Layer-selective deep representation to improve esophageal cancer classification.

Even though artificial intelligence and machine learning have demonstrated remarkable performances in medical image computing, their accountability and transparency level must be improved to transfer this success into clinical practice. The reliability of machine learning decisions must be explained and interpreted, especially for supporting the medical diagnosis. For this task, the deep learning techniques' black-box nature must somehow be lightened up to clarify its promising results. Hence, we aim to investigate the impact of the ResNet-50 deep convolutional design for Barrett's esophagus and adenocarcinoma classification. For such a task, and aiming at proposing a two-step learning technique, the output of each convolutional layer that composes the ResNet-50 architecture was trained and classified for further definition of layers that would provide more impact in the architecture. We showed that local information and high-dimensional features are essential to improve the classification for our task. Besides, we observed a significant improvement when the most discriminative layers expressed more impact in the training and classification of ResNet-50 for Barrett's esophagus and adenocarcinoma classification, demonstrating that both human knowledge and computational processing may influence the correct learning of such a problem.

Expert assessment of infiltration depth and recommendation of endoscopic resection technique in early Barrett cancer.

BACKGROUND: Early Barrett cancer can be curatively treated by endoscopic resection. The choice of the resection technique, however-endoscopic mucosal resection (EMR) or submucosal dissection (ESD)-largely depends on the assumed infiltration depth as judged by the endoscopist. However, the accuracy of endoscopic diagnosis of the degree of cancer infiltration is not known. METHODS: Three to four high-quality images (both in overview and close-up) from 202 of early Barrett esophagus cancer cases (82% men, mean age 66.9 years) were selected from our endoscopy database (73.3% stage T1a and 26.7% in stage T1b). Images were shown to 9 Barrett esophagus experts, with patients' clinical data (age, sex, Barrett esophagus length) and biopsy results. The experts were asked to predict infiltration depth (T1b vs. T1a), and to suggest the appropriate endoscopic resection technique (EMR or ESD, or surgery). Interobserver variability (kappa values) was also determined for these parameters. RESULTS: Overall positive (PPV) and negative predictive values (NPV) to diagnose T1b versus T1a infiltration were 40.7% (95% CI: 36.7, 44.8) and 79.8% (95% CI: 77.5, 81.9), respectively; kappa value was 0.41. Paris classification (kappa 0.51) and suggested treatment also varied between experts. In a post hoc analysis, only the correlation between lesions classified as invisible or flat according to the Paris classification (IIB; 25% of all cases) and the suggested resection technique was better: In this subgroup, EMR was recommended in >80% of cases, with a high complete (basal R0) resection rate (mean of 88.1%). CONCLUSIONS: Precise endoscopic distinction between mucosal and submucosal involvement of Barrett esophagus cancer by experts as a basis for choosing the resection technique has limited predictive values and high interobserver variability. It seems that mainly invisible/flat lesions may result in good resection outcomes when treated by EMR, but this stratification strategy has to be assessed in further studies.

Automatic three-dimensional reconstruction of the oesophagus in achalasia patients undergoing POEM: an innovative approach for evaluating treatment outcomes.

BACKGROUND AND AIMS: Peroral endoscopic myotomy (POEM) is a standard treatment option for achalasia patients. Treatment response varies due to factors such as achalasia type, degree of dilatation, pressure and distensibility indices. We present an innovative approach for treatment response prediction based on an automatic three-dimensional (3-D) reconstruction of the tubular oesophagus (TE) and the lower oesophageal sphincter (LES) in patients undergoing POEM for achalasia. METHODS: A software was developed, integrating data from high-resolution manometry, timed barium oesophagogram and endoscopic images to automatically generate 3-D reconstructions of the TE and LES. Novel normative indices for TE (volume×pressure) and LES (volume/pressure) were automatically integrated, facilitating pre-POEM and post-POEM comparisons. Treatment response was evaluated by changes in volumetric and pressure indices for the TE and the LES before as well as 3 and 12 months after POEM. In addition, these values were compared with normal value indices of non-achalasia patients. RESULTS: 50 treatment-naive achalasia patients were enrolled prospectively. The mean TE index decreased significantly (p<0.0001) and the mean LES index increased significantly 3 months post-POEM (p<0.0001). In the 12-month follow-up, no further significant change of value indices between 3 and 12 months post-POEM was seen. 3 months post-POEM mean LES index approached the mean LES of the healthy control group (p=0.077). CONCLUSION: 3-D reconstruction provides an interactive, dynamic visualisation of the oesophagus, serving as a comprehensive tool for evaluating treatment response. It may contribute to refining our approach to achalasia treatment and optimising treatment outcomes. TRIAL REGISTRATION NUMBER: 22-0149.

