.61; 95% CI, 1.43-4.77) and versus normal colonoscopy findings (RR: 3.52; 95% CI, 2.17-5.69); and in SPs with dysplasia versus normal colonoscopy findings (RR: 2.71; 95% CI, 2.00-3.67). No increase in CRC or advanced polyp risk was found in patients with proximal versus distal SPs, nor in ≥3 SPs versus 1 or 2 SPs. CONCLUSIONS: CRC risk is significantly higher in patients with baseline advanced SPs after 4.9 years of follow-up, with risk magnitudes similar to those described for AA, supporting the current recommendation for 3-year surveillance in patients with advanced SPs.
ABSTRACT
BACKGROUND: The efficacy of endoscopic ultrasound-guided liver biopsy (EUS-LB) compared to percutaneous liver biopsy (PC-LB) remains uncertain. METHODS: Our data consist of randomized controlled trials (RCTs) comparing EUS-LB to PC-LB, found through a literature search via PubMed/Medline and Embase. The primary outcome was sample adequacy, whereas secondary outcomes were longest and total lengths of tissue specimens, diagnostic accuracy, and number of complete portal tracts (CPTs). RESULTS: Sample adequacy did not significantly differ between EUS-LB and PC-LB (risk ratio [RR] 1.18; 95% confidence interval [CI] 0.58-2.38; p = 0.65), with very low evidence quality and inadequate sample size as per trial sequential analysis (TSA). The two techniques were equivalent with respect to diagnostic accuracy (RR: 1; CI: 0.95-1.05; p = 0.88), mean number of complete portal tracts (mean difference: 2.29, -4.08 to 8.66; p = 0.48), and total specimen length (mean difference: -0.51, -20.92 to 19.9; p = 0.96). The mean maximum specimen length was significantly longer in the PC-LB group (mean difference: -3.11, -5.51 to -0.71; p = 0.01), and TSA showed that the required information size was reached. CONCLUSION: EUS-LB and PC-LB are comparable in terms of diagnostic performance although PC-LB provides longer non-fragmented specimens.
ABSTRACT
Detection and diagnosis of colon polyps are key to preventing colorectal cancer. Recent evidence suggests that AI-based computer-aided detection (CADe) and computer-aided diagnosis (CADx) systems can enhance endoscopists' performance and boost colonoscopy effectiveness. However, most available public datasets primarily consist of still images or video clips, often at a down-sampled resolution, and do not accurately represent real-world colonoscopy procedures. We introduce the REAL-Colon (Real-world multi-center Endoscopy Annotated video Library) dataset: a compilation of 2.7 M native video frames from sixty full-resolution, real-world colonoscopy recordings across multiple centers. The dataset contains 350k bounding-box annotations, each created under the supervision of expert gastroenterologists. Comprehensive patient clinical data, colonoscopy acquisition information, and polyp histopathological information are also included in each video. With its unprecedented size, quality, and heterogeneity, the REAL-Colon dataset is a unique resource for researchers and developers aiming to advance AI research in colonoscopy. Its openness and transparency facilitate rigorous and reproducible research, fostering the development and benchmarking of more accurate and reliable colonoscopy-related algorithms and models.
Subject(s)
Colonic Polyps , Colonoscopy , Colonoscopy/methods , Humans , Colonic Polyps/diagnosis , Diagnosis, Computer-Assisted , Artificial Intelligence , Video Recording , Colorectal Neoplasms/diagnosisABSTRACT
BACKGROUND: Computer-aided diagnosis (CADx) allows prediction of polyp histology during colonoscopy, which may reduce unnecessary removal of nonneoplastic polyps. However, the potential benefits and harms of CADx are still unclear. PURPOSE: To quantify the benefit and harm of using CADx in colonoscopy for the optical diagnosis of small (≤5-mm) rectosigmoid polyps. DATA SOURCES: Medline, Embase, and Scopus were searched for articles published before 22 December 2023. STUDY SELECTION: Histologically verified diagnostic accuracy studies that evaluated the real-time performance of physicians in predicting neoplastic change of small rectosigmoid polyps without or with CADx assistance during colonoscopy. DATA EXTRACTION: The clinical benefit and harm were estimated on the basis of accuracy values of the endoscopist before and after CADx assistance. The certainty of evidence was assessed using the GRADE (Grading of Recommendations Assessment, Development and Evaluation) framework. The outcome measure for benefit was the proportion of polyps predicted to be nonneoplastic that would avoid removal with the use of CADx. The outcome measure for harm was the proportion of neoplastic polyps that would be not resected and left in situ due to an incorrect diagnosis with the use of CADx. Histology served as the reference standard for both outcomes. DATA SYNTHESIS: Ten studies, including 3620 patients with 4103 small rectosigmoid polyps, were analyzed. The studies that assessed the performance of CADx alone (9 studies; 3237 polyps) showed a sensitivity of 87.3% (95% CI, 79.2% to 92.5%) and specificity of 88.9% (CI, 81.7% to 93.5%) in predicting neoplastic change. In the studies that compared histology prediction performance before versus after CADx assistance (4 studies; 2503 polyps), there was no difference in the proportion of polyps predicted to be nonneoplastic that would avoid removal (55.4% vs. 58.4%; risk ratio [RR], 1.06 [CI, 0.96 to 1.17]; moderate-certainty evidence) or in the proportion of neoplastic polyps that would be erroneously left in situ (8.2% vs. 7.5%; RR, 0.95 [CI, 0.69 to 1.33]; moderate-certainty evidence). LIMITATION: The application of optical diagnosis was only simulated, potentially altering the decision-making process of the operator. CONCLUSION: Computer-aided diagnosis provided no incremental benefit or harm in the management of small rectosigmoid polyps during colonoscopy. PRIMARY FUNDING SOURCE: European Commission. (PROSPERO: CRD42023402197).
