Facilitating standardized COVID-19 suspicion prediction based on computed tomography radiomics in a multi-demographic setting.

OBJECTIVE: To develop an automatic COVID-19 Reporting and Data System (CO-RADS)-based classification in a multi-demographic setting. METHODS: This multi-institutional review boards-approved retrospective study included 2720 chest CT scans (mean age, 58 years [range 18-100 years]) from Italian and Russian patients. Three board-certified radiologists from three countries assessed randomly selected subcohorts from each population and provided CO-RADS-based annotations. CT radiomic features were extracted from the selected subcohorts after preprocessing steps like lung lobe segmentation and automatic noise reduction. We compared three machine learning models, logistic regression (LR), multilayer perceptron (MLP), and random forest (RF) for the automated CO-RADS classification. Model evaluation was carried out in two scenarios, first, training on a mixed multi-demographic subcohort and testing on an independent hold-out dataset. In the second scenario, training was done on a single demography and externally validated on the other demography. RESULTS: The overall inter-observer agreement for the CO-RADS scoring between the radiologists was substantial (k = 0.80). Irrespective of the type of validation test scenario, suspected COVID-19 CT scans were identified with an accuracy of 84%. SHapley Additive exPlanations (SHAP) interpretation showed that the "wavelet_(LH)_GLCM_Imc1" feature had a positive impact on COVID prediction both with and without noise reduction. The application of noise reduction improved the overall performance between the classifiers for all types. CONCLUSION: Using an automated model based on the COVID-19 Reporting and Data System (CO-RADS), we achieved clinically acceptable performance in a multi-demographic setting. This approach can serve as a standardized tool for automated COVID-19 assessment. KEYPOINTS: • Automatic CO-RADS scoring of large-scale multi-demographic chest CTs with mean AUC of 0.93 ± 0.04. • Validation procedure resembles TRIPOD 2b and 3 categories, enhancing the quality of experimental design to test the cross-dataset domain shift between institutions aiding clinical integration. • Identification of COVID-19 pneumonia in the presence of community-acquired pneumonia and other comorbidities with an AUC of 0.92.

Corticosteroid therapy for the management of paradoxical inflammatory reaction in patients with pulmonary tuberculosis.

BACKGROUND: Paradoxical reaction after the initiation of tuberculosis treatment is defined as increased inflammation following effective antimycobacterial treatment. This is a phenomenon that can severely complicate a patient's recovery, potentially leading to further morbidity and residual deficits. Paradoxical reaction remains poorly understood regarding its pathophysiology and management. Only a limited number of reports look critically at the available therapeutic options, with evidence of the efficacy of prednisolone therapy being primarily limited to extrapulmonary PR only. CASE: We describe two HIV negative patients who were admitted to our department with pulmonary tuberculosis, presenting with inflammatory patterns attributable to PR and their response to adjunctive steroid therapy. DISCUSSION AND CONCLUSIONS: The presented cases further highlight the need for immunological studies and randomized trials for corticosteroid therapy are needed to better understand this phenomenon as well as provide an evidence-base for anti-inflammatory treatment. Furthermore, by means of this case series, we are also able to highlight the potential variability in the symptomatology of the lesser known PR phenomenon, in which we observed a hypotensive shock-like syndrome not previously described in literature.

Effect of emphysema on AI software and human reader performance in lung nodule detection from low-dose chest CT.

