Search

Most gastric neuroendocrine tumors (NETs) develop from enterochromaffin-like (ECL) cells. ECL-cell NETs are classically categorized into three types according to their etiology. A 50-year-old woman presented with submucosal tumor-like lesions in the stomach, which were identified via esophagogastroduodenoscopy. Although esophagogastroduodenoscopy and pathological findings of biopsy specimens showed an absence of mucosal atrophy in the body of the stomach, sticky, adherent, dense mucus was observed. All lesions were diagnosed as ECL-cell NETs based on histological examination findings; however, ECL-cell NETs did not apply to any of the classic types I-III categorization based on laboratory, computed tomography, and 24-h intragastric pH monitoring test findings. Endoscopic submucosal dissection of the tumor was performed. Pathological findings of the excised specimen indicated that parietal cell hyperplasia with a protrusion, dilated fundic glands, and endocrine cell hyperplasia in the background mucosa, and parietal cells were not immunostained for the α-subunits of H

Introduction: Although endocytoscopy (EC) with narrow-band imaging (NBI) is effective in diagnosing gastric cancer, no diagnostic system has been validated. We explored a specific diagnostic system for gastric cancer using EC with NBI. Methods: Equal numbers of images from cancerous and noncancerous areas (114 images each) were assessed by endoscopists with (development group: 33) and without (validation group: 28) specific training in magnifying endoscopy with NBI. Microvascular and microsurface patterns (MS) in each image were evaluated. Lesions were diagnosed as cancerous when patterns were deemed “irregular.” The accuracy, sensitivity, specificity, positive predictive value, and negative predictive value of a diagnosis according to patterns on EC with NBI (microvascular pattern [MV] alone, MS alone, and both) were evaluated and compared between groups to determine the diagnostic performance. Results: In the development and validation groups, diagnoses based on the MV alone had significantly higher accuracy (91.7% vs. 76.3%, p < 0.0001 and 92.5% vs. 67.5%, p < 0.0001, respectively) and sensitivity (88.6% vs. 68.3%, p < 0.0001 and 89.5% vs. 38.6%, p < 0.0001, respectively) than those based on the MS alone. In both groups, there were no significant differences in diagnostic accuracy between using the MV alone and both patterns. Discussion/Conclusion: Evaluation of the MV alone is a simple and accurate diagnostic method for gastric cancer. This system could find widespread applications in clinical practice.

Background An association between specific endoscopic findings and high-grade dysplasia/carcinoma in superficial nonampullary duodenal epithelial tumors ≤ 5 mm in size has not been reported. We sought to identify the endoscopic findings associated with high-grade dysplasia/carcinoma in patients with superficial nonampullary duodenal epithelial tumors ≤ 5 mm. Methods We retrospectively assessed the data of 84 patients (88 lesions; low-grade dysplasia: n = 35, high-grade dysplasia/carcinoma: n = 53) with superficial nonampullary duodenal epithelial tumors who underwent initial treatment at a single center (from July 2009 to April 2021). All the patients had lesions sized ≤ 5 mm. We assumed that the endoscopic findings were independently associated with high-grade dysplasia/carcinoma and determined the accuracy, sensitivity, and specificity of a combination of independent factors for diagnosing high-grade dysplasia/carcinoma and low-grade dysplasia. Results Multivariate logistic regression of significant factors in the univariate analysis revealed that lesions with depressed morphology (odds ratio: 23.9, 95% confidence interval: 2.8-204.2; p = 0.0037) and a reddish color (odds ratio: 175.7, 95% confidence interval: 11.4-2697.1; p = 0.0002) were independently associated with high-grade dysplasia/carcinoma. McNemar's test revealed that combining the macroscopic type and color provided significantly higher sensitivity for diagnosing high-grade dysplasia/carcinoma than color alone (98.1%, 95% confidence interval: 90.1-99.7 vs. 71.7%, 95% confidence interval: 58.4-82.0; p = 0.0002). Conclusions Reddish and depressed-type lesions before treatment were associated with high-grade dysplasia/carcinoma. Combining the macroscopic type and color can help detect high-grade dysplasia/carcinoma. These findings could help clinicians determine the best therapeutic strategy for patients with smaller (≤ 5 mm) superficial nonampullary duodenal epithelial tumors in clinical settings.