Endoscopic Assessment of Local Resectability of Colorectal Malignancies.

Background: The endoscopic assessment of colorectal malignancies primarily aims at deciding on the local resectability. Local resectability is defined by the risk of lymphonodal metastasis. Summary: The gross morphology as well as the surface and vessel patterns provide valuable information prior to endoscopic resection. Various classifications have been standardized to assist endoscopists during endoscopic assessment. Key Message: The macroscopic assessment of colorectal malignancies should include the Paris and laterally spreading tumor (LST) classification as well as chromoendoscopic assessments such as Kudo's pit pattern and the Japanese NBI Expert Team classifications to describe the vessel and surface patterns.

Cold Versus Hot Snare Endoscopic Resection of Large Nonpedunculated Colorectal Polyps: Randomized Controlled German CHRONICLE Trial.

BACKGROUND & AIMS: Endoscopic mucosal resection (EMR) is standard therapy for nonpedunculated colorectal polyps ≥20 mm. It has been suggested recently that polyp resection without current (cold resection) may be superior to the standard technique using cutting/coagulation current (hot resection) by reducing adverse events (AEs), but evidence from a randomized trial is missing. METHODS: In this randomized controlled multicentric trial involving 19 centers, nonpedunculated colorectal polyps ≥20 mm were randomly assigned to cold or hot EMR. The primary outcome was major AE (eg, perforation or postendoscopic bleeding). Among secondary outcomes, major AE subcategories, postpolypectomy syndrome, and residual adenoma were most relevant. RESULTS: Between 2021 and 2023, there were 396 polyps in 363 patients (48.2% were female) enrolled for the intention-to-treat analysis. Major AEs occurred in 1.0% of the cold group and in 7.9% of the hot group (P = .001; odds ratio [OR], 0.12; 95% CI, 0.03-0.54). Rates for perforation and postendoscopic bleeding were significantly lower in the cold group, with 0% vs 3.9% (P = .007) and 1.0% vs 4.4% (P = .040). Postpolypectomy syndrome occurred with similar frequency (3.1% vs 4.4%; P = .490). After cold resection, residual adenoma was found more frequently, with 23.7% vs 13.8% (P = .020; OR, 1.94; 95% CI, 1.12-3.38). In multivariable analysis, lesion diameter of ≥4 cm was an independent predictor both for major AEs (OR, 3.37) and residual adenoma (OR, 2.47) and high-grade dysplasia/cancer for residual adenoma (OR, 2.92). CONCLUSIONS: Cold resection of large, nonpedunculated colorectal polyps appears to be considerably safer than hot EMR; however, at the cost of a higher residual adenoma rate. Further studies have to confirm to what extent polyp size and histology can determine an individualized approach. German Clinical Trials Registry (Deutsches Register Klinischer Studien), Number DRKS00025170.

Endoscopic Submucosal Dissection for Early Esophageal Adenocarcinoma: Low Rates of Metastases in Mucosal Cancers with Poor Differentiation.

BACKGROUND AND AIMS: Endoscopic resection (ER) is accepted as standard treatment for intramucosal esophageal adenocarcinoma (EAC) with well or moderate differentiation. Poor differentiation (PD) is judged as a risk factor for lymph node metastasis (LNM) and surgery is recommended. However, the evidence for this recommendation is weak. Study aim was to analyze the clinical course of patients after ER of EAC with PD. PATIENTS AND METHODS: Patients undergoing endoscopic submucosal dissection for EAC were included from 16 German centers. Inclusion criteria were PD in the resection specimen, R0 resection and endoscopic follow-up. Primary outcome was the metastasis rate during follow-up. Analysis was performed retrospectively in a prospectively collected database. RESULTS: 25 patients with PD as single risk factor (group A) and 15 patients with PD and additional risk factors (submucosal invasion and/or lymphovascular invasion) were included. The metastasis rate was was 1/25 (4.0%; 95%CI 0.4-17.2) in group A and 3/15 (20.0%; 95%CI 6.0-44.4%) in group B, respectively (p=0.293). The rate of EAC-associated deaths was 1/25 (4%; 95%CI 0.4-17.2%) versus 3/15 (20%; 95%CI 6.0-44.4%) in group B (p=0.293) while the overall death rate was 7/25 (28.0%; 95%CI 13.5-47.3%) versus 3/15 (20%; 95%CI 6.0-44.4%) (p=0.715). Median follow-up was 30 months (IQR 15-53). CONCLUSIONS: During long-term follow-up the risk of metastasis is low after ER of mucosal EAC with PD as single risk factor. A conservative approach seems justified in this small patient group. However, the treatment strategy has to be determined on an individualized basis until further prospective data are available.

