Boring biopsy with rapid on-site evaluation for gastric gastrointestinal stromal tumor: A pilot study.

One of 4 patients (25%) and 2 of 3 patients (67%) were correctly diagnosed by conventional and hot boring biopsies, respectively. The median procedure time of conventional and hot boring biopsies was 21 (range, 13-33) and 17 (range, 16-23) min, respectively. Rapid on-site evaluation was performed totally 12 times in 7 patients. The positive and negative predictive values of rapid on-site evaluation for gastrointestinal stromal tumor were 0.5 (0.12-0.88) and 1.0 (0.42-1.00), respectively.

Concomitant Clostridioides difficile and Edwardsiella tarda Infection in a Patient with Ulcerative Colitis.

Infectious enteritis is common in patients with inflammatory bowel disease (IBD). This case presented a young woman who underwent remission maintenance therapy for ulcerative colitis (UC). She was suspected of having concomitant Clostridioides difficile and Edwardsiella tarda infections. The patient's symptoms did not improve after initial antibiotic therapy; thus, the treatment strategy was modified to include an intravenous corticosteroid to treat the UC flare-up. Her symptoms significantly improved after corticosteroid administration. This is the first report of a case in which concomitant C. difficile and E. tarda infections occurred with UC flare-up.

Rapid endothermal development of juvenile pacific bluefin tuna.

An important trait of Pacific bluefin tuna (PBT) is their ability to maintain their body temperature above the ambient temperature, which allows them to occupy a wider ecological niche. However, the size at which this ability in nature develops is unclear. Therefore, this study aimed to clarify this point by monitoring the body temperature and the surrounding ambient temperature as the fish grew. PBT with fork lengths (FLs) ranging from 19.5 to 28.0 cm were implanted with archival electronic tags and released into the ocean. Data from 41 fish were obtained (recorded body and water temperatures, light level, and swimming depth (pressure) at 30-s intervals) and analyzed to elucidate the development of the ability of PBT to maintain a high body temperature. Body temperature of a PBT (< FL of ca. 40 cm) decreased in response to a vertical movement down to cooler depths, but higher body temperatures were maintained as the fish grew. The body temperature was then continuously maintained above ambient temperatures and fluctuated independently when fish attained more than 40 cm FL. Estimation of the whole-body heat-transfer coefficient and heat-production rate indicated that the latter decreased slowly with growth, while the former decreased by one order of magnitude when tuna reached 52 cm FL. Additionally, in the daytime, the whole-body heat-transfer coefficient was significantly higher than that at nighttime. Unlike other fishes including other Thunnus species, inhabiting tropical/subtropical waters, PBT rapidly acquire higher thermo-conservation ability when young, allowing capture of high-quality prey abundant in temperate waters to support high growth rates during early life.

Endoscopic submucosal dissection in the right lateral position for early gastric cancer in the fornix.

Video 1The right lateral position facilitated endoscopic submucosal dissection of early gastric cancer in the fornix.

Clinical features of superficial esophagus squamous cell carcinoma according to alcohol-degrading enzyme ADH1B and ALDH2 genotypes.

BACKGROUND: Inactivated alcohol dehydrogenase (ADH) and aldehyde dehydrogenase (ALDH) are related to esophageal carcinogenesis. We aimed to clarify the clinical features associated with the alcohol-degrading enzyme genotypes, ADH1B and ALDH2. We also investigated the risk factors for metachronous esophageal squamous cell carcinoma (ESCC) and head and neck SCC (HNSCC). METHODS: We conducted a single-center, retrospective study including patients with ESCC treated by endoscopic resection. Patients were recruited between October 2020 and September 2021. Buccal mucosal swabs were obtained from them to analyze the genetic polymorphisms affecting ADH (ADH1B) and ALDH (ALDH2) activity. Patients were categorized into three groups: both inactivated = double-inactivated group; inactivated ADH1B or ALDH2 = single-inactivated group; and both activated = activated group. RESULTS: Among the 297 enrolled patients, patients in the double-inactivated group were significantly younger (P < 0.001) and 60% of them were ≤ 50 years old. This group also had more ESCCs located in the upper esophagus (P < 0.001) and more simultaneous multiple ESCCs (P = 0.044). More than half of the patients had multiple Lugol-voiding lesions (LVLs) (P < 0.001) and heavy alcohol consumers (P = 0.012). Metachronous ESCC and HNSCC were more common in the double-inactivated group (P < 0.001, P = 0.001). Multivariate analysis identified located in the upper esophagus, multiple LVLs and history of HNSCC as risk factors for metachronous ESCC. CONCLUSIONS: Activation patterns of alcohol-metabolizing enzymes were related to age at ESCC onset, lesion location, and metachronous ESCC and HNSCC. Different approaches to the prophylaxis and treatment of esophageal cancer should be considered, depending on the enzyme activity pattern.