BACKGROUND: Emphysema influences the appearance of lung tissue in computed tomography (CT). We evaluated whether this affects lung nodule detection by artificial intelligence (AI) and human readers (HR). METHODS: Individuals were selected from the "Lifelines" cohort who had undergone low-dose chest CT. Nodules in individuals without emphysema were matched to similar-sized nodules in individuals with at least moderate emphysema. AI results for nodular findings of 30-100 mm3 and 101-300 mm3 were compared to those of HR; two expert radiologists blindly reviewed discrepancies. Sensitivity and false positives (FPs)/scan were compared for emphysema and non-emphysema groups. RESULTS: Thirty-nine participants with and 82 without emphysema were included (n = 121, aged 61 ± 8 years (mean ± standard deviation), 58/121 males (47.9%)). AI and HR detected 196 and 206 nodular findings, respectively, yielding 109 concordant nodules and 184 discrepancies, including 118 true nodules. For AI, sensitivity was 0.68 (95% confidence interval 0.57-0.77) in emphysema versus 0.71 (0.62-0.78) in non-emphysema, with FPs/scan 0.51 and 0.22, respectively (p = 0.028). For HR, sensitivity was 0.76 (0.65-0.84) and 0.80 (0.72-0.86), with FPs/scan of 0.15 and 0.27 (p = 0.230). Overall sensitivity was slightly higher for HR than for AI, but this difference disappeared after the exclusion of benign lymph nodes. FPs/scan were higher for AI in emphysema than in non-emphysema (p = 0.028), while FPs/scan for HR were higher than AI for 30-100 mm3 nodules in non-emphysema (p = 0.009). CONCLUSIONS: AI resulted in more FPs/scan in emphysema compared to non-emphysema, a difference not observed for HR. RELEVANCE STATEMENT: In the creation of a benchmark dataset to validate AI software for lung nodule detection, the inclusion of emphysema cases is important due to the additional number of FPs. KEY POINTS: • The sensitivity of nodule detection by AI was similar in emphysema and non-emphysema. • AI had more FPs/scan in emphysema compared to non-emphysema. • Sensitivity and FPs/scan by the human reader were comparable for emphysema and non-emphysema. • Emphysema and non-emphysema representation in benchmark dataset is important for validating AI.

Detection and size quantification of pulmonary nodules in ultralow-dose versus regular-dose CT: a comparative study in COPD patients.

OBJECTIVE: To evaluate detectability and semi-automatic diameter and volume measurements of pulmonary nodules in ultralow-dose CT (ULDCT) vs regular-dose CT (RDCT). METHODS: Fifty patients with chronic obstructive pulmonary disease (COPD) underwent RDCT on 64-multidetector CT (120 kV, filtered back projection), and ULDCT on third-generation dual source CT (100 kV with tin filter, advanced modeled iterative reconstruction). One radiologist evaluated the presence of nodules on both scans in random order, with discrepancies judged by two independent radiologists and consensus reading. Sensitivity of nodule detection on RDCT and ULDCT was compared to reader consensus. Systematic error in semi-automatically derived diameter and volume, and 95% limits of agreement (LoA) were evaluated. Nodule classification was compared by κ statistics. RESULTS: ULDCT resulted in 83.1% (95% CI: 81.0-85.2) dose reduction compared to RDCT (p < 0.001). 45 nodules were present, with diameter range 4.0-25.3 mm and volume range 16.0-4483.0 mm3. Detection sensitivity was non-significant (p = 0.503) between RDCT 88.8% (95% CI: 76.0-96.3) and ULDCT 95.5% (95% CI: 84.9-99.5). No systematic bias in diameter measurements (median difference: -0.2 mm) or volumetry (median difference: -6 mm3) was found for ULDCT compared to RDCT. The 95% LoA for diameter and volume measurements were ±3.0 mm and ±33.5%, respectively. κ value for nodule classification was 0.852 for diameter measurements and 0.930 for volumetry. CONCLUSION: ULDCT based on Sn100 kV enables comparable detectability of solid pulmonary nodules in COPD patients, at 83% reduced radiation dose compared to RDCT, without relevant difference in nodule measurement and size classification. ADVANCES IN KNOWLEDGE: Pulmonary nodule detectability and measurements in ULDCT are comparable to RDCT.

The additional diagnostic value of virtual bronchoscopy navigation in patients with pulmonary nodules - The NAVIGATOR study.

BACKGROUND: The number of solitary pulmonary nodules to be evaluated is expected to increase and therefore we need to improve diagnostic and therapeutic tools to approach these nodules. To prevent patients from futile invasive procedures and receiving treatment without histological confirmation of cancer, we evaluated the value of virtual bronchoscopy navigation to obtain a diagnosis of the solitary pulmonary nodule in a real-world clinical setting. METHODS: In the NAVIGATOR single center, prospective, observational cohort study patients underwent a virtual bronchoscopy navigation procedure with or without guide sheet tunnelling to assess a solitary pulmonary nodule. Nodules were considered not accessible if a diagnosis could not be obtained by either by CT-guided transthoracic biopsy or conventional bronchoscopy. RESULTS: Between February 2021 and January 2022 35 patients underwent the virtual bronchoscopy navigation procedure. The overall diagnostic yield was 77% and was dependent on size of the nodule and chosen path, with highest yield in lesions with an airway path. Adverse events were few and manageable. CONCLUSION: Virtual bronchoscopy navigation with or without sheet tunnelling is a new technique with a good diagnostic yield, also in patients in whom previously performed procedures failed to establish a diagnosis and/or alternative procedures are considered not feasible based on expected yield and/or safety. Preventing futile or more invasive procedures like surgery or transthoracic punctures with a higher complication rate is beneficial for patients, and allowed treatment adaptation in two-third of the analyzed patient population.