OBJECTIVES Endoscopic resection (ER) is indicated for a wide range of superficial esophageal squamous cell carcinomas (ESCCs). We examined the long-term outcomes in patients with pathological (p) invasion of ESCC into the T1a-muscularis mucosae (MM) and T1b-submucosa (SM) after ER, for which data on prognosis are limited. METHODS Of the 1217 patients with superficial ESCC who underwent ER, 225 patients with a pathological diagnosis of ESCC invasion into the MM, minute submucosal invasion ≤200 µm (SM1), or massive submucosal invasion (SM2) were included. In patients with lymphovascular invasion, droplet infiltration, or SM2 invasion, additional treatments, including chemoradiation (CRT) or esophagectomy with two- to three-field lymph node dissection, were recommended. The median observation period was 66 months (interquartile range 48-91 months). RESULTS In total, there were 151, 28, and 46 pT1a-MM, pT1b-SM1, and pT1b-SM2 cases, respectively. Metastatic recurrence was observed in 1.3%, 10.7%, and 6.5% patients with pT1a-MM, pT1b-SM1, and pT1b-SM2 ESCCs, respectively. Of the eight patients with metastatic recurrence, six were successfully treated, and two died of ESCC. The 5-year overall survival rates were 84.1%, 71.4%, and 67.4%, the 5-year relapse-free survival rates were 82.8%, 64.3%, and 65.2%, and the 5-year disease-specific survival rates were 100%, 96.4%, and 99.1% in patients with pT1a-MM, pT1b-SM1, and pT1b-SM2 ESCCs, respectively. Multivariate analysis showed that additional CRT and esophagectomy, and T1b-SM2 were positively and negatively associated with overall survival, respectively. CONCLUSIONS Endoscopic resection preceding appropriate additional treatments resulted in favorable outcomes. Many cases of metastatic recurrence in this cohort could be successfully treated.

Aims To compare endoscopy gastric cancer images diagnosis rate between artificial intelligence (AI) and expert endoscopists. Patients and methods We used the retrospective data of 500 patients, including 100 with gastric cancer, matched 1:1 to diagnosis by AI or expert endoscopists. We retrospectively evaluated the noninferiority (prespecified margin 5 %) of the per-patient rate of gastric cancer diagnosis by AI and compared the per-image rate of gastric cancer diagnosis. Results Gastric cancer was diagnosed in 49 of 49 patients (100 %) in the AI group and 48 of 51 patients (94.12 %) in the expert endoscopist group (difference 5.88, 95 % confidence interval: −0.58 to 12.3). The per-image rate of gastric cancer diagnosis was higher in the AI group (99.87 %, 747 /748 images) than in the expert endoscopist group (88.17 %, 693 /786 images) (difference 11.7 %). Conclusions Noninferiority of the rate of gastric cancer diagnosis by AI was demonstrated but superiority was not demonstrated.

We herein report two cases of early esophageal adenocarcinoma derived from non-Barrett's columnar epithelium. Both patients, a 65-year-old woman and 60-year-old man, had elevated lesions on white-light imaging. Magnifying endoscopy revealed slightly irregular surface and vessel patterns, and both patients were successfully treated with endoscopic submucosal dissection. Histopathologically, both lesions comprised of well-differentiated gastric mucin phenotype adenocarcinoma. One lesion was accompanied by ectopic gastric mucosa, but the other was speculated to be ectopic gastric mucosa according to the tumor locus at the upper thoracic esophagus. Despite its rarity, endoscopists should consider the existence of adenocarcinoma derived from non-Barrett's columnar epithelium.