Utility of an artificial intelligence system for classification of esophageal lesions when simulating its clinical use.

Previous reports have shown favorable performance of artificial intelligence (AI) systems for diagnosing esophageal squamous cell carcinoma (ESCC) compared with endoscopists. However, these findings don't reflect performance in clinical situations, as endoscopists classify lesions based on both magnified and non-magnified videos, while AI systems often use only a few magnified narrow band imaging (NBI) still images. We evaluated the performance of the AI system in simulated clinical situations. We used 25,048 images from 1433 superficial ESCC and 4746 images from 410 noncancerous esophagi to construct our AI system. For the validation dataset, we took NBI videos of suspected superficial ESCCs. The AI system diagnosis used one magnified still image taken from each video, while 19 endoscopists used whole videos. We used 147 videos and still images including 83 superficial ESCC and 64 non-ESCC lesions. The accuracy, sensitivity and specificity for the classification of ESCC were, respectively, 80.9% [95% CI 73.6-87.0], 85.5% [76.1-92.3], and 75.0% [62.6-85.0] for the AI system and 69.2% [66.4-72.1], 67.5% [61.4-73.6], and 71.5% [61.9-81.0] for the endoscopists. The AI system correctly classified all ESCCs invading the muscularis mucosa or submucosa and 96.8% of lesions ≥ 20 mm, whereas even the experts diagnosed some of them as non-ESCCs. Our AI system showed higher accuracy for classifying ESCC and non-ESCC than endoscopists. It may provide valuable diagnostic support to endoscopists.

Positive predictive value of the clinical diagnosis of T1a-epithelial/lamina propria esophageal cancer depends on lesion size.

OBJECTIVES: Endoscopic resection (ER) is a minimally invasive treatment for esophageal squamous cell carcinoma (ESCC). However, stricture may develop after ER for widespread lesions. Application of ER is justified if these cancers are pathological T1a-epithelial/lamina propria (pEP/LPM) cancers that can be cured by ER. We conducted a study to clarify the association between pathological invasion depth and lesion size or circumference in clinical (c) EP/LPM cancers. METHODS: From our database, we identified patients diagnosed with cEP/LPM ESCC via endoscopic examination who underwent endoscopic or surgical tumor resection. The accuracy of the cEP/LPM ESCC diagnosis was determined by histologically diagnosing cancer invasion depth as a reference standard. RESULTS: Between January 2015 and December 2019, 1271 cancer patients were diagnosed with cEP/LPM ESCC, of which 1195 (94.0%) were correctly diagnosed with pEP/LPM cancer. The positive predictive value (PPV) classified according to lesion sizes of ≤25, 26-49, and ≥50 mm was 95.8% (981/1024 lesions), 89.7% (191/213 lesions), and 67.6% (23/34 lesions), respectively. PPV according to the circumferential extent of <3/4, ≥3/4, and <1, and whole was 94.6% (1164/1230 lesions), 75.0% (24/32 lesions), and 77.8% (7/9 lesions), respectively. In multivariate analysis, the PPV of cEP/LPM ESCC was significantly associated with lesion size (P < 0.001) and male sex. CONCLUSIONS: Between January 2015 and December 2019, 1271 cancer patients were diagnosed with cEP/LPM ESCC, of which 1195 (94.0%) were correctly diagnosed with pEP/LPM cancer. The PPV of cEP/LPM ESCC was related to lesion size. Treatment should be determined considering the high risk of cancer invasion into the muscularis mucosa or deeper in cEP/LPM cancers with a lesion size of ≥50 mm.

Geriatric nutritional risk index as a prognostic indicator in elderly patients with early colorectal cancer undergoing endoscopic submucosal dissection.

