Validation of a deep learning, value-based care model to predict mortality and comorbidities from chest radiographs in COVID-19.

We validate a deep learning model predicting comorbidities from frontal chest radiographs (CXRs) in patients with coronavirus disease 2019 (COVID-19) and compare the model's performance with hierarchical condition category (HCC) and mortality outcomes in COVID-19. The model was trained and tested on 14,121 ambulatory frontal CXRs from 2010 to 2019 at a single institution, modeling select comorbidities using the value-based Medicare Advantage HCC Risk Adjustment Model. Sex, age, HCC codes, and risk adjustment factor (RAF) score were used. The model was validated on frontal CXRs from 413 ambulatory patients with COVID-19 (internal cohort) and on initial frontal CXRs from 487 COVID-19 hospitalized patients (external cohort). The discriminatory ability of the model was assessed using receiver operating characteristic (ROC) curves compared to the HCC data from electronic health records, and predicted age and RAF score were compared using correlation coefficient and absolute mean error. The model predictions were used as covariables in logistic regression models to evaluate the prediction of mortality in the external cohort. Predicted comorbidities from frontal CXRs, including diabetes with chronic complications, obesity, congestive heart failure, arrhythmias, vascular disease, and chronic obstructive pulmonary disease, had a total area under ROC curve (AUC) of 0.85 (95% CI: 0.85-0.86). The ROC AUC of predicted mortality for the model was 0.84 (95% CI,0.79-0.88) for the combined cohorts. This model using only frontal CXRs predicted select comorbidities and RAF score in both internal ambulatory and external hospitalized COVID-19 cohorts and was discriminatory of mortality, supporting its potential use in clinical decision making.

Predicting Prolonged Hospitalization and Supplemental Oxygenation in Patients with COVID-19 Infection from Ambulatory Chest Radiographs using Deep Learning.

RATIONALE AND OBJECTIVES: The clinical prognosis of outpatients with coronavirus disease 2019 (COVID-19) remains difficult to predict, with outcomes including asymptomatic, hospitalization, intubation, and death. Here we determined the prognostic value of an outpatient chest radiograph, together with an ensemble of deep learning algorithms predicting comorbidities and airspace disease to identify patients at a higher risk of hospitalization from COVID-19 infection. MATERIALS AND METHODS: This retrospective study included outpatients with COVID-19 confirmed by reverse transcription-polymerase chain reaction testing who received an ambulatory chest radiography between March 17, 2020 and October 24, 2020. In this study, full admission was defined as hospitalization within 14 days of the COVID-19 test for > 2 days with supplemental oxygen. Univariate analysis and machine learning algorithms were used to evaluate the relationship between the deep learning model predictions and hospitalization for > 2 days. RESULTS: The study included 413 patients, 222 men (54%), with a median age of 51 years (interquartile range, 39-62 years). Fifty-one patients (12.3%) required full admission. A boosted decision tree model produced the best prediction. Variables included patient age, frontal chest radiograph predictions of morbid obesity, congestive heart failure and cardiac arrhythmias, and radiographic opacity, with an internally validated area under the curve (AUC) of 0.837 (95% CI: 0.791-0.883) on a test cohort. CONCLUSION: Deep learning analysis of single frontal chest radiographs was used to generate combined comorbidity and pneumonia scores that predict the need for supplemental oxygen and hospitalization for > 2 days in patients with COVID-19 infection with an AUC of 0.837 (95% confidence interval: 0.791-0.883). Comorbidity scoring may prove useful in other clinical scenarios.

Evaluating the Patient with Right Lower Quadrant Pain.

Right lower quadrant pain is one of the most common indications for imaging evaluation of the abdomen in the emergency department setting. This article reviews important imaging findings associated with acute appendicitis as well as major differential considerations including: mesenteric adenitis, Meckel diverticulum, neutropenic colitis, right-sided diverticulitis, epiploic appendagitis, omental infarction, and inflammatory bowel diseaseRight lower quadrant pain is one of the most common indications for imaging evaluation of the abdomen in the emergency department setting. This article reviews important imaging findings associated with acute appendicitis as well as major differential considerations including: mesenteric adenitis, Meckel diverticulum, neutropenic colitis, right-sided diverticulitis, epiploic appendagitis, omental infarction, and inflammatory bowel disease.

