RESUMO
Artificial intelligence (AI) has been found to assist in optical differentiation of hyperplastic and adenomatous colorectal polyps. We investigated whether AI can improve the accuracy of endoscopists' optical diagnosis of polyps with advanced features. We introduced our AI system distinguishing polyps with advanced features with more than 0.870 of accuracy in the internal and external validation datasets. All 19 endoscopists with different levels showed significantly lower diagnostic accuracy (0.410-0.580) than the AI. Prospective randomized controlled study involving 120 endoscopists into optical diagnosis of polyps with advanced features with or without AI demonstration identified that AI improved endoscopists' proportion of polyps with advanced features correctly sent for histological examination (0.960 versus 0.840, p < 0.001), and the proportion of polyps without advanced features resected and discarded (0.490 versus 0.380, p = 0.007). We thus developed an AI technique that significantly increases the accuracy of colorectal polyps with advanced features.
RESUMO
BACKGROUND & AIMS: Infection with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) causes coronavirus disease 2019 (COVID-19), which has been characterized by fever, respiratory, and gastrointestinal symptoms as well as shedding of virus RNA into feces. We performed a systematic review and meta-analysis of published gastrointestinal symptoms and detection of virus in stool and also summarized data from a cohort of patients with COVID-19 in Hong Kong. METHODS: We collected data from the cohort of patients with COVID-19 in Hong Kong (N = 59; diagnosis from February 2 through February 29, 2020),and searched PubMed, Embase, Cochrane, and 3 Chinese databases through March 11, 2020, according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. We analyzed pooled data on the prevalence of overall and individual gastrointestinal symptoms (loss of appetite, nausea, vomiting, diarrhea, and abdominal pain or discomfort) using a random effects model. RESULTS: Among the 59 patients with COVID-19 in Hong Kong, 15 patients (25.4%) had gastrointestinal symptoms, and 9 patients (15.3%) had stool that tested positive for virus RNA. Stool viral RNA was detected in 38.5% and 8.7% among those with and without diarrhea, respectively (P = .02). The median fecal viral load was 5.1 log10 copies per milliliter in patients with diarrhea vs 3.9 log10 copies per milliliter in patients without diarrhea (P = .06). In a meta-analysis of 60 studies comprising 4243 patients, the pooled prevalence of all gastrointestinal symptoms was 17.6% (95% confidence interval [CI], 12.3-24.5); 11.8% of patients with nonsevere COVID-19 had gastrointestinal symptoms (95% CI, 4.1-29.1), and 17.1% of patients with severe COVID-19 had gastrointestinal symptoms (95% CI, 6.9-36.7). In the meta-analysis, the pooled prevalence of stool samples that were positive for virus RNA was 48.1% (95% CI, 38.3-57.9); of these samples, 70.3% of those collected after loss of virus from respiratory specimens tested positive for the virus (95% CI, 49.6-85.1). CONCLUSIONS: In an analysis of data from the Hong Kong cohort of patients with COVID-19 and a meta-analysis of findings from publications, we found that 17.6% of patients with COVID-19 had gastrointestinal symptoms. Virus RNA was detected in stool samples from 48.1% patients, even in stool collected after respiratory samples had negative test results. Health care workers should therefore exercise caution in collecting fecal samples or performing endoscopic procedures in patients with COVID-19, even during patient recovery.