Influence of artificial intelligence on the diagnostic performance of endoscopists in the assessment of Barrett's esophagus: a tandem randomized and video trial.

BACKGROUND: This study evaluated the effect of an artificial intelligence (AI)-based clinical decision support system on the performance and diagnostic confidence of endoscopists in their assessment of Barrett's esophagus (BE). METHODS: 96 standardized endoscopy videos were assessed by 22 endoscopists with varying degrees of BE experience from 12 centers. Assessment was randomized into two video sets: group A (review first without AI and second with AI) and group B (review first with AI and second without AI). Endoscopists were required to evaluate each video for the presence of Barrett's esophagus-related neoplasia (BERN) and then decide on a spot for a targeted biopsy. After the second assessment, they were allowed to change their clinical decision and confidence level. RESULTS: AI had a stand-alone sensitivity, specificity, and accuracy of 92.2%, 68.9%, and 81.3%, respectively. Without AI, BE experts had an overall sensitivity, specificity, and accuracy of 83.3%, 58.1%, and 71.5%, respectively. With AI, BE nonexperts showed a significant improvement in sensitivity and specificity when videos were assessed a second time with AI (sensitivity 69.8% [95%CI 65.2%-74.2%] to 78.0% [95%CI 74.0%-82.0%]; specificity 67.3% [95%CI 62.5%-72.2%] to 72.7% [95%CI 68.2%-77.3%]). In addition, the diagnostic confidence of BE nonexperts improved significantly with AI. CONCLUSION: BE nonexperts benefitted significantly from additional AI. BE experts and nonexperts remained significantly below the stand-alone performance of AI, suggesting that there may be other factors influencing endoscopists' decisions to follow or discard AI advice.

Western outcomes of circumferential endoscopic submucosal dissection for early esophageal squamous cell carcinoma.

BACKGROUND AND AIMS: Circumferential endoscopic submucosal dissection (cESD) in the esophagus has been reported to be feasible in small Eastern case series. We assessed the outcomes of cESD in the treatment of early esophageal squamous cell carcinoma (ESCC) in Western countries. METHODS: We conducted an international study at 25 referral centers in Europe and Australia using prospective databases. We included all patients with ESCC treated with cESD before November 2022. Our main outcomes were curative resection according to European guidelines and adverse events. RESULTS: A total of 171 cESDs were performed on 165 patients. En bloc and R0 resections rates were 98.2% (95% confidence interval [CI], 95.0-99.4) and 69.6% (95% CI, 62.3-76.0), respectively. Curative resection was achieved in 49.1% (95% CI, 41.7-56.6) of the lesions. The most common reason for noncurative resection was deep submucosal invasion (21.6%). The risk of stricture requiring 6 or more dilations or additional techniques (incisional therapy/stent) was high (71%), despite the use of prophylactic measures in 93% of the procedures. The rates of intraprocedural perforation, delayed bleeding, and adverse cardiorespiratory events were 4.1%, 0.6%, and 4.7%, respectively. Two patients died (1.2%) of a cESD-related adverse event. Overall and disease-free survival rates at 2 years were 91% and 79%. CONCLUSIONS: In Western referral centers, cESD for ESCC is curative in approximately half of the lesions. It can be considered a feasible treatment in selected patients. Our results suggest the need to improve patient selection and to develop more effective therapies to prevent esophageal strictures.

Texture and color enhancement imaging versus high definition white-light endoscopy for detection of colorectal neoplasia: a randomized trial.