Radiation-induced hemorrhagic gastritis is a relatively uncommon complication of irradiation that can be severe. However, appropriate treatment guidelines have not yet been established because of the small number of known cases. At our hospital, we encountered nine cases of radiation-induced hemorrhagic gastritis between July 2005 and July 2018. All patients initially underwent argon plasma coagulation (APC) for hemostasis. The treatment was highly effective, and hemostasis was successfully achieved in eight of the cases. Hemostasis could not be achieved in one case treated with APC; therefore, surgical resection was required. This patient had risk factors, such as liver cirrhosis and a history of abdominal surgery. Our case series suggests that APC is an effective hemostatic method that should be considered as the initial treatment option for radiation-induced hemorrhagic gastritis; however, surgical resection may be considered when the patient is at high risk for rebleeding.

Helicobacter pylori (HP) is a major etiologic driver of diffuse-type gastric cancer (DGC). However, improvements in hygiene have led to an increase in the prevalence of HP-naïve DGC; that is, DGC that occurs independent of HP. Although multiple genomic cohort studies for gastric cancer have been conducted, including studies for DGC, distinctive genomic differences between HP-exposed and HP-naïve DGC remain largely unknown. Here, we employed exome and RNA sequencing with immunohistochemical analyses to perform binary comparisons between 36 HP-exposed and 27 HP-naïve DGCs from sporadic, early-stage, and intramucosal or submucosal tumor samples. Among the samples, 33 HP-exposed and 17 HP-naïve samples had been preserved as fresh-frozen samples. HP infection status was determined using stringent criteria. HP-exposed DGCs exhibited an increased single nucleotide variant burden (HP-exposed DGCs; 1.97 [0.48-7.19] and HP-naïve DGCs; 1.09 [0.38-3.68] per megabase; p = 0.0003) and a higher prevalence of chromosome arm-level aneuploidies (p 0.0001). CDH1 was mutated at similar frequencies in both groups, whereas the RHOA-ARHGAP pathway misregulation was exclusive to HP-exposed DGCs (p = 0.0167). HP-exposed DGCs showed gains in chromosome arms 8p/8q (p 0.0001), 7p (p = 0.0035), and 7q (p = 0.0354), and losses in 16q (p = 0.0167). Immunohistochemical analyses revealed a higher expression of intestinal markers such as CD10 (p 0.0001) and CDX2 (p = 0.0002) and a lower expression of the gastric marker, MUC5AC (p = 0.0305) among HP-exposed DGCs. HP-naïve DGCs, on the other hand, had a purely gastric marker phenotype. This work reveals that HP-naïve and HP-exposed DGCs develop along different molecular pathways, which provide a basis for early detection strategies in high incidence settings. © 2022 The Authors. The Journal of Pathology published by John WileySons Ltd on behalf of The Pathological Society of Great Britain and Ireland.

Whether antithrombotic drugs increase the risk of post-esophageal endoscopic resection bleeding is unknown. This study examined the effect of antithrombotic drugs, aspirin, thienopyridine, direct oral anticoagulants (DOAC), and warfarin, on post-esophageal endoscopic resection bleeding.We enrolled 957 patients (1202 esophageal tumors) treated with endoscopic resection and classified them based on antithrombotic drug use as no use, aspirin, thienopyridine, DOAC, and warfarin. Patients using antiplatelet drugs (i.e. aspirin and thienopyridine) were further sub-classified based on their continued or discontinued use before endoscopic resection. The bleeding rates were compared between these groups to assess the effects of antithrombotic drug use and interruption of antiplatelet therapy on post-esophageal endoscopic resection bleeding.The post-endoscopic resection bleeding rate was 0.3% (95% CI, 0.1-1) in the group without antithrombotic drug use, 4.5% (95% CI, 0.1-23) in the aspirin-continued group, 2.9% (95% CI, 0.1-15) in the aspirin-discontinued group, 0% (95% CI, 0-78) in the replaced thienopyridine with aspirin group, 0% (95% CI, 0-26) in the thienopyridine-discontinued group, 13% (95% CI, 1.6-38) in the DOAC group, and 0% (95% CI, 0-45) in the warfarin group. The post-endoscopic resection bleeding rate in the DOAC group was significantly higher than that in the group without antithrombotic drugs (P = 0.003). The post-endoscopic resection bleeding rates did not differ between the other groups.Our results suggest that discontinuing aspirin is not necessary for esophageal endoscopic resection while we must be careful regarding DOAC.