OBJECTIVES: Data on the long-term outcomes of endoscopic submucosal dissection (ESD) performed in elderly patients with early colorectal cancer (CRC) are limited. We analyzed the prognosis of elderly CRC patients, not only from the viewpoint of treatment curability but also from the patients' baseline physical condition assessed by several indexes. METHODS: A retrospective analysis of 729 patients aged ≥75 years who underwent ESD for Tis/T1 CRC in 16 institutions was conducted. The patients were classified into three groups based on curability: curative ESD (Group A, n = 582), non-curative ESD with additional surgery (Group B, n = 60), and non-curative ESD without additional surgery (Group C, n = 87). Overall survival (OS) was compared among the groups, and factors associated with reduced OS were investigated. RESULTS: The median follow-up periods in Groups A, B, and C were 41, 49, and 46 months, respectively (P = 0.62), during which 92 patients died. Two patients (0.3%) in Group A, none (0%) in Group B, and three (3.4%) in Group C died of CRC. Three-year OS rates in Groups A, B, and C were 93.9%, 96.1%, and 90.1%, respectively, without a significant difference (P = 0.07). Multivariate analysis indicated low (<96.3) geriatric nutritional risk index (GNRI) as the sole independent predictor for reduced OS (hazard ratio 3.37; 95% confidence interval 2.18-5.22; P < 0.0001). CONCLUSIONS: Low GNRI, but not the curability attained by ESD, was independently associated with reduced OS in patients with early CRC aged ≥75 years.

Endoscopists' diagnostic accuracy in detecting upper gastrointestinal neoplasia in the framework of artificial intelligence studies.

BACKGROUND: Estimates on miss rates for upper gastrointestinal neoplasia (UGIN) rely on registry data or old studies. Quality assurance programs for upper GI endoscopy are not fully established owing to the lack of infrastructure to measure endoscopists' competence. We aimed to assess endoscopists' accuracy for the recognition of UGIN exploiting the framework of artificial intelligence (AI) validation studies. METHODS: Literature searches of databases (PubMed/MEDLINE, EMBASE, Scopus) up to August 2020 were performed to identify articles evaluating the accuracy of individual endoscopists for the recognition of UGIN within studies validating AI against a histologically verified expert-annotated ground-truth. The main outcomes were endoscopists' pooled sensitivity, specificity, positive and negative predictive value (PPV/NPV), and area under the curve (AUC) for all UGIN, for esophageal squamous cell neoplasia (ESCN), Barrett esophagus-related neoplasia (BERN), and gastric adenocarcinoma (GAC). RESULTS: Seven studies (2 ESCN, 3 BERN, 1 GAC, 1 UGIN overall) with 122 endoscopists were included. The pooled endoscopists' sensitivity and specificity for UGIN were 82 % (95 % confidence interval [CI] 80 %-84 %) and 79 % (95 %CI 76 %-81 %), respectively. Endoscopists' accuracy was higher for GAC detection (AUC 0.95 [95 %CI 0.93-0.98]) than for ESCN (AUC 0.90 [95 %CI 0.88-0.92]) and BERN detection (AUC 0.86 [95 %CI 0.84-0.88]). Sensitivity was higher for Eastern vs. Western endoscopists (87 % [95 %CI 84 %-89 %] vs. 75 % [95 %CI 72 %-78 %]), and for expert vs. non-expert endoscopists (85 % [95 %CI 83 %-87 %] vs. 71 % [95 %CI 67 %-75 %]). CONCLUSION: We show suboptimal accuracy of endoscopists for the recognition of UGIN even within a framework that included a higher prevalence and disease awareness. Future AI validation studies represent a framework to assess endoscopist competence.

Delineating the extent of esophageal squamous cell carcinoma.

BACKGROUND: Lugol chromoendoscopy has been conventionally used for the detection and delineation of esophageal squamous cell carcinoma (SCC). However, the boundaries of some lesions are unclear even with Lugol chromoendoscopy, and there is a risk of residual lesions or over-excision. This study aimed to evaluate the utility of narrow-band imaging (NBI) for the delineation of esophageal SCC in endoscopic resection. METHODS: Among 367 esophageal SCCs endoscopically resected between January and December 2019 at our institute, this retrospective study included consecutive lesions, which were first marked with NBI, followed by Lugol chromoendoscopy. The proportion of residual cancer, which was defined as histologically proven cancer confirmed adjacent to the scar within 1 year after endoscopic resection, was evaluated. To evaluate whether the marks added by Lugol chromoendoscopy after NBI marking were more reliable, we evaluated the presence of cancer in the iodine-unstained area outside the NBI-determined marks, i.e., the cancerous area missed by NBI. The presence of cancer in the iodine-stained areas inside the NBI-determined marks, i.e., the cancerous area missed by Lugol, was also evaluated. These were compared to assess the risk of residual cancer in endoscopic resection with NBI and Lugol chromoendoscopy. RESULTS: Among 304 lesions, 2 (0.7%) residual cancers were detected. The cancerous area missed by NBI and the cancerous area missed by Lugol were identified in 18 (6%) and 43 (14%) lesions, respectively (P = 0.001). CONCLUSIONS: NBI might be acceptable for delineating the extent of esophageal SCCs that are difficult to delineate with Lugol chromoendoscopy.