Informatics in radiology: postprocessing pitfalls in using CT for automatic and semiautomatic determination of global left ventricular function.

Recent advances in technical capabilities of computed tomographic (CT) scanners, including an increasing number of detector rows, improved spatial and temporal resolution, and the development of retrospective gating, have allowed the acquisition of four-dimensional (4D) datasets of the beating heart. As a result, the heart can be visualized in different phases and CT datasets can be used to assess cardiac function. Many software packages currently exist that allow automatic or semiautomatic evaluation of left ventricular function on the basis of 4D CT datasets. The level of automation varies from extensive, completely manual segmentation by the user to fully automatic evaluation of left ventricular function without any user interaction. Although the reproducibility of functional parameter assessment is reported to be high and intersoftware variability low for larger groups of patients, significant differences can exist among measurements obtained with different software tools from the same dataset. Thus, careful review of automatically or semiautomatically obtained results is required.

Sarcoidosis presenting with glazy mucoid sputum and dyspnea: a case report.

BACKGROUND: Patients with pulmonary sarcoidosis commonly present with a dry cough; a productive cough suggests a complicating airway infection or an alternative diagnosis such as tuberculosis or bronchiectasis. CASE PRESENTATION: A 36-year-old European (Frisian) woman recently diagnosed with pulmonary sarcoidosis presented with debilitating exertional dyspnea and cough productive of glazy mucoid sputum. Several different attempts including video-assisted thoracoscopic biopsies failed to reach a second or alternative diagnosis including an infectious, autoimmune or collagen-vascular condition. She responded to steroids but with poor tolerance to this treatment, which could not be tapered. After she was started on anti-tumor necrosis factor alpha (TNF-α) therapy with infliximab, 200 mg at three-monthly intervals, she has been fine for well over a decade. CONCLUSIONS: In this patient with sarcoidosis who had a productive cough accompanied by fever, an extensive workup and prolonged follow-up, an alternative or second diagnosis could be ruled out; we therefore conclude that this highly unusual presentation is part of the clinical spectrum of sarcoidosis.

Evaluation of global left ventricular function assessment by dual-source computed tomography compared with MRI.

Left ventricular (LV) function assessment by dual-source computed tomography (DSCT) was compared with the reference standard method using magnetic resonance imaging (MRI). Accurate assessment of LV function is essential for the prediction of prognosis in cardiac disease. Thirty-four patients undergoing DSCT examination of the heart for various clinical indications underwent MRI after DSCT. Short-axis cine images were reconstructed from the DSCT datasets and were analyzed using a dedicated post-processing software-tool to generate global left ventricular function parameters. Five DSCT datasets were considered to be of insufficient image quality. DSCT showed a small overestimation of end-diastolic and end-systolic volumes of 11.0 ml and 3.5 ml, respectively. Myocardial mass assessed by DSCT showed an average underestimation of 0.2 g. DSCT showed a small overestimation of LV ejection fraction (LVEF) of 0.4%-point with a Bland-Altman interval of [-8.67 (0.40) 9.48]. Global LV functional parameters calculated from DSCT datasets acquired in daily clinical practice correlated well with MRI and may be considered interchangeable. However, visual assessment of the image quality of the short-axis cine slices should be performed to detect any artifacts in the DSCT data which could influence accuracy.

Accuracy of noninvasive coronary stenosis quantification of different commercially available dedicated software packages.

