Background and study aim: We aimed to develop a convolutional neural network (CNN) model for detecting neoplastic lesions during real-time digital single-operator cholangioscopy (DSOC) and clinically validated the model through comparisons with DSOC experts and nonexpert endoscopists.
Patients and methods: In this two-stage study, we first developed and validated CNN1. Then, we performed a multicenter diagnostic trial to compare four DSOC experts and nonexperts against an improved model (CNN2). They classified the lesions in neoplastic and nonneoplastic in accordance with Carlos Robles-Medranda (CRM) and Mendoza disaggregated criteria. The final diagnosis of neoplasia was based on histopathology and twelve-month follow-up outcomes.
Results: In stage I, CNN2 achieved a mAP of 0.88, an IoU of 83.24%, and a total loss of 0.0975. For clinical validation, a total of 170 videos from newly included patients were analyzed with the CNN2. 50% of cases had neoplastic lesions. This model achieved significant accuracy values for neoplastic diagnosis cases, with a 90.5% sensitivity, 68.2% specificity, and 74.0% and 87.8% positive and negative predictive values, respectively. CNN2 model outperformed nonexpert #2 ([ROC-CRM: 0.657 vs ROC-CNN2: 0.794, P<0.05 and ROC-Mendoza: 0.657 vs ROC-CNN2: 0.794, P<0.05]), nonexpert #4 (ROC-CRM: 0.683 vs ROC-CNN2: 0.791, P<0.05), and expert #4 ([ROC-CRM: 0.755 vs ROC-CNN2: 0.848, P<0,05 and ROC-Mendoza: 0.753 vs ROC-CNN2: 0.848, P<0,05]).
Conclusions: The proposed CNN models distinguish neoplastic bile duct lesions with good accuracy and outperformed two nonexperts and one expert endoscopists.
Background: Chronic esophageal conditions (CEC) are associated with significant disease-related burden, disability, and costs. Health-related quality of life (HRQOL) constructs are intended to capture the physical, mental, social, and emotional aspects of a patient's life and how health status impacts these domains. The Northwestern Esophageal Quality of Life (NEQOL) can be used among esophageal diseases while maintaining sensitivity to specific conditions. We aimed to translate, cross-cultural adapt, and validate the NEQOL into Spanish.Methods: After language and cross-cultural adaptation, the NEQOL was applied to an outpatient clinic-based population in a single tertiary center. We analyzed the internal consistency, construct, criterion validity, and test-retest reliability of the questionnaire. The criterion validity was tested against the SF-12 questionnaire. Key Results: After completing the translation process, no item was considered problematic. A total of 385 patients were included in the validation study. The internal consistency (Cronbach's alpha) for the total NEQOL-S score was 0.89. The NEQOL-S questionnaire showed moderate test-retest reliability (ICC = 0.828; 95% CI 0.755-0.881; p < 0.001). Criterion validity showed good coherence when correlated with the SF-12 survey (R 2 = 0.538; 95% CI 0.491-0.585, p < 0.001).
Conclusions and Inferences:The translated and cross-culturally adapted NEQOL-S showed good psychometric properties that allow its use in Spanish-speaking patients suffering from CEC.
BACKGROUND
Endoscopic ultrasound (EUS) can detect small lesions throughout the digestive tract; however, it remains challenging to accurately identify malignancies with this approach. EUS elastography measures tissue hardness, by which malignant and nonmalignant pancreatic masses (PMs) and lymph nodes (LNs) can be differentiated. However, there is currently little information regarding the strain ratio (SR) cutoff in Hispanic populations.
AIM
To determine the diagnostic accuracy of EUS elastography for PMs and LNs with an SR cutoff value in Hispanics.
METHODS
A retrospective study of patients who underwent EUS elastography for PMs between December 2013 and December 2014. A qualitative (analysis of color maps) and quantitative (SR) analysis of PMs and their associated LNs was performed. The accuracy of EUS elastography in identifying malignant PMs and LNs and cutoff value for SR were analyzed. A PM and/or its associated LNs were considered malignant based on histopathological findings from fine-needle aspiration biopsy samples.
RESULTS
A sample of 121 patients was included, 45.4% of whom were female. 69 (57.0%) PMs were histologically malignant, with a median SR of 50.4
vs
33.0 for malignant
vs
nonmalignant masses (
P
< 0.001). EUS evaluation identified associated LNs in 43/121 patients (35.5%), in whom 22/43 (51.2%) patients had histologically confirmed malignant diagnosis, with a median SR of 30
vs
40 for malignant
vs
nonmalignant LNs (
P =
0.7182). In detecting malignancy in PMs, an SR cutoff value of > 21.5 yielded a sensitivity of 94.2%, while a cutoff value of > 121 yielded a specificity of 96.2.2%. There were significant differences in the Giovannini scores, a previously established elastic score system, between the patients grouped by their final histology results (
P
< 0.001). For LNs, SR cutoff values of > 14.0 and > 155 yielded a sensitivity of 90.9% and a specificity of 95.2%, respectively, in detecting malignancy.
CONCLUSION
EUS elastography is a helpful technique for the diagnosis of solid PMs and their associated LNs. The proposed SR cutoff values have a high sensitivity and specificity for the detection of malignancy.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.