Background and Aim Helicobacter pylori antibody levels in the blood are currently measured using an ELISA. In April 2016, FUJIFILM Wako Pure Chemical Corporation launched the “l‐type Wako Helicobacter pylori antibody J" test, which is based on the latex agglutination turbidimetric immunoassay. In this study, we investigated the usefulness of the Wako test. Methods We measured H. pylori antibody levels using both the ELISA and Wako test in 180 patients who underwent upper gastrointestinal endoscopy at our hospital between September 2017 and February 2019. Ninety patients were infected with H. pylori. We calculated the diagnostic accuracy, sensitivity, and specificity of each test and the concordance rate between the ELISA and Wako test. If lower limits of 90% confidence intervals (CIs) for each diagnostic validity exceeded the 85% threshold, the usefulness of the diagnostic test was confirmed. Results The diagnostic accuracy, sensitivity, and specificity were 94.4% (90% CI, 90.8–97.0%), 94.4% (90% CI, 88.7–97.8%), and 94.4% (90% CI, 88.7–97.8%), respectively, when the Wako test was used, and 94.4% (90% CI, 90.8–97.0%), 88.9% (90% CI, 81.9–93.8%), and 100% (90% CI, 96.0–100%), respectively, when the ELISA was used. The concordance rate between the two tests was high (κ = 0.8444). Conclusions We confirmed the usefulness of the Wako test, especially when screening for H. pylori infection, due to its high sensitivity., “l‐type Wako Helicobacter pylori antibody J” test is based on the latex agglutination turbidimetric immunoassay. Many samples can be analyzed simultaneously, rapidly, and more economically using this kit. l‐HP J test showed a high diagnosis performance, and is not inferior to ELISA. The l‐HP J test is useful, especially when screening for Helicobacter pylori infection, due to its high sensitivity.

Background and Aim l‐Menthol has smooth muscle‐relaxing and antiperistaltic effects. We examined its effectiveness against peristalsis resumption during endoscopic submucosal dissection (ESD) of gastric tumors. Methods We retrospectively examined clinical data of 485 patients (501 lesions) who underwent ESD for upper gastrointestinal tumors in 2017. We included 119 patients (127 lesions) in whom peristaltic movement resumed during ESD and l‐menthol was applied; 366 patients (374 lesions) without l‐menthol application were used as controls. Video recordings were reviewed to determine whether l‐menthol suppressed peristalsis resumption. Results In cases with l‐menthol application, 2 (2.9%), 36 (14.3%), and 89 (71.2%) lesions were found in the upper (U), middle (M), and lower (L) regions, respectively. In the control group, the corresponding values were 66 (17.6%), 215 (57.5%), and 93 (24.9%), respectively. l‐Menthol efficacy was observed in 116 of the 127 treated lesions (91.3%), over 90% of which were in the posterior wall of the U region, anterior wall and greater curvature of the M region, and anterior wall and lesser curvature of the L region. The most and least effective areas for l‐menthol application were the anterior wall of gastric antrum and posterior wall of the M region, respectively. The mean time from application to peristalsis inhibition was 8.7 s. No adverse effects were observed; perforation and secondary hemorrhage were not significantly different between the groups. Conclusion Direct l‐menthol application to the submucosal layer during mucosal resection affects smooth muscles and rapidly inhibits peristalsis resumption. Clinically, l‐Menthol can be used to suppress peristalsis recurrence during ESD, without adverse effects., Direct l‐menthol application to the submucosal layer during mucosal resection. When peristaltic resumption was seen, we sprayed l‐menthol on submucosal layer directly. Peristalsis stopped rapidly (8.7 seconds). l‐menthol was effective in 91.5% cases.

