Application of an artificial intelligence system for endoscopic diagnosis of superficial esophageal squamous cell carcinoma.

BACKGROUND: Upper gastrointestinal endoscopy is critical for esophageal squamous cell carcinoma (ESCC) detection; however, endoscopists require long-term training to avoid missing superficial lesions. AIM: To develop a deep learning computer-assisted diagnosis (CAD) system for endoscopic detection of superficial ESCC and investigate its application value. METHODS: We configured the CAD system for white-light and narrow-band imaging modes based on the YOLO v5 algorithm. A total of 4447 images from 837 patients and 1695 images from 323 patients were included in the training and testing datasets, respectively. Two experts and two non-expert endoscopists reviewed the testing dataset independently and with computer assistance. The diagnostic performance was evaluated in terms of the area under the receiver operating characteristic curve, accuracy, sensitivity, and specificity. RESULTS: The area under the receiver operating characteristics curve, accuracy, sensitivity, and specificity of the CAD system were 0.982 [95% confidence interval (CI): 0.969-0.994], 92.9% (95%CI: 89.5%-95.2%), 91.9% (95%CI: 87.4%-94.9%), and 94.7% (95%CI: 89.0%-97.6%), respectively. The accuracy of CAD was significantly higher than that of non-expert endoscopists (78.3%, P < 0.001 compared with CAD) and comparable to that of expert endoscopists (91.0%, P = 0.129 compared with CAD). After referring to the CAD results, the accuracy of the non-expert endoscopists significantly improved (88.2% vs 78.3%, P < 0.001). Lesions with Paris classification type 0-IIb were more likely to be inaccurately identified by the CAD system. CONCLUSION: The diagnostic performance of the CAD system is promising and may assist in improving detectability, particularly for inexperienced endoscopists.

Endoscopic submucosal dissection for proximal duodenal subepithelial lesions: a retrospective cohort study.

BACKGROUND AND AIM: Endoscopic submucosal dissection (ESD) has been used to remove subepithelial lesions (SELs) in recent years; however, duodenal ESD is associated with high rates of immediate or delayed bleeding and perforation. Whether ESD can be recommended for the treatment of duodenal SELs remains controversial. Therefore, we evaluated the efficacy and safety of ESD for duodenal SELs. METHODS: We conducted a retrospective cohort study in 62 patients (62 lesions) who underwent ESD for duodenal SELs between January 2012 and December 2020. The therapeutic outcomes from ESD for duodenal SELs and procedure-related complications were analyzed. RESULTS: En bloc resection and complete resection rates associated with duodenal ESD were 90.3% and 100%, respectively; four patients had a positive microscopic margin on pathologic examination. The median procedure time was 45 min (range 20-106 min). During the procedure, two patients received emergency surgery for uncontrolled bleeding and perforation, respectively. After the procedure, delayed bleeding occurred in three patients (4.8%), which was successfully managed by clipping, and delayed perforation occurred in two patients (3.2%) and needed emergency surgery. Risk factors related to complications were analyzed. Lesion size was found to be significantly associated with the complications (P = 0.028). No recurrences were detected, and no distant metastasis was observed in any patient during a median follow-up period of 45.5 months (range, 6-103 months). CONCLUSION: Duodenal ESD is relatively safe and feasible for duodenal SELs, especially for lesions no more than 2 cm in size.

Shear Wave Elastography of Invasive Ductal Carcinoma: Correlations between Shear Wave Velocity and Histological Prognostic Factors.

The correlations between shear wave velocity (SWV) calculated from virtual touch tissue imaging quantification (VTIQ) technique and histological prognostic factors of invasive ductal carcinoma was investigated. A total of 76 breast tumors histologically confirmed as invasive ductal carcinomas were included in this study. SWV values were measured by VTIQ for each lesion preoperatively or prior to breast biopsy. The maximum values were recorded for statistical analysis. Medical records were reviewed to determine tumor size, histological grade, lymph node status and immunohistochemical results. Tumor subtypes were categorized as luminal A, luminal B, human epidermal growth factor receptor 2 (HER2) positive and triple negative. The correlations between SWV and histological prognostic factors were analyzed. It was found that tumor size showed positive association with SWV (r=0.465, P<0.001). Larger tumors had significantly higher SWV than smaller ones (P=0.001). Histological grade 1 tumors had significantly lower SWV values than those with higher histological grade (P=0.015). The Ki67 expression, tumor subtypes and lymph node status showed no statistically significant correlations with SWV, although triple negative tumors and lymph node-positive tumors showed higher SWV values. It was concluded that tumor size was significantly associated with SWV. Higher histological grade was associated with increased SWV. There was no statistically significant correlations between SWV and other histological prognostic factors.

Laser-induced interstitial thermotherapy via a single-needle delivery system: Optimal conditions of ablation, pathological and ultrasonic changes.

This study aimed to examine the optimal conditions of laser-induced interstitial thermotherapy (LITT) via a single-needle delivery system, and the ablation-related pathological and ultrasonic changes. Ultrasound (US)-guided LITT (EchoLaser system) was performed at the output power of 2-4 Wattage (W) for 1-10 min in ex vivo bovine liver. Based on the results of the ex vivo study, the output power of 3 and 4 W with different durations was applied to in vivo rabbit livers (n=24), and VX2 tumors implanted in the hind limbs of rabbits (n=24). The ablation area was histologically determined by hematoxylin-eosin (HE) staining. Traditional US and contrast enhanced ultrasound (CEUS) were used to evaluate the treatment outcomes. The results showed: (1) In the bovine liver, ablation disruption was grossly seen, including a strip-like ablation crater, a carbonization zone anteriorly along the fiber tip, and a surrounding gray-white coagulation zone. The coagulation area, 1.2 cm in length and 1.0 cm in width, was formed in the bovine liver subjected to the ablation at 3 W for 5 min and 4 W for 4 min, and it extended slightly with the ablation time. (2) In the rabbit liver, after LITT at 3 W for 3 min and more, the coagulation area with length greater than or equal to 1.2 cm, and width greater than or equal to 1.0 cm, was found. Similar coagulation area was seen in the implanted VX2 carcinoma at 3 W for 5 min. (3) Gross examination of the liver and carcinoma showed three distinct regions: ablation crater/carbonization, coagulation and congestion distributed from the center outwards. (4) Microscopy revealed four zones after LITT, including ablation crater/carbonization, coagulation, edema and congestion from the center outwards. A large area with coagulative necrosis was observed around a vessel in the peripheral area with edema and hyperemia. (5) The size of coagulation was consistent well to the CEUS findings. It was concluded that EchoLaser system at low power can produce a coagulation area larger than 1.0 cm×1.0 cm during a short time period. The real-time US imaging can be used to effectively guide and assess the treatment.