Bowel Obstruction.

Small bowel obstruction and large bowel obstruction account for approximately 20% of cases of acute abdominal surgical conditions. The role of the radiologist is to answer several key questions: Is obstruction present? What is the level of the obstruction? What is the cause of the obstruction? What is the severity of the obstruction? Is the obstruction simple or closed loop? Is strangulation, ischemia, or perforation present? In this presentation, the radiologic approach to and imaging findings of patients with known or suspected bowel obstruction are presented.

MR imaging of the elbow in the injured athlete.

This article summarizes key MR imaging findings in common athletic elbow injuries including little leaguer's elbow, Panner disease, osteochondritis dissecans, olecranon stress fracture, occult fracture, degenerative osteophyte formation, flexor-pronator strain, ulnar collateral ligament tear, lateral ulnar collateral ligament and radial collateral ligament tear, lateral epicondylitis, medial epicondylitis, biceps tear, bicipitoradial bursitis, triceps tear, olecranon bursitis, ulnar neuropathy, posterior interosseous nerve syndrome, and radial tunnel syndrome. The article also summarizes important technical considerations in elbow MR imaging that enhance image quality and contribute to the radiologist's success.

The incidental cystic pancreas mass: a practical approach.

Technical advances in cross-sectional imaging have led to the discovery of incidental cystic pancreatic lesions in the oncology and non-oncology population that in the past remained undetected. These lesions have created a diagnostic and management dilemma for both clinicians and radiologists: should these lesions be ignored, watched, aspirated, or removed? In this review, recommendations concerning the assessment of the more common pancreatic cystic incidental lesions are presented.

That liver lesion on MDCT in the oncology patient: is it important?

Multidetector-row computed tomography (MDCT) has become the primary imaging test for the staging and follow-up of most malignancies that originate outside of the central nervous system. Technical advances in this imaging technique have led to significant improvement in the detection of metastatic disease to the liver. An unintended by-product of this improving diagnostic acumen is the discovery of incidental hepatic lesions in oncology patients that in the past remained undetected. These ubiquitous, incidentally identified hepatic lesions have created a management dilemma for both clinicians and radiologists: are these lesions benign or do they represent metastases? Naturally, the answer to this question has profound prognostic and therapeutic implications. In this review, guidelines concerning the diagnosis and management of some of the more common hepatic incidental lesions detected in patients with extrahepatic malignancies are presented.

Mucosal immunization of rhesus monkeys against respiratory syncytial virus subgroups A and B and human parainfluenza virus type 3 by using a live cDNA-derived vaccine based on a host range-attenuated bovine parainfluenza virus type 3 vector backbone.

Reverse genetics was used to develop a two-component, trivalent live attenuated vaccine against human parainfluenza virus type 3 (HPIV3) and respiratory syncytial virus (RSV) subgroups A and B. The backbone for each of the two components of this vaccine was the attenuated recombinant bovine/human PIV3 (rB/HPIV3), a recombinant BPIV3 in which the bovine HN and F protective antigens are replaced by their HPIV3 counterparts (48). This chimera retains the well-characterized host range attenuation phenotype of BPIV3, which appears to be appropriate for immunization of young infants. The open reading frames (ORFs) for the G and F major protective antigens of RSV subgroup A and B were each placed under the control of PIV3 transcription signals and inserted individually or in homologous pairs as supernumerary genes in the promoter proximal position of rB/HPIV3. The level of replication of rB/HPIV3-RSV chimeric viruses in the respiratory tract of rhesus monkeys was similar to that of their parent virus rB/HPIV3, and each of the chimeras induced a robust immune response to both RSV and HPIV3. RSV-neutralizing antibody titers induced by rB/HPIV3-RSV chimeric viruses were equivalent to those induced by infection with wild-type RSV, and HPIV3-specific antibody responses were similar to, or slightly less than, after infection with the rB/HPIV3 vector itself. This study describes a novel vaccine strategy against RSV in which vaccine viruses with a common attenuated backbone, specifically rB/HPIV3 derivatives expressing the G and/or F major protective antigens of RSV subgroup A and of RSV subgroup B, are used to immunize by the intranasal route against RSV and HPIV3, which are the first and second most important viral agents of pediatric respiratory tract disease worldwide.