To evaluate the diagnostic accuracy of a deep learning (DL) algorithm predicting hemodynamically significant coronary artery disease (CAD) by using a rest dataset of myocardial computed tomography perfusion (CTP) as compared to invasive evaluation. One hundred and twelve consecutive symptomatic patients scheduled for clinically indicated invasive coronary angiography (ICA) underwent CCTA plus static stress CTP and ICA with invasive fractional flow reserve (FFR) for stenoses ranging between 30 and 80%. Subsequently, a DL algorithm for the prediction of significant CAD by using the rest dataset (CTP-DLrest) and stress dataset (CTP-DLstress) was developed. The diagnostic accuracy for identification of significant CAD using CCTA, CCTA + CTP stress, CCTA + CTP-DLrest, and CCTA + CTP-DLstress was measured and compared. The time of analysis for CTP stress, CTP-DLrest, and CTP-DLStress was recorded. Patient-specific sensitivity, specificity, NPV, PPV, accuracy, and area under the curve (AUC) of CCTA alone and CCTA + CTPStress were 100%, 33%, 100%, 54%, 63%, 67% and 86%, 89%, 89%, 86%, 88%, 87%, respectively. Patient-specific sensitivity, specificity, NPV, PPV, accuracy, and AUC of CCTA + DLrest and CCTA + DLstress were 100%, 72%, 100%, 74%, 84%, 96% and 93%, 83%, 94%, 81%, 88%, 98%, respectively. All CCTA + CTP stress, CCTA + CTP-DLRest, and CCTA + CTP-DLStress significantly improved detection of hemodynamically significant CAD compared to CCTA alone (p < 0.01). Time of CTP-DL was significantly lower as compared to human analysis (39.2 ± 3.2 vs. 379.6 ± 68.0 s, p < 0.001). Evaluation of myocardial ischemia using a DL approach on rest CTP datasets is feasible and accurate. This approach may be a useful gatekeeper prior to CTP stress..

Pierino Spadafora,1 Andrea Esposito,2 Caterina Giannitto,3 Letizia Di Meglio,1 Nunzio Paolo Nuzzi,3 Gianpaolo Carrafiello2 1Postgraduate School of Diagnostic and Interventional Radiology, University of Milan, Milan, 20122, Italy; 2Foundation IRCCS Ca’ Granda Maggiore Policlinico Hospital, Radiology Department, Milan, Italy; 3Humanitas Research Hospital, Radiology Department, Milan, 20013, ItalyCorrespondence: Pierino SpadaforaPostgraduate School of Diagnostic and Interventional Radiology, University of Milan, Via Festa del Perdono 7, Milan, 20122, ItalyEmail pierino.spadafora@unimi.itAndrea EspositoFoundation IRCCS Ca’ Granda Maggiore Policlinico Hospital, Radiology Department, Via F. Sforza 35, Milan, 20122, ItalyEmail andrea.esposito@policlinico.mi.itAbstract: Dysphonia is a frequent and often disabling condition that can be caused by a multitude of circumstances. Differential diagnosis of dysphonia comprehends many different etiologies and many causative agents (neoplasms, inflammations, traumatic injuries) that can occur in a large anatomical space (from the encephalic trunk to the upper mediastinum). It is fundamental to remember that vascular etiologies are responsible for some rare cases of dysphonia. In the radiological database of two urban academic hospitals, from 2012 to 2020, we sought patients who underwent CT or MRI for dysphonia, selecting only the ones with an underlying clinically confirmed vascular etiology. We present three emblematic cases with different vascular etiologies: a ductus arteriosus aneurysm, a left internal carotid artery dilatation, a laryngeal arteriovenous malformation (AVM). Vascular causes of dysphonia are rare, but especially in these cases an accurate and prompt diagnosis is fundamental, in particular considering that the underlying cardiovascular anomaly can often pose a higher risk for the patient than the hoarseness itself. Diagnostic imaging plays a fundamental role in detecting the most common causes of dysphonia but it is very important that radiologists take the vascular causes into account so as not to miss them and to obtain a correct diagnosis.Keywords: dysphonia, thoracic aorta, arteriovenous malformation, carotid artery, patent ductus arteriosus

Purpose Radiologic criteria for the diagnosis of primary graft dysfunction (PGD) after lung transplantation are nonspecific and can lead to misinterpretation. The primary aim of our study was to assess the interobserver agreement in the evaluation of chest X-rays (CXRs) for PGD diagnosis and to establish whether a specific training could have an impact on concordance rates. Secondary aim was to analyze causes of interobserver discordances. Material and methods We retrospectively enrolled 164 patients who received bilateral lung transplantation at our institution, between February 2013 and December 2019. Three radiologists independently reviewed postoperative CXRs and classified them as suggestive or not for PGD. Two of the Raters performed a specific training before the beginning of the study. A senior thoracic radiologist subsequently analyzed all discordant cases among the Raters with the best agreement. Statistical analysis to calculate interobserver variability was percent agreement, Cohen’s kappa and intraclass correlation coefficient. Results A total of 473 CXRs were evaluated. A very high concordance among the two trained Raters, 1 and 2, was found ( K = 0.90, ICC = 0.90), while a poorer agreement was found in the other two pairings (Raters 1 and 3: K = 0.34, ICC = 0.40; Raters 2 and 3: K = 0.35, ICC = 0.40). The main cause of disagreement (52.4% of discordant cases) between Raters 1 and 2 was the overestimation of peribronchial thickening in the absence of unequivocal bilateral lung opacities or the incorrect assessment of unilateral alterations. Conclusion To properly identify PGD, it is recommended for radiologists to receive an adequate specific training.