Characterizing incidental mass lesions in abdominal dual-energy CT compared to conventional contrast-enhanced CT.

BACKGROUND: Incidental findings are common in abdominal computed tomography (CT) and often warrant further investigations with economic implications as well as implications for patients. PURPOSE: To evaluate the potential of dual-energy CT (DECT) in the identification and/or characterization of abdominal incidental mass lesions compared to conventional contrast-enhanced CT. MATERIAL AND METHODS: This retrospective study from a major tertiary hospital included 96 patients, who underwent contrast-enhanced abdominal DECT. Incidental lesions in adrenals, kidneys, liver, and pancreas were evaluated by two board-certified abdominal radiologists. Observer 1 only had access to standard CT reconstructions, while observer 2 had access to standard CT as well as DECT reconstructions. Disagreements were resolved by consensus review and used as a reference for observers using McNemar's test. RESULTS: Observers 1 and 2 identified a total of 40 and 34 findings, respectively. Furthermore, observer 1 registered 13 lesions requiring follow-up, of which seven (two renal and five adrenal lesions) were resolved following consensus review using DECT (P = 0.008). The inter-observer agreement was near perfect (κ = 0.82). CONCLUSION: DECT has the potential to improve the immediate characterization of incidental findings when compared to conventional CT for abdominal imaging.

COVID-19 Assessment with Bedside Lung Ultrasound in a Population of Intensive Care Patients Treated with Mechanical Ventilation and ECMO.

The COVID-19 pandemic has increased the need for an accessible, point-of-care and accurate imaging modality for pulmonary assessment. COVID-19 pneumonia is mainly monitored with chest X-ray, however, lung ultrasound (LUS) is an emerging tool for pulmonary evaluation. In this study, patients with verified COVID-19 disease hospitalized at the intensive care unit and treated with ventilator and extracorporal membrane oxygenation (ECMO) were evaluated with LUS for pulmonary changes. LUS findings were compared to C-reactive protein (CRP) and ventilator settings. Ten patients were included and scanned the day after initiation of ECMO and thereafter every second day until, if possible, weaned from ECMO. In total 38 scans adding up to 228 cineloops were recorded and analyzed off-line with the use of a constructed LUS score. The study indicated that patients with a trend of lower LUS scores over time were capable of being weaned from ECMO. LUS score was associated to CRP (R = 0.34; p < 0.03) and compliance (R = 0.60; p < 0.0001), with the strongest correlation to compliance. LUS may be used as a primary imaging modality for pulmonary assessment reducing the use of chest X-ray in COVID-19 patients treated with ventilator and ECMO.

The Performance of Deep Learning Algorithms on Automatic Pulmonary Nodule Detection and Classification Tested on Different Datasets That Are Not Derived from LIDC-IDRI: A Systematic Review.

The aim of this study was to systematically review the performance of deep learning technology in detecting and classifying pulmonary nodules on computed tomography (CT) scans that were not from the Lung Image Database Consortium and Image Database Resource Initiative (LIDC-IDRI) database. Furthermore, we explored the difference in performance when the deep learning technology was applied to test datasets different from the training datasets. Only peer-reviewed, original research articles utilizing deep learning technology were included in this study, and only results from testing on datasets other than the LIDC-IDRI were included. We searched a total of six databases: EMBASE, PubMed, Cochrane Library, the Institute of Electrical and Electronics Engineers, Inc. (IEEE), Scopus, and Web of Science. This resulted in 1782 studies after duplicates were removed, and a total of 26 studies were included in this systematic review. Three studies explored the performance of pulmonary nodule detection only, 16 studies explored the performance of pulmonary nodule classification only, and 7 studies had reports of both pulmonary nodule detection and classification. Three different deep learning architectures were mentioned amongst the included studies: convolutional neural network (CNN), massive training artificial neural network (MTANN), and deep stacked denoising autoencoder extreme learning machine (SDAE-ELM). The studies reached a classification accuracy between 68-99.6% and a detection accuracy between 80.6-94%. Performance of deep learning technology in studies using different test and training datasets was comparable to studies using same type of test and training datasets. In conclusion, deep learning was able to achieve high levels of accuracy, sensitivity, and/or specificity in detecting and/or classifying nodules when applied to pulmonary CT scans not from the LIDC-IDRI database.