BACKGROUND: Texture and color enhancement imaging (TXI) was recently proposed as a substitute for standard high definition white-light imaging (WLI) to increase lesion detection during colonoscopy. This international, multicenter randomized trial assessed the efficacy of TXI in detection of colorectal neoplasia. METHODS: Consecutive patients aged ≥ 40 years undergoing screening, surveillance, or diagnostic colonoscopies at five centers (Italy, Germany, Japan) between September 2021 and May 2022 were enrolled. Patients were randomly assigned (1:1) to TXI or WLI. Primary outcome was adenoma detection rate (ADR). Secondary outcomes were adenomas per colonoscopy (APC) and withdrawal time. Relative risks (RRs) adjusted for age, sex, and colonoscopy indication were calculated. RESULTS: We enrolled 747 patients (mean age 62.3 [SD 9.5] years, 50.2 % male). ADR was significantly higher with TXI (221/375, 58.9 %) vs. WLI (159/372, 42.7 %; adjusted RR 1.38 [95 %CI 1.20-1.59]). This was significant for ≤ 5 mm (RR 1.42 [1.16-1.73]) and 6-9 mm (RR 1.36 [1.01-1.83]) adenomas. A higher proportion of polypoid (151/375 [40.3 %] vs. 104/372 [28.0 %]; RR 1.43 [1.17-1.75]) and nonpolypoid (136/375 [36.3 %] vs. 102/372 [27.4 %]; RR 1.30 [1.05-1.61]) adenomas, and proximal (143/375 [38.1 %] vs. 111/372 [29.8 %]; RR 1.28 [1.05-1.57]) and distal (144/375 [38.4 %] vs. 98/372 [26.3 %]; RR 1.46 [1.18-1.80]) lesions were found with TXI. APC was higher with TXI (1.36 [SD 1.79] vs. 0.89 [SD 1.35]; incident rate ratio 1.53 [1.25-1.88]). CONCLUSIONS: TXI increased ADR and APC among patients undergoing colonoscopy for various indications. TXI increased detection of polyps < 10 mm, both in the proximal and distal colon, and may help to improve colonoscopy quality indicators.

Diagnostic Value of Adding Magnifying Chromoendoscopy to Magnifying Narrow-Band Imaging Endoscopy for Colorectal Polyps.

BACKGROUND & AIMS: This study examined the additional value of magnifying chromoendoscopy (MCE) on magnifying narrow-band imaging endoscopy (M-NBI) in the optical diagnosis of colorectal polyps. METHODS: A multicenter prospective study was conducted at 9 facilities in Japan and Germany. Patients with colorectal polyps scheduled for resection were included. Optical diagnosis was performed by M-NBI first, followed by MCE. Both diagnoses were made in real time. MCE was performed on all type 2B lesions classified according to the Japan NBI Expert Team classification and other lesions at the discretion of endoscopists. The diagnostic accuracy and confidence of M-NBI and MCE for colorectal cancer (CRC) with deep invasion (≥T1b) were compared on the basis of histologic findings after resection. RESULTS: In total, 1173 lesions were included between February 2018 and December 2020, with 654 (5 hyperplastic polyp/sessile serrated lesion, 162 low-grade dysplasia, 403 high-grade dysplasia, 97 T1 CRCs, and 32 ≥T2 CRCs) examined using MCE after M-NBI. In the diagnostic accuracy for predicting CRC with deep invasion, sensitivity, specificity, positive predictive value, negative predictive value, and accuracy for M-NBI were 63.1%, 94.2%, 61.6%, 94.5%, and 90.2%, respectively, and for MCE they were 77.4%, 93.2%, 62.5%, 96.5%, and 91.1%, respectively. The sensitivity was significantly higher in MCE (P < .001). However, these additional values were limited to lesions with low confidence in M-NBI or the ones diagnosed as ≥T1b CRC by M-NBI. CONCLUSIONS: In this multicenter prospective study, we demonstrated the additional value of MCE on M-NBI. We suggest that additional MCE be recommended for lesions with low confidence or the ones diagnosed as ≥T1b CRC. Trials registry number: UMIN000031129.

Detection of duodenal villous atrophy on endoscopic images using a deep learning algorithm.

BACKGROUND AND AIMS: Celiac disease with its endoscopic manifestation of villous atrophy (VA) is underdiagnosed worldwide. The application of artificial intelligence (AI) for the macroscopic detection of VA at routine EGD may improve diagnostic performance. METHODS: A dataset of 858 endoscopic images of 182 patients with VA and 846 images from 323 patients with normal duodenal mucosa was collected and used to train a ResNet18 deep learning model to detect VA. An external dataset was used to test the algorithm, in addition to 6 fellows and 4 board-certified gastroenterologists. Fellows could consult the AI algorithm's result during the test. From their consultation distribution, a stratification of test images into "easy" and "difficult" was performed and used for classified performance measurement. RESULTS: External validation of the AI algorithm yielded values of 90%, 76%, and 84% for sensitivity, specificity, and accuracy, respectively. Fellows scored corresponding values of 63%, 72%, and 67% and experts scored 72%, 69%, and 71%, respectively. AI consultation significantly improved all trainee performance statistics. Although fellows and experts showed significantly lower performance for difficult images, the performance of the AI algorithm was stable. CONCLUSIONS: In this study, an AI algorithm outperformed endoscopy fellows and experts in the detection of VA on endoscopic still images. AI decision support significantly improved the performance of nonexpert endoscopists. The stable performance on difficult images suggests a further positive add-on effect in challenging cases.