Diagnosis using artificial intelligence (AI) with deep learning could be useful in endoscopic examinations. We investigated the ability of AI to detect superficial esophageal squamous cell carcinoma (ESCC) from esophagogastroduodenoscopy (EGD) videos. We retrospectively collected 8428 EGD images of esophageal cancer to develop a convolutional neural network through deep learning. We evaluated the detection accuracy of the AI diagnosing system compared with that of 18 endoscopists. We used 144 EGD videos for the two validation sets. First, we used 64 EGD observation videos of ESCCs using both white light imaging (WLI) and narrow-band imaging (NBI). We then evaluated the system using 80 EGD videos from 40 patients (20 with superficial ESCC and 20 with non-ESCC). In the first set, the AI system correctly diagnosed 100% ESCCs. In the second set, it correctly detected 85% (17/20) ESCCs. Of these, 75% (15/20) and 55% (11/22) were detected by WLI and NBI, respectively, and the positive predictive value was 36.7%. The endoscopists correctly detected 45% (25–70%) ESCCs. With AI real-time assistance, the sensitivities of the endoscopists were significantly improved without AI assistance (p

Introduction: Curative rates of endoscopic treatment for undifferentiated-type early gastric cancer (EGC), particularly mixed poorly differentiated adenocarcinoma (MIXED-POR), are lower than those of endoscopic treatment for the differentiated type. Magnifying endoscopy with narrow-band imaging (ME-NBI) is useful for diagnoses of the histological type. This study aimed to investigate the detection rates of MIXED-POR among undifferentiated-type EGCs using biopsy and ME-NBI in order to improve curative rates through endoscopic treatment. Methods: We analyzed 267 lesions initially subjected to endoscopic submucosal resection (ESD) and histologically diagnosed as undifferentiated-type EGCs between July 2005 and December 2016 at our hospital. We obtained written informed consent from all participants. Biopsy and ME-NBI findings were compared to distinguish pure signet ring cell carcinoma (PURE-SIG) and MIXED-POR. ME-NBI findings were divided into 2 categories depending on the presence of irregular vessels. Results of biopsy and ME-NBI (combination method) were also analyzed, and detection rates of MIXED-POR and PURE-SIG were evaluated in terms of sensitivity, specificity, and accuracy. Results: Overall, 114 lesions were analyzed. Fifty-eight lesions (50.9%) were identified as MIXED-POR. With biopsy, the detection rate of MIXED-POR was significantly lower than that of PURE-SIG (p < 0.0001). ME-NBI detected significantly more MIXED-POR with irregular vessels than PURE-SIG (p < 0.0001). The combination method could detect significantly more MIXED-POR than PURE-SIG (p < 0.0001). The sensitivity and accuracy for MIXED-POR diagnosis were significantly higher with the combination method than with biopsy alone (p < 0.0001). Discussion/Conclusion: Combining biopsy and ME-NBI improved the accuracy of pretreatment diagnosis before ESD in undifferentiated-type cancer.

Image recognition using artificial intelligence (AI) has progressed significantly due to innovative technologies such as machine learning and deep learning. In the field of gastric cancer (GC) management, research on AI-based diagnosis such as anatomical classification of endoscopic images, diagnosis of Helicobacter pylori infection, and detection and qualitative diagnosis of GC is being conducted, and an accuracy equivalent to that of physicians has been reported. It is expected that AI will soon be introduced in the field of endoscopic diagnosis and management of gastric cancer as a supportive tool for physicians, thus improving the quality of medical care.