Curative value of underwater endoscopic mucosal resection for submucosally invasive colorectal cancer.

BACKGROUND AND AIM: Occasionally, colorectal tumors without characteristics of deep submucosal invasion are found to be invasive upon pathological evaluation after endoscopic resection (ER). Because the resection depth for underwater endoscopic mucosal resection (UEMR) has not been clarified, we evaluated the feasibility of UEMR for pathologically invasive colorectal cancer (pT1-CRC). METHODS: We retrospectively investigated data on the backgrounds and outcomes of patients with pT1-CRC who underwent UEMR between January 2014 and June 2019 at our institute. As a reference standard, the backgrounds and outcomes of pT1-CRCs that had undergone conventional EMR (CEMR) were also investigated. RESULTS: Thirty-one patients (median age, 68 years [range, 32-88 years]; 22 men [71%]) were treated with UEMR. Median lesion size was 17 mm (range, 6-50 mm). The endoscopic complete resection rate was 100%. The overall en bloc resection rate was 77%, and the VM0, HM0, and R0 resection rates were 81%, 58%, and 55%, respectively. In cases of pT1a (invasion <1000 µm)-CRC (n = 14), the en bloc, VM0, and R0 resection rates were 92%, 100%, and 71%, respectively. Seventeen patients (five with risk factors for lymph node metastasis and 12 without) were followed up, and no local recurrence and distant metastasis were observed during the follow-up period (median follow-up period, 18 months [range, 6-62 months]) after UEMR. The outcomes of UEMR seemed to be comparable with those of CEMR (n = 32). CONCLUSIONS: The VM0 rate of UEMR for pT1-CRC, especially for pT1a-CRC, without characteristics of deep submucosal invasion seems feasible.

Esophageal metal stent for malignant obstruction after prior radiotherapy.

The association between severe adverse events (SAEs) and prior radiotherapy or stent type remains controversial. Patients with esophageal or esophagogastric junctional cancer who underwent stent placement (2005-2019) were enrolled in this retrospective study conducted at a tertiary cancer institute in Japan. The exclusion criteria were follow-up period of < 1 month and insufficient data on stent type or cancer characteristics. We used Mann-Whitney's U test for quantitative data and Fisher's exact test for categorical data. Multivariate analysis was performed using a logistic regression model. 107 stents were placed. Low radial-force stents (L group) were used in 51 procedures and high radial-force stents (H group) in 56 procedures. SAEs developed after nine procedures, the median interval from stent placement being 6 days (range, 1-141 days). SAEs occurred more frequently in the H (14%: 8/56) than in the L group (2%: 1/51) (P = 0.03). In patients who had undergone prior radiotherapy, SAEs were more frequent in the H (36%: 4/11) than in the L group (0%: 0/13) (P = 0.03). Re-obstruction and migration occurred after 16 and three procedures, respectively; these rates did not differ significantly between groups (P = 0.59, P = 1, respectively). Low radial-force stents may reduce the risk of SAEs after esophageal stenting.

Usefulness of an artificial intelligence system for the detection of esophageal squamous cell carcinoma evaluated with videos simulating overlooking situation.