Role of artificial intelligence in diagnosing Barrett's esophagus-related neoplasia.

Barrett's esophagus is associated with an increased risk of adenocarcinoma. Thorough screening during endoscopic surveillance is crucial to improve patient prognosis. Detecting and characterizing dysplastic or neoplastic Barrett's esophagus during routine endoscopy are challenging, even for expert endoscopists. Artificial intelligence-based clinical decision support systems have been developed to provide additional assistance to physicians performing diagnostic and therapeutic gastrointestinal endoscopy. In this article, we review the current role of artificial intelligence in the management of Barrett's esophagus and elaborate on potential artificial intelligence in the future.

Endoscopic submucosal dissection for superficial esophageal squamous cell carcinoma: long-term follow-up in a Western center.

BACKGROUND/AIMS: Endoscopic submucosal dissection (ESD) has been established as a treatment modality for superficial esophageal squamous cell carcinoma (ESCC). Long-term follow-up data are lacking in Western countries. The aim of this study was to analyze long-term survival in a Western center. METHODS: Patients undergoing ESD for ESCC were included. The analysis was performed retrospectively using a prospectively collected database. RESULTS: R0 resection rate was 96.7% (59/61 lesions in 58 patients). Twenty-seven patients (46.6%) fulfilled the curative resection criteria (M1/M2) (group A), 11 patients (19.0%) had M3 lesions without lymphovascular invasion (LVI) (group B), and 20 patients (34.5%) had lesions with submucosal invasion or LVI (group C). Additional treatment was recommended after non-curative resection. It was not performed in 20/31 patients (64.5%), mainly because of comorbidities (75%). Twenty-nine out of 58 (50.0%) patients died during a mean follow-up of 3.7 years. Death was related to ESCC in 17.2% (5/29) of patients. The disease-specific survival rate after curative resection was 100%. Overall survival rates after 5 years were 61.5%, 63.6% and 28.1% for groups A, B, and C, respectively. The overall survival was significantly worse after non-curative resection (p=0.038). CONCLUSION: Non-curative resection is frequent after ESD for ESCC in Western patients. The long-term prognosis is limited and mainly determined by comorbidity. Early diagnosis and pre-interventional assessments need to be improved.

GWAS meta-analysis of 16 790 patients with Barrett's oesophagus and oesophageal adenocarcinoma identifies 16 novel genetic risk loci and provides insights into disease aetiology beyond the single marker level.

OBJECTIVE: Oesophageal cancer (EC) is the sixth leading cause of cancer-related deaths. Oesophageal adenocarcinoma (EA), with Barrett's oesophagus (BE) as a precursor lesion, is the most prevalent EC subtype in the Western world. This study aims to contribute to better understand the genetic causes of BE/EA by leveraging genome wide association studies (GWAS), genetic correlation analyses and polygenic risk modelling. DESIGN: We combined data from previous GWAS with new cohorts, increasing the sample size to 16 790 BE/EA cases and 32 476 controls. We also carried out a transcriptome wide association study (TWAS) using expression data from disease-relevant tissues to identify BE/EA candidate genes. To investigate the relationship with reported BE/EA risk factors, a linkage disequilibrium score regression (LDSR) analysis was performed. BE/EA risk models were developed combining clinical/lifestyle risk factors with polygenic risk scores (PRS) derived from the GWAS meta-analysis. RESULTS: The GWAS meta-analysis identified 27 BE and/or EA risk loci, 11 of which were novel. The TWAS identified promising BE/EA candidate genes at seven GWAS loci and at five additional risk loci. The LDSR analysis led to the identification of novel genetic correlations and pointed to differences in BE and EA aetiology. Gastro-oesophageal reflux disease appeared to contribute stronger to the metaplastic BE transformation than to EA development. Finally, combining PRS with BE/EA risk factors improved the performance of the risk models. CONCLUSION: Our findings provide further insights into BE/EA aetiology and its relationship to risk factors. The results lay the foundation for future follow-up studies to identify underlying disease mechanisms and improving risk prediction.