We aimed to clarify the long-term outcomes of endoscopic resection (ER) for early gastric cancers (EGCs) based on pathological curability in a multicenter prospective cohort study.We analyzed the long-term outcomes of 9054 patients with 10,021 EGCs undergoing ER between July 2010 and June 2012. Primary endpoint was the 5-year overall survival (OS). The hazard ratio for all-cause mortality was calculated using the Cox proportional hazards model. We also compared the 5-year OS with the expected one calculated for the surgically resected patients with EGC. If the lower limit of the 95% confidence interval (CI) of the 5-year OS exceeded the expected 5-year OS minus a margin of 5% (threshold 5-year OS), ER was considered to be effective. Pathological curability was categorized into en bloc resection, negative margins, and negative lymphovascular invasion: differentiated-type, pT1a, ulcer negative, ≤2 cm (Category A1); differentiated-type, pT1a, ulcer negative,2 cm or ulcer positive, ≤3 cm (Category A2); undifferentiated-type, pT1a, ulcer negative, ≤2 cm (Category A3); differentiated-type, pT1b (SM1), ≤3 cm (Category B); or noncurative resections (Category C).Overall, the 5-year OS was 89.0% (95% CI, 88.3%-89.6%). In a multivariate analysis, no significant differences were observed when the hazard ratio of Categories A2, A3, and B were compared with that of A1. In all the pathological curability categories, the lower limit of the 95% CI for the 5-year OS exceeded the threshold 5-year OS.ER can be recommended as a standard treatment for patients with EGCs fulfilling Category A2, A3, and B, as well as A1 (UMIN Clinical Trial Registry, UMIN000005871).

Endoscopists' abilities to diagnose early gastric cancers (EGCs) vary, especially between specialists and nonspecialists. We developed an artificial intelligence (AI)-based diagnostic support tool "Tango" to differentiate EGCs and compared its performance with that of endoscopists.The diagnostic performances of Tango and endoscopists (34 specialists, 42 nonspecialists) were compared using still images of 150 neoplastic and 165 non-neoplastic lesions. Neoplastic lesions included EGCs and adenomas. The primary outcome was to show the noninferiority of Tango (based on sensitivity) over specialists. The secondary outcomes were the noninferiority of Tango (based on accuracy) over specialists and the superiority of Tango (based on sensitivity and accuracy) over nonspecialists. The lower limit of the 95% confidence interval (CI) of the difference between Tango and the specialists for sensitivity was calculated, with-10% defined as noninferiority and0% defined as superiority in the primary outcome. The comparable differences between Tango and the endoscopists for each performance were calculated, with10% defined as superiority and0% defined as noninferiority in the secondary outcomes.Tango achieved superiority over the specialists based on sensitivity (84.7% vs. 65.8%, difference 18.9%, 95% CI 12.3-25.3%) and demonstrated noninferiority based on accuracy (70.8% vs. 67.4%). Tango achieved superiority over the nonspecialists based on sensitivity (84.7% vs. 51.0%) and accuracy (70.8% vs. 58.4%).The AI-based diagnostic support tool for EGCs demonstrated a robust performance and may be useful to reduce misdiagnosis.

The Japan Gastroenterological Endoscopy Society developed the Guideline for Endoscopic Diagnosis of Early Gastric Cancer based on scientific methods. Endoscopy for the diagnosis of early gastric cancer has been acknowledged as a useful and highly precise examination, and its use has become increasingly more common in recent years. However, the level of evidence in this field is low, and it is often necessary to determine recommendations based on expert consensus only. This clinical practice guideline consists of the following sections to provide the current guideline: [I] Risk stratification of gastric cancer before endoscopic examination, [II] Detection of early gastric cancer, [III] Qualitative diagnosis of early gastric cancer, [IV] Diagnosis to choose the therapeutic strategy for gastric cancer, [V] Risk stratification after endoscopic examination, and [VI] Surveillance of early gastric cancer.

Artificial intelligence (AI) that surpasses human ability in image recognition is expected to be applied in the field of gastrointestinal endoscopes. Accordingly, its research and development (RD) is being actively conducted. With the development of endoscopic diagnosis, there is a shortage of specialists who can perform high-precision endoscopy. We will examine whether AI with excellent image recognition ability can overcome this problem.Since 2016, papers on artificial intelligence using convolutional neural network (CNN in other word Deep Learning) have been published. CNN is generally capable of more accurate detection and classification than conventional machine learning. This is a review of papers using CNN in the gastrointestinal endoscopy area, along with the reasons why AI is required in clinical practice. We divided this review into four parts: stomach, esophagus, large intestine, and capsule endoscope (small intestine).Potential applications for the AI include colorectal polyp detection and differentiation, gastric and esophageal cancer detection, and lesion detection in capsule endoscopy. The accuracy of endoscopic diagnosis will increase if the AI and endoscopist perform the endoscopy together.