PURPOSE: The purpose of this study was to investigate the noninvasive quantification of coronary artery stenosis using cardiac software packages and vessel phantoms with known stenosis severity. MATERIALS AND METHODS: Four different sizes of vessel phantoms were filled with contrast agent and scanned on a 64-slice multidetector computed tomography. Diameter and area stenosis were evaluated by 2 observers blinded from the true measures using 5 different software packages. Measurements were compared with the true measure of the vessel phantoms. The absolute difference in stenosis measurements and intraobserver and interobserver variabilities were assessed. RESULTS: All software packages show a trend toward larger differences for the smaller vessel phantoms. The absolute difference of the automatic measurements was significantly higher compared with that of the manual measurements in all 5 evaluated software packages for all vessel phantoms (P < 0.05). CONCLUSION: Manual stenosis measurements are significantly more accurate compared with automatic measurements, and therefore, manual adjustments are still essential for noninvasive assessment of coronary artery stenosis.

Influence of lung nodule margin on volume- and diameter-based reader variability in CT lung cancer screening.

OBJECTIVE:: To evaluate the influence of nodule margin on inter- and intrareader variability in manual diameter measurements and semi-automatic volume measurements of solid nodules detected in low-dose CT lung cancer screening. METHODS:: 25 nodules of each morphological category (smooth, lobulated, spiculated and irregular) were randomly selected from 93 participants of the Dutch-Belgian Randomized Lung Cancer Screening Trial (NELSON). Semi-automatic volume measurements were performed using Syngo LungCARE® software (Version Somaris/5 VB10A-W, Siemens, Forchheim, Germany). Three radiologists independently measured mean diameters manually. Impact of nodule margin on interreader variability was evaluated based on systematic error and 95% limits of agreement. Interreader variability was compared with the nodule growth cut-off as used in Lung CT Screening Reporting and Data System (LungRADS; +1.5-mm diameter) and the Dutch-Belgian Randomized Lung Cancer Screening Trial(acronym: NELSON) /British Thoracic Society (+25% volume). RESULTS:: For manual diameter measurements, a significant systematic error (up to 1.2 mm) between readers was found in all morphological categories. For semi-automatic volume measurements, no statistically significant systematic error was found. The interreader variability in mean diameter measurements exceeded the 1.5-mm cut-off for nodule growth for all morphological categories [smooth: ±1.9 mm (+27%), lobulated: ±2.0 mm (+33%), spiculated: ±3.5 mm (+133%), irregular: ±4.5 mm (+200%)]. The 25% vol growth cut-off was exceeded slightly for spiculated [28% (+12%)] and irregular [27% (+8%)] nodules. CONCLUSION:: Lung nodule sizing based on manual diameter measurement is affected by nodule margin. Interreader variability increases especially for nodules with spiculated and irregular margins, and causes substantial misclassification of nodule growth. This effect is almost neglectable for semi-automated volume measurements. Semi-automatic volume measurements are superior for both size and growth determination of pulmonary nodules. ADVANCES IN KNOWLEDGE:: Nodule assessment based on manual diameter measurements is susceptible to nodule margin. This effect is almost neglectable for semi-automated volume measurements. The larger interreader variability for manual diameter measurement results in inaccurate lung nodule growth detection and size classification.

Coronary fly-through or virtual angioscopy using dual-source MDCT data.

Coronary fly-through or virtual angioscopy (VA) has been studied ever since its invention in 2000. However, application was limited because it requires an optimal computed tomography (CT) scan and time-consuming post-processing. Recent advances in post-processing software facilitate easy construction of VA, but until now image quality was insufficient in most patients. The introduction of dual-source multidetector CT (MDCT) could enable VA in all patients. Twenty patients were scanned using a dual-source MDCT (Definition, Siemens, Forchheim, Germany) using a standard coronary artery protocol. Post-processing was performed on an Aquarius Workstation (TeraRecon, San Mateo, Calif.). Length travelled per major branch was recorded in millimetres, together with the time required in minutes. VA could be performed in every patient for each of the major coronary arteries. The mean (range) length of the automated fly-through was 80 (32-107) mm for the left anterior descending (LAD), 75 (21-116) mm for the left circumflex artery (LCx), and 109 (21-190) mm for the right coronary artery (RCA). Calcifications and stenoses were visualised, as well as most side branches. The mean time required was 3 min for LAD, 2.5 min for LCx, and 2 min for the RCA. Dual-source MDCT allows for high quality visualisation of the coronary arteries in every patient because scanning with this machine is independent of the heart rate. This is clearly shown by the successful VA in all patients. Potential clinical value of VA should be determined in the near future.