Search

BACKGROUND & AIMS: We aimed to develop and validate a deep-learning computer-aided detection (CAD) system, suitable for use in real time in clinical practice, to improve endoscopic detection of early neoplasia in patients with Barrett's esophagus (BE). METHODS: We developed a hybrid ResNet-UNet model CAD system using 5 independent endoscopy data sets. We performed pretraining using 494,364 labeled endoscopic images collected from all intestinal segments. Then, we used 1704 unique esophageal high-resolution images of rigorously confirmed early-stage neoplasia in BE and nondysplastic BE, derived from 669 patients. System performance was assessed by using data sets 4 and 5. Data set 5 was also scored by 53 general endoscopists with a wide range of experience from 4 countries to benchmark CAD system performance. Coupled with histopathology findings, scoring of images that contained early-stage neoplasia in data sets 2-5 were delineated in detail for neoplasm position and extent by multiple experts whose evaluations served as the ground truth for segmentation. RESULTS: The CAD system classified images as containing neoplasms or nondysplastic BE with 89% accuracy, 90% sensitivity, and 88% specificity (data set 4, 80 patients and images). In data set 5 (80 patients and images) values for the CAD system vs those of the general endoscopists were 88% vs 73% accuracy, 93% vs 72% sensitivity, and 83% vs 74% specificity. The CAD system achieved higher accuracy than any of the individual 53 nonexpert endoscopists, with comparable delineation performance. CAD delineations of the area of neoplasm overlapped with those from the BE experts in all detected neoplasia in data sets 4 and 5. The CAD system identified the optimal site for biopsy of detected neoplasia in 97% and 92% of cases (data sets 4 and 5, respectively). CONCLUSIONS: We developed, validated, and benchmarked a deep-learning computer-aided system for primary detection of neoplasia in patients with BE. The system detected neoplasia with high accuracy and near-perfect delineation performance. The Netherlands National Trials Registry, Number: NTR7072. ispartof: GASTROENTEROLOGY vol:158 issue:4 pages:915-+ ispartof: location:United States status: published

Background and study aims: Early neoplasia in Barrett’s esophagus is difficult to detect and often overlooked during Barrett’s surveillance. An automatic detection system could be beneficial, by assisting endoscopists with detection of early neoplastic lesions. The aim of this study was to assess the feasibility of a computer system to detect early neoplasia in Barrett’s esophagus.Patients and methods: Based on 100 images from 44 patients with Barrett’s esophagus, a computer algorithm, which employed specific texture, color filters, and machine learning, was developed for the detection of early neoplastic lesions in Barrett’s esophagus. The evaluation by one endoscopist, who extensively imaged and endoscopically removed all early neoplastic lesions and was not blinded to the histological outcome, was considered the gold standard. For external validation, four international experts in Barrett’s neoplasia, who were blinded to the pathology results, reviewed all images.Results: The system identified early neoplastic lesions on a per-image analysis with a sensitivity and specificity of 0.83. At the patient level, the system achieved a sensitivity and specificity of 0.86 and 0.87, respectively. A trade-off between the two performance metrics could be made by varying the percentage of training samples that showed neoplastic tissue.Conclusion: The automated computer algorithm developed in this study was able to identify early neoplastic lesions with reasonable accuracy, suggesting that automated detection of early neoplasia in Barrett’s esophagus is feasible. Further research is required to improve the accuracy of the system and prepare it for real-time operation, before it can be applied in clinical practice.

BACKGROUND & AIMS: Acute perforations of the gastrointestinal tract are rare, severe complications of endoscopy that usually require surgical repair. Endoscopic repair of perforations would reduce the need For surgeries; we evaluated the efficacy and safety of endoscopic closure of acute perforations of the gastrointestinal tract by using a new clip device. METHODS: We conducted a prospective, international, multicenter study of 36 consecutive patients (15 male) with acute iatrogenic perforations (5 esophageal, 6 gastric, 12 duodenal, and 13 colonic perforation). Endoscopic repair was performed by using the Over-the-Scope-Clip according to a standardized operating procedure. Primary end point was successful closure, which was determined as endoscopic successful closure without leakage (detected by water-soluble contrast x-ray analysis), and absence of adverse events within 30 days after the procedure. RESULTS: Immediate closure was endoscopically successful in 33 patients (92%). One patient developed an esophageal perforation while the cap was introduced, and in 2 patients the perforations did not close; these 3 patients were successfully treated with surgery. None of the patients had leakage of soluble contrast on the basis of contrast x-ray. One patient with a closed colonic perforation deteriorated clinically within 6 hours after the procedure. Despite surgery, the patient died within 36 hours. The remaining 32 patients had successful endoluminal closures; the overall success rate was 89% (95% confidence interval, 75%-96%). The mean endoscopic closure time was 5 minutes 44 seconds 4 minutes 15 seconds. CONCLUSIONS: The Over-the-Scope-Clip is effective for endoluminal closure of acute iatrogenic perforations. It allows patients to avoid surgery, and 89% of patients had successful closures without adverse events.

To evaluate the efficacy of radiofrequency ablation(RFA) combined with endoscopic resection(ER) for eradicating widespread early non-flat type esophageal squamous cell carcinoma (ESCC) and precancerous lesions.Retrospective analysis was performed on the clinical data of 4 patients with early non-flat type ESCC and precancerous lesions in January 2010 at the Cancer Institute and Hospital, Chinese Academy of Medical Sciences. Proportion of patients with histological complete response (CR) 3 months, 12 months to 5 years after operation and adverse events were observed.These 4 patients were all male, aged from 47 to 71 (mean age 62) years, including 2 of ESCC, 1 of HGIN, 1 of MGIN confirmed by pathology. USL length was 6-12 (mean 8.5) cm. Treatment area (TA) length was 8-14 (mean 10.5) cm. Three cases were 0-II a (mean length 2 cm), and 1 case 0-II c (mean length 4 cm). Lesions of 2 cases were complete cycle, and other 2 cases occupied 3/4 circumference. Four patients completed their operations successfully. Total operation time was 42-105 (mean 66.8) min, RFA time was 3-12 (mean 8.25) min, and ER time was 6-20 (10.25) min, without bleeding and perforation. The mean hospital stay was 3 days. Pathology examination showed that 2 cases were ESCC G2 (lesion length 12, 8 cm; non-flat type lesion length 3, 4 cm), 1 was HGIN (lesion length 12 cm; non-flat type lesion length 1 cm) and 1 was MGIN (lesion length 6 cm; non-flat type lesion length 2 cm). Three cases were CR 3 months, 1 to 5 years after operation. One case had HGIN at 3-month and MGIN at 1-year and 3-year during follow up, and was CR after treatment with HALO. Postoperative esophageal stenosis occurred in 4 cases. Among them, 2 cases were mild without treatment, and 2 were severe, who were relieved by endoscopic water sac dilation for 5-8 (mean 6.5) times.RFA combined with ER is effective and safe in the treatment of patients with early non-flat esophageal squamous cell carcinoma and precancerous lesions.