Management strategies for primary non-ampullary duodenal adenocarcinoma (NADAC) in early stage are not well established given its low incidence. This study aimed to elucidate clinicopathological features of early NADAC, including risk for lymph nodal metastasis (LNM). In total, 166 patients with early NADAC underwent initial treatment at our institution between 2006 and 2019, of whom 153 had intramucosal (M-) and 13 had submucosal (SM-) NADAC. These endoscopic and pathological features were retrospectively analyzed. Risk factors for LNM were evaluated in 46 early NADAC patients who underwent surgery with lymph node dissection. Compared with M-NADAC, SM-NADAC was significantly more frequently located at the proximal side of the papilla, with mixed elevated and depressed macroscopic type, histologically poorly differentiated tumor and lymphovascular invasion (LVI) (85% vs. 47%, P = 0.009; 54% vs. 5%, P

Background We previously reported for the first time the usefulness of artificial intelligence (AI) systems in detecting gastric cancers. However, the “original convolutional neural network (O-CNN)” employed in the previous study had a relatively low positive predictive value (PPV). Therefore, we aimed to develop an advanced AI-based diagnostic system and evaluate its applicability for the classification of gastric cancers and gastric ulcers. Methods We constructed an “advanced CNN” (A-CNN) by adding a new training dataset (4453 gastric ulcer images from 1172 lesions) to the O-CNN, which had been trained using 13 584 gastric cancer and 373 gastric ulcer images. The diagnostic performance of the A-CNN in terms of classifying gastric cancers and ulcers was retrospectively evaluated using an independent validation dataset (739 images from 100 early gastric cancers and 720 images from 120 gastric ulcers) and compared with that of the O-CNN by estimating the overall classification accuracy. Results The sensitivity, specificity, and PPV of the A-CNN in classifying gastric cancer at the lesion level were 99.0 % (95 % confidence interval [CI] 94.6 %−100 %), 93.3 % (95 %CI 87.3 %−97.1 %), and 92.5 % (95 %CI 85.8 %−96.7 %), respectively, and for classifying gastric ulcers were 93.3 % (95 %CI 87.3 %−97.1 %), 99.0 % (95 %CI 94.6 %−100 %), and 99.1 % (95 %CI 95.2 %−100 %), respectively. At the lesion level, the overall accuracies of the O- and A-CNN for classifying gastric cancers and gastric ulcers were 45.9 % (gastric cancers 100 %, gastric ulcers 0.8 %) and 95.9 % (gastric cancers 99.0 %, gastric ulcers 93.3 %), respectively. Conclusion The newly developed AI-based diagnostic system can effectively classify gastric cancers and gastric ulcers.

Objectives Detecting early gastric cancer is difficult, and it may even be overlooked by experienced endoscopists. Recently, artificial intelligence based on deep learning through convolutional neural networks (CNNs) has enabled significant advancements in the field of gastroenterology. However, it remains unclear whether a CNN can outperform endoscopists. In this study, we evaluated whether the performance of a CNN in detecting early gastric cancer is better than that of endoscopists. Methods The CNN was constructed using 13,584 endoscopic images from 2639 lesions of gastric cancer. Subsequently, its diagnostic ability was compared to that of 67 endoscopists using an independent test dataset (2940 images from 140 cases). Results The average diagnostic time for analyzing 2940 test endoscopic images by the CNN and endoscopists were 45.5 ± 1.8 s and 173.0 ± 66.0 min, respectively. The sensitivity, specificity, and positive and negative predictive values for the CNN were 58.4%, 87.3%, 26.0%, and 96.5%, respectively. These values for the 67 endoscopists were 31.9%, 97.2%, 46.2%, and 94.9%, respectively. The CNN had a significantly higher sensitivity than the endoscopists (by 26.5%; 95% confidence interval, 14.9-32.5%). Conclusion The CNN detected more early gastric cancer cases in a shorter time than the endoscopists. The CNN needs further training to achieve higher diagnostic accuracy. However, a diagnostic support tool for gastric cancer using a CNN will be realized in the near future.

