RESUMO
BACKGROUND: Missing occult cancer lesions accounts for the most diagnostic errors in retrospective radiology reviews as early cancer can be small or subtle, making the lesions difficult to detect. Second-observer is the most effective technique for reducing these events and can be economically implemented with the advent of artificial intelligence (AI). AIM: To achieve appropriate AI model training, a large annotated dataset is necessary to train the AI models. Our goal in this research is to compare two methods for decreasing the annotation time to establish ground truth: Skip-slice annotation and AI-initiated annotation. METHODS: We developed a 2D U-Net as an AI second observer for detecting colorectal cancer (CRC) and an ensemble of 5 differently initiated 2D U-Net for ensemble technique. Each model was trained with 51 cases of annotated CRC computed tomography of the abdomen and pelvis, tested with 7 cases, and validated with 20 cases from The Cancer Imaging Archive cases. The sensitivity, false positives per case, and estimated Dice coefficient were obtained for each method of training. We compared the two methods of annotations and the time reduction associated with the technique. The time differences were tested using Friedman's two-way analysis of variance. RESULTS: Sparse annotation significantly reduces the time for annotation particularly skipping 2 slices at a time (P < 0.001). Reduction of up to 2/3 of the annotation does not reduce AI model sensitivity or false positives per case. Although initializing human annotation with AI reduces the annotation time, the reduction is minimal, even when using an ensemble AI to decrease false positives. CONCLUSION: Our data support the sparse annotation technique as an efficient technique for reducing the time needed to establish the ground truth.
RESUMO
CONTEXT: COVID-19 caused a worldwide pandemic, and there are still many uncertainties about the disease. C-reactive protein (CRP) levels could be utilized as a prognosticator for disease severity in COVID-19 patients. OBJECTIVES: This study aims to determine whether CRP levels are correlated with COVID-19 patient outcomes and length of stay (LoS). METHODS: A retrospective cohort study was conducted utilizing data obtained between March and May 2020. Data were collected by abstracting past medical records through electronic medical records at 10 hospitals within CommonSpirit Health. Patients were included if they had a positive COVID-19 test from a nasopharyngeal swab sample, and if they were admitted and then discharged alive or had in-hospital mortality and were ≥18 years. A total of 541 patients had CRP levels measured and were included in this report. Patient outcome and LoS were the endpoints measured. RESULTS: The 541 patients had their CRP levels measured, as well as the demographic and clinical data required for analysis. While controlling for body mass index (BMI), number of comorbidities, and age, the first CRP was significantly predictive of mortality (p<0.001). The odds ratio for first CRP indicates that for each one-unit increase in CRP, the odds of death increased by 0.007. For LoS, the first CRP was a significant predictor (p<0.001), along with age (p=0.002). The number of comorbidities also predicted LoS (p=0.007), but BMI did not. The coefficient for the first CRP indicates that, for each one-unit increase in CRP, LoS increased 0.003 days. CONCLUSIONS: The results indicate that there is a positive correlation between the CRP levels of COVID-19 patients and their respective outcomes with regard to death and LoS.