Subject(s)
Colonic Polyps , Colonoscopy , Diagnosis, Computer-Assisted , Humans , Colonic Polyps/pathology , Colonic Polyps/diagnostic imaging , Colorectal Neoplasms/pathology , Colorectal Neoplasms/diagnosisABSTRACT
BACKGROUND: Real-time prediction of histologic features of small colorectal polyps may prevent resection and/or pathologic evaluation and therefore decrease colonoscopy costs. Previous studies showed that computer-aided diagnosis (CADx) was highly accurate, though it did not outperform expert endoscopists. OBJECTIVE: To assess the diagnostic performance of histologic predictions by general endoscopists before and after assistance from CADx in a real-life setting. DESIGN: Prospective, multicenter, single-group study. (ClinicalTrials.gov: NCT04437615). SETTING: 6 centers across the United States. PARTICIPANTS: 1252 consecutive patients undergoing colonoscopy and 49 general endoscopists with variable experience in real-time prediction of polyp histologic features. INTERVENTION: Real-time use of CADx during routine colonoscopy. MEASUREMENTS: The primary end points were the sensitivity and specificity of CADx-unassisted and CADx-assisted histologic predictions for adenomas measuring 5 mm or less. For clinical purposes, additional estimates according to location and confidence level were provided. RESULTS: The CADx device made a diagnosis for 2695 polyps measuring 5 mm or less (96%) in 1252 patients. There was no difference in sensitivity between the unassisted and assisted groups (90.7% vs. 90.8%; P = 0.52). Specificity was higher in the CADx-assisted group (59.5% vs. 64.7%; P < 0.001). Among all 2695 polyps measuring 5 mm or less, 88.2% and 86.1% (P < 0.001) in the CADx-assisted and unassisted groups, respectively, could be resected and discarded without pathologic evaluation. Among 743 rectosigmoid polyps measuring 5 mm or less, 49.5% and 47.9% (P < 0.001) in the CADx-assisted and unassisted groups, respectively, could be left in situ without resection. LIMITATION: Decision making based on CADx might differ outside a clinical trial. CONCLUSION: CADx assistance did not result in increased sensitivity of optical diagnosis. Despite a slight increase, the specificity of CADx-assisted diagnosis remained suboptimal. PRIMARY FUNDING SOURCE: Olympus America Corporation served as the clinical study sponsor.
Subject(s)
Artificial Intelligence , Colonic Polyps , Colonoscopy , Diagnosis, Computer-Assisted , Sensitivity and Specificity , Humans , Colonic Polyps/pathology , Prospective Studies , Female , Male , Middle Aged , Aged , Adenoma/pathology , Adenoma/diagnosis , Colorectal Neoplasms/pathology , Clinical Competence , AdultABSTRACT
Background and study aims Artificial Intelligence (AI) systems could make the optical diagnosis (OD) of diminutive colorectal polyps (DCPs) more reliable and objective. This study was aimed at prospectively evaluating feasibility and diagnostic performance of AI-standalone and AI-assisted OD of DCPs in a real-life setting by using a white light-based system (GI Genius, Medtronic Co, Minneapolis, Minnesota, United States). Patients and methods Consecutive colonoscopy outpatients with at least one DCP were evaluated by 11 endoscopists (5 experts and 6 non-experts in OD). DCPs were classified in real time by AI (AI-standalone OD) and by the endoscopist with the assistance of AI (AI-assisted OD), with histopathology as the reference standard. Results Of the 480 DCPs, AI provided the outcome "adenoma" or "non-adenoma" in 81.4% (95% confidence interval [CI]: 77.5-84.6). Sensitivity, specificity, positive and negative predictive value, and accuracy of AI-standalone OD were 97.0% (95% CI 94.0-98.6), 38.1% (95% CI 28.9-48.1), 80.1% (95% CI 75.2-84.2), 83.3% (95% CI 69.2-92.0), and 80.5% (95% CI 68.7-82.8%), respectively. Compared with AI-standalone, the specificity of AI-assisted OD was significantly higher (58.9%, 95% CI 49.7-67.5) and a trend toward an increase was observed for other diagnostic performance measures. Overall accuracy and negative predictive value of AI-assisted OD for experts and non-experts were 85.8% (95% CI 80.0-90.4) vs. 80.1% (95% CI 73.6-85.6) and 89.1% (95% CI 75.6-95.9) vs. 80.0% (95% CI 63.9-90.4), respectively. Conclusions Standalone AI is able to provide an OD of adenoma/non-adenoma in more than 80% of DCPs, with a high sensitivity but low specificity. The human-machine interaction improved diagnostic performance, especially when experts were involved.
ABSTRACT
Despite the evolution in tools and techniques, perforation is still one of the most pernicious adverse events of therapeutic endoscopy with potentially huge consequences. As advanced endoscopic resection techniques are worldwide spreading, endoscopists must be ready to manage intraprocedural perforations. In fact, immediate endoscopic closure through a prompt diagnosis represents the first-line option, saving patients from surgery, long hospitalizations and worse outcomes. Traditional and novel endoscopic closure modalities, including clips, suturing devices, stents and vacuum therapy, are increasingly expanding the therapeutic armamentarium for closing these defects. Nevertheless, available literature on this topic is currently limited. In this review our goal is to give an overview on the management of perforations occurring during endoscopic resections, with particular attention to characteristics, advantages, disadvantages and new horizons of endoscopic closure tools.