With the application of artificial intelligence (AI) in deep learning, it has become possible to develop an AI that can be used clinically even in the field of upper endoscopy, which has been said to be difficult to diagnose. This review summarizes current studies on upper gastrointestinal tract based on AI and deep learning. At present, AI research on gastric cancer detection, Heliobactor pylori infection diagnosis, and esophagus cancer detection is progressing, and there is a possibility that AI can be used to assist in diagnosing invasion of depth of gastric caners and esophageal cancers. The studies reviewed show evidence that the use of AI in diagnosing cancer in the upper gastrointestinal tract as well as in the lower gastrointestinal tract, where research has been advanced, will be introduced to the clinical site in a form that contributes to detecting suspected cancer lesions, determining treatment policies, and improving examination accuracy.

Objectives The prognosis for pharyngeal cancer is relatively poor. It is usually diagnosed in an advanced stage. Although the recent development of narrow-band imaging (NBI) and increased awareness among endoscopists have enabled detection of superficial pharyngeal cancer, these techniques are still not prevalent worldwide. Nevertheless, artificial intelligence (AI)-based deep learning has led to significant advancements in various medical fields. Here, we demonstrate the diagnostic ability of AI-based detection of pharyngeal cancer from endoscopic images in esophagogastroduodenoscopy. Methods We retrospectively collected 5403 training images of pharyngeal cancer from 202 superficial cancers and 45 advanced cancers from the Cancer Institute Hospital, Tokyo, Japan. Using these images, we developed an AI-based diagnostic system with convolutional neural networks. We prepared 1912 validation images from 35 patients with 40 pharyngeal cancers and 40 patients without pharyngeal cancer to evaluate our system. Results Our AI-based diagnostic system correctly detected all pharyngeal cancer lesions (40/40) in the patients with cancer, including three small lesions smaller than 10 mm. For each image, the AI-based system correctly detected pharyngeal cancers in images obtained via NBI with a sensitivity of 85.6%, much higher sensitivity than that for images obtained via white light imaging (70.1%). The novel diagnostic system took only 28 s to analyze 1912 validation images. Conclusions The novel AI-based diagnostic system detected pharyngeal cancer with high sensitivity. It could facilitate early detection, thereby leading to better prognosis and quality of life for patients with pharyngeal cancers in the near future.

Introduction: We aimed to investigate the safety and efficacy of self-expandable metallic stent (SEMS) placement in patients with prior radiotherapy (RT) using the Niti-S stent, which is characterized by low radial force, in comparison to patients without prior RT. Methods: A consecutive series of 83 patients who were treated by SEMS placement using Niti-S stent for severe malignant esophageal obstruction or fistula were enrolled. The adverse event rates and efficacy were retrospectively compared between patients with/without prior RT before SEMS placement (RT group [n = 32] versus non-RT group [n = 51]). Results: The incidence rate of major adverse events in the RT group was 6.3% and was not significantly different from that in the non-RT group (5.9%, p = 0.95). Among the RT group, 84.4% were able to resume oral intake within a median of 2 days. Among the patients with fistula, 78.6% could resume oral intake and survive for 73 days after SEMS placement. Cox proportional hazard regression analysis identified significant factors affecting overall survival to be prior RT (hazard ratio [HR]: 1.96), low performance status (HR: 3.87), and subsequent anticancer treatment after SEMS placement (HR: 0.41). However, compared to the non-RT group, the RT group had received longer duration of anticancer treatment before SEMS placement and a lower rate of subsequent anticancer treatment after SEMS placement. Conclusions: With the Niti-S stent, the incidence of major adverse events was sufficiently low even for patients after RT. SEMS with low radial force would be an effective palliative treatment option for patients, regardless of prior RT.