RESUMEN
BACKGROUND: Whether the sensitivity of Deep Learning (DL) models to screen chest radiographs (CXR) for CoVID-19 can approximate that of radiologists, so that they can be adopted and used if real-time review of CXRs by radiologists is not possible, has not been explored before. OBJECTIVE: To evaluate the diagnostic performance of a doctor-trained DL model (Svita_DL8) to screen for COVID-19 on CXR, and to compare the performance of the DL model with that of expert radiologists. MATERIALS AND METHODS: We used a pre-trained convolutional neural network to develop a publicly available online DL model to evaluate CXR examinations saved in .jpeg or .png format. The initial model was subsequently curated and trained by an internist and a radiologist using 1062 chest radiographs to classify a submitted CXR as either normal, COVID-19, or a non-COVID-19 abnormal. For validation, we collected a separate set of 430 CXR examinations from numerous publicly available datasets from 10 different countries, case presentations, and two hospital repositories. These examinations were assessed for COVID-19 by the DL model and by two independent radiologists. Diagnostic performance was compared between the model and the radiologists and the correlation coefficient calculated. RESULTS: For detecting COVID-19 on CXR, our DL model demonstrated sensitivity of 91.5%, specificity of 55.3%, PPV 60.9%, NPV 77.9%, accuracy 70.1%, and AUC 0.73 (95% CI: 0.86, 0.95). There was a significant correlation (r = 0.617, P = 0.000) between the results of the DL model and the radiologists' interpretations. The sensitivity of the radiologists is 96% and their overall diagnostic accuracy is 90% in this study. CONCLUSIONS: The DL model demonstrated high sensitivity for detecting COVID-19 on CXR. CLINICAL IMPACT: The doctor trained DL tool Svita_DL8 can be used in resource-constrained settings to quickly triage patients with suspected COVID-19 for further in-depth review and testing.
RESUMEN
AIMS AND OBJECTIVES: There is a paucity of data regarding the outcomes of Heart transplantation in children from the Indian subcontinent. The data of patients under the age of 18 undergoing an isolated heart transplantation was analyzed for patient clinical profiles and risk factors for early and medium-term mortality. Hospital mortality was defined as death within 90 days of transplantation and medium-term survival as follow up of up to 6 years. MATERIALS AND METHODS: A total of 97 patients operated between March 2014 and October 2019 were included in this study. Data was collected about their INTERMACS status, pulmonary vascular resistance, donor heart ischemic times, donor age, donor to recipient weight ratio and creatinine levels. RESULTS: The age range was from 1 to 18 with a mean of 10.6 ± 4.6 years. 67 % patients were in INTERMACS category 3 or less.12 children were on mechanical circulatory support at the time of transplant. The 90 day survival was 89 %. The risk factors for hospital mortality was lower INTERMACS category (odd's ratio 0.2143, P = 0.026), elevated creatinine (odd's ratio 5.42, P = 0.076) and elevated right atrial pressure (odd's ratio 1.19, P = 0.015). Ischemic time, pulmonary vascular resistance (PVR) and PVR index (PVRI) had no effect on 90 day survival. Kaplan Meier estimates for 5 year survival was 73 %. The medium term survival was affected by INTERMACS category (Hazard ratio 0.7, P = .078), donor age > 25 (Hazard ratio 1.6, P = 0.26) and raised serum creatinine values.(Hazard ratio 2.7, P = 0.012). All the survivors are in good functional class. CONCLUSIONS: Excellent outcomes are possible after heart transplantation in a pediatric population even in a resource constrained environment of a developing economy. More efforts are needed to promote pediatric organ donation and patients need to be referred in better INTERMACS category for optimal outcomes.