Cluster of microcalcifications can be an early sign of breast cancer. In this paper, we propose a novel approach based on convolutional neural networks for the detection and segmentation of microcalcification clusters. In this work, we used 283 mammograms to train and validate our model, obtaining an accuracy of 99.99% on microcalcification detection and a false positive rate of 0.005%. Our results show how deep learning could be an effective tool to effectively support radiologists during mammograms examination.

Percutaneous left atrial appendage occlusion (LAAO) with the Amplatzer Cardiac Plug (ACP) emerged as a valid alternative in patients with a formal contraindication to oral anticoagulant therapy. Transoesophageal echocardiography (TEE), cardiac computed tomography angiography (CCTA), intracardiac echocardiography (ICE), and conventional cardiac angiography (CCA) are used to evaluate LAA diameters. The aim of our study was to compare pre- and intraprocedural imaging techniques in determining the correct selection of the device size, with a retrospective evaluation of the results obtained at post-procedural CCTA follow-up. Between September 2009 and July 2013, 66 consecutive patients underwent to LAAO with the ACP at our institution. Preoperative LAA evaluation was realized with TEE, CCTA, ICE, and CCA. Fifty-eight (58) patients underwent to post-procedural CCTA to confirm the LAA complete exclusion, the number and extent of the residual leaks, and the positioning of the device. LAA diameters measured by CCTA correlate with the diameters obtained with CCA and ICE, but they are sized slightly larger than the others. TEE has a lower correlation with every other imaging method and a likely tendency to underestimate. The distribution of the leaks and the positioning of the device in post-procedural CCTA show no substantial differences between the devices used greater or equal to the one selected with CCTA in terms of LAA exclusion. The sizing of the device decided using CCTA in the phase of maximum LAA expansion reduces the risk of high-flow leaks and device malposition due to undersizing.

Arterial hypertension is a main determinant of arterial remodelling and atherosclerosis. Coronary artery calcium score and carotid intima-media thickness are recognized indices of vascular remodelling. Established biohumoral markers for the diagnosis of atherosclerosis are still lacking in asymptomatic subjects with hypertension. We aimed to test the association of plasma N-terminal pro B-type natriuretic peptide concentrations with either coronary artery calcium score or carotid intima-media thickness in asymptomatic hypertensive subjects. We conducted a case–control study on 436 hypertensi.ve and 436 age/sex-matched normotensive subjects from the population of the Montignoso HEart and Lung Project, a community-based study of asymptomatic general population ≥45 years. Subjects underwent N-terminal pro B-type natriuretic peptide measurement, echocardiography and evaluation of coronary artery calcium score and carotid intima-media thickness. Hypertensive subjects had higher median coronary artery calcium score (60 (interquartile range, 30–112) vs. 15 (interquartile range 3–70) Agatson units, p = 0.007), carotid intima-media thickness (8.6 (interquartile range 7.5–9.1) vs. 7.9 (7.1–8.4) µm, p < 0.001) and indexed left ventricular mass (101 (interquartile range 82–126) vs. 87 (63–91) mg/m2, p = 0.03) than controls, with no differences in left ventricular ejection fraction, diameters, E/E′, left atrial area. N-terminal pro B-type natriuretic peptide concentrations were higher in hypertensive subjects with either coronary artery calcium score (p = 0.008) or carotid intima-media thickness >75th (p < 0.006) percentile and highest in combined coronary artery calcium score/carotid intima-media thickness >75th percentile (p = 0.021). In multivariable analysis, N-terminal pro B-type natriuretic peptide independently predicted either coronary artery calcium score or carotid intima-media thickness >75th percentile, but only in hypertensive subjects (odds ratio = 1.87, 95% confidence interval 1.30–2.74, p = 0.001 and odds ratio = 1.99, 95% confidence interval 1.43–2.76, p = 0.001). In asymptomatic subjects with hypertension, N-terminal pro B-type natriuretic peptide is a marker of hypertension-mediated preclinical vascular disease.