OBJECTIVES: Artificial intelligence (AI) systems have shown favorable performance in the detection of esophageal squamous cell carcinoma (ESCC). However, previous studies were limited by the quality of their validation methods. In this study, we evaluated the performance of an AI system with videos simulating situations in which ESCC has been overlooked. METHODS: We used 17,336 images from 1376 superficial ESCCs and 1461 images from 196 noncancerous and normal esophagi to construct the AI system. To record validation videos, the endoscope was passed through the esophagus at a constant speed without focusing on the lesion to simulate situations in which ESCC has been missed. Validation videos were evaluated by the AI system and 21 endoscopists. RESULTS: We prepared 100 video datasets, including 50 superficial ESCCs, 22 noncancerous lesions, and 28 normal esophagi. The AI system had sensitivity of 85.7% (54 of 63 ESCCs) and specificity of 40%. Initial evaluation by endoscopists conducted with plain video (without AI support) had average sensitivity of 75.0% (47.3 of 63 ESCC) and specificity of 91.4%. Subsequent evaluation by endoscopists was conducted with AI assistance, which improved their sensitivity to 77.7% (P = 0.00696) without changing their specificity (91.6%, P = 0.756). CONCLUSIONS: Our AI system had high sensitivity for the detection of ESCC. As a support tool, the system has the potential to enhance detection of ESCC without reducing specificity. (UMIN000039645).

Validity of endoscopic resection for clinically diagnosed T1a-MM/T1b-SM1 N0 M0 esophageal squamous cell carcinoma.

BACKGROUND: Previous guidelines have not described clear recommendations for performing endoscopic resection (ER) of T1a-muscularis mucosa (MM)/T1b-submucosal (SM1) cancers that have invaded ≤ 200 µm because these are considered to have a non-negligible risk of metastasis based on previous analyses of pathologically diagnosed (p)MM/SM1 cancers. Considering that the indication for ER is determined based on a clinical diagnosis, the applicability of ER should be investigated in clinical (c)MM/SM1 but not pMM/SM1 cancers. This study aimed to evaluate validity of ER for cMM/SM1 cancers. METHODS: In total, 175 cMM/SM1 esophageal squamous cell carcinoma cases that were endoscopically or surgically resected between January 2008 and December 2018 were identified from a prospectively maintained database. We histologically evaluated resected specimens and divided them into low- (n = 92) and high-risk (n = 83) cancers for metastasis. RESULTS: Univariate analysis showed that longer tumor length and larger circumferential extent were significantly correlated with high-risk cancer (P < 0.001). Multivariate analysis showed that tumor circumference was an independent predictor of high-risk cancer (P = 0.036). The proportion of low-risk cancers among cases with ≤ 3/4, > 3/4 and < 1, and whole circumferential extent were 59, 17, and 14%, respectively, and the post-ER stricture rates of these groups were 12, 33, and 100%, respectively. CONCLUSION: ER is the first-line treatment for cMM/SM1 cancers with ≤ 3/4 circumferential extent considering that 59% of cMM/SM1 cancers were low-risk cancers for which ER is mostly curative. ER is not recommended for whole circumferential cMM/SM1 cancers given the low proportion of low-risk cancers and the high risk of stricture after ER.

Endoscopic findings in the soft palatal mucosa are associated with the risk of esophageal squamous cell carcinoma.

BACKGROUND AND AIM: We investigated endoscopic findings of the soft palatal mucosa to identify factors associated with esophageal squamous cell carcinoma (ESCC). METHODS: This study was conducted during endoscopic examinations of subjects at Osaka International Cancer Institute from January 2020 through May 2020. We took endoscopic images of the soft palate under non-magnifying and mild-magnifying observations. Subjects with ESCC or a history of ESCC were defined as the ESCC group. Two endoscopists who were blinded to subjects' clinical information interpreted 10 endoscopic findings: melanosis, brownish changes, whitish epithelium, vasodilation, circular alignment of dilated vessels, uneven surface, uneven epithelial color, uneven vessel visibility, palate ridge, and erosion. Subjects were interviewed about their alcohol use, smoking, and flushing reactions. RESULTS: Two hundred eighty-two subjects, including 151 in the ESCC group and 131 in the non-ESCC group, were included in the analysis. Univariate analyses and multivariate logistic regression demonstrated that melanosis, whitish epithelium, and vasodilation were significantly associated with ESCC. The positive likelihood ratios (PLRs) of melanosis, whitish epithelium, and vasodilation were 3.3, 4.2, and 2.8, respectively. Additionally, the PLRs for three of the endoscopic findings in subjects with drinking and smoking habits were higher than in those without these habits-PLRs ranging from 7.23 to 19.1. High PLRs for three endoscopic findings suggested a high possibility of ESCC. Interobserver agreement was substantial for whitish epithelium, moderate for melanosis, and fair for vasodilation. CONCLUSIONS: Three endoscopic findings in soft palate were considered to be useful as alarming signs that indicate ESCC risk.

Diagnosis of pharyngeal cancer on endoscopic video images by Mask region-based convolutional neural network.

OBJECTIVES: We aimed to develop an artificial intelligence (AI) system for the real-time diagnosis of pharyngeal cancers. METHODS: Endoscopic video images and still images of pharyngeal cancer treated in our facility were collected. A total of 4559 images of pathologically proven pharyngeal cancer (1243 using white light imaging and 3316 using narrow-band imaging/blue laser imaging) from 276 patients were used as a training dataset. The AI system used a convolutional neural network (CNN) model typical of the type used to analyze visual imagery. Supervised learning was used to train the CNN. The AI system was evaluated using an independent validation dataset of 25 video images of pharyngeal cancer and 36 video images of normal pharynx taken at our hospital. RESULTS: The AI system diagnosed 23/25 (92%) pharyngeal cancers as cancers and 17/36 (47%) non-cancers as non-cancers. The transaction speed of the AI system was 0.03 s per image, which meets the required speed for real-time diagnosis. The sensitivity, specificity, and accuracy for the detection of cancer were 92%, 47%, and 66% respectively. CONCLUSIONS: Our single-institution study showed that our AI system for diagnosing cancers of the pharyngeal region had promising performance with high sensitivity and acceptable specificity. Further training and improvement of the system are required with a larger dataset including multiple centers.

Magnifying endoscopy with crystal violet staining for immune checkpoint inhibitor-associated colitis.

BACKGROUND AND AIM: There exists no evidence on the relationship between endoscopic and histologic findings. Furthermore, even after multiple biopsy specimens were obtained, histologic examination usually fails to show the characteristic features of immune checkpoint inhibitor-associated colitis. In this study, we explored the endoscopic and histologic findings of immune checkpoint inhibitor-associated colitis. METHODS: Patients diagnosed with immune checkpoint inhibitor-associated colitis at our hospital between March 2018 and December 2018 were retrospectively assessed. The degree of mucosal inflammation was evaluated using endoscopic inflammation grade (inactive, mild, moderate, or severe disease) and further observed using magnifying endoscopy with crystal violet staining. Pit structures were classified into three types: regularly arranged pits with circular or elliptical shape (R type), irregularly arranged pits with inhomogeneous size and morphology (IR type), and pits with reduced density or pits that partially disappeared (AD type). RESULTS: Eleven patients (median age, 71 years; range, 44-83 years) were diagnosed with immune checkpoint inhibitor-associated colitis. All characteristic histologic findings, including crypt distortion, crypt abscesses, and apoptotic bodies, were observed at sites with moderate-to-severe endoscopic inflammation but not at sites with inactive-to-mild endoscopic inflammation. Characteristic histologic features were observed in 0%, 50%, and 100% of R-type, IR-type, and AD-type mucosa, respectively. CONCLUSIONS: We revealed the possible utility of endoscopic images for selecting suitable target sites for biopsy and showed that endoscopic findings could reduce the time lag associated with tissue diagnosis and sampling errors due to biopsy.

Artificial intelligence for the detection of esophageal and esophagogastric junctional adenocarcinoma.

BACKGROUND AND AIM: Conventional endoscopy for the early detection of esophageal and esophagogastric junctional adenocarcinoma (E/J cancer) is limited because early lesions are asymptomatic, and the associated changes in the mucosa are subtle. There are no reports on artificial intelligence (AI) diagnosis for E/J cancer from Asian countries. Therefore, we aimed to develop a computerized image analysis system using deep learning for the detection of E/J cancers. METHODS: A total of 1172 images from 166 pathologically proven superficial E/J cancer cases and 2271 images of normal mucosa in esophagogastric junctional from 219 cases were used as the training image data. A total of 232 images from 36 cancer cases and 43 non-cancerous cases were used as the validation test data. The same validation test data were diagnosed by 15 board-certified specialists (experts). RESULTS: The sensitivity, specificity, and accuracy of the AI system were 94%, 42%, and 66%, respectively, and that of the experts were 88%, 43%, and 63%, respectively. The sensitivity of the AI system was favorable, while its specificity for non-cancerous lesions was similar to that of the experts. Interobserver agreement among the experts for detecting superficial E/J was fair (Fleiss' kappa = 0.26, z = 20.4, P < 0.001). CONCLUSIONS: Our AI system achieved high sensitivity and acceptable specificity for the detection of E/J cancers and may be a good supporting tool for the screening of E/J cancers.