Deep blue meets scarred lung: computer- based analysis of a fibrosing lung disease

Deep blue meets scarred lung: computer- based analysis of a fibrosing lung disease

S2 Abstract / Clinical Radiology 71 (2016) S1eS10 Purpose: The purpose of this article is to describe incidental extramammary findings identified on b...

42KB Sizes 0 Downloads 5 Views


Abstract / Clinical Radiology 71 (2016) S1eS10

Purpose: The purpose of this article is to describe incidental extramammary findings identified on breast magnetic resonance imaging (MRI). Methods and materials: A retrospective review of 545 breast MRI performed between 2013e2015 was conducted across two district general hospitals. Each hospital used 1.5T MR units. Results: 542/545 (99%) were female. 37/545 (7%) patients had incidental findings. In this group, 37/37 (100%) were female with an average age of 56.5 years. 26/37 (70%) had no history of cancer. 7/37 (19%) had previous breast cancer. 4/37 (11%) were undergoing breast cancer treatment at the time of the MRI scan. Benign extramammary findings were common (33/37; 89%). The liver was major site with frequently reported simple cysts (28/37; 76%) and haemangioma (2/37; 3%). Eight MRI reports of ’high intensity focus’ in the liver were not followed-up with advised ultrasonography clarification. A multi-nodular goitre was referred to the ear, nose and throat (ENT) department and monitored. A 1.2 cm focus was noted in the spleen, which was possibly a simple cyst. 3/37 (8%) findings were significant. Two new breast cancer recurrences with pulmonary metastatic spread were found (2/37; 5%). They were referred to oncology. One non-toxic multi0nodular thyroid goitre was referred to ENT and had a total thyroidectomy. Conclusion: Incidental extramammary MRI findings are, overall, uncommon. However, when found, they should warrant further follow up. Three patients in our series had significant findings, which resulted in further treatment or a new diagnosis. Two of these patients had a previous breast cancer history. Hence, careful scrutiny for additional pathology outside the breast and the axilla is important and findings should be further clarified.

Purpose: The sophistication of computed tomography (CT) quantitation using computers is now comparable to visual CT evaluation. We compared a novel computer algorithm (computer aided lung informatics for pathology evaluation and rating [CALIPER]), with semi-quantitative visual scoring to examine discordances in evaluating disease extent in patients with idiopathic pulmonary fibrosis (IPF). Mortality prediction using CALIPER, visual scores and pulmonary function indices (PFI) were compared. Methods and materials: 283 consecutive patients with a multidisciplinary IPF diagnosis had various CT parenchymal patterns evaluated visually by two experienced thoracic radiologists and CALIPER: ground glass opacity (GGO), reticular pattern (RET), honeycombing and emphysema. CALIPER alone quantified pulmonary vessel volume (PVV); visual scores evaluated traction bronchiectasis. Results: Fibrosis-related volume loss resulted in underestimation of disease extent using CALIPERs volumetric scoring method relative to lobar visual scores. Univariate Cox analysis predictors of mortality: CALIPER: GGO (HR 1.03; p <0.0001), RET (HR 1.10; p <0.0001), honeycombing (HR 1.11; p¼0.006), PVV (HR 1.52; p <0.0001) VISUAL: RET (HR 1.02; p¼0.002), honeycombing (HR 1.03; p <0.0001), traction bronchiectasis (HR 1.11; p <0.0001) PFI: composite physiologic index (CPI) (HR 1.07; p ¼<0.001) Multivariate Cox regression analysis: CPI (HR 1.05; p <0.0001), and CALIPER-derived emphysema (HR 1.06; p <0.0001), honeycombing (HR 1.16; p¼0.005) and PVV (HR 1.27; p <0.0001) were independent predictors of mortality. Conclusion: Volumetric scores may underestimate fibrosis extent. Optimal parameters for prognostication in IPF are the CPI and novel CALIPER-derived (not visually scored) CT features. The PVV, a new CALIPER index with no visual correlate, is the strongest CT predictor of mortality in IPF.

Cardiac MR assessment of lipid-containing cardiac pathologies

Spectrum of viral infections in post-transplant lung

Authors: Katharine Alison Grace Garfath-Cox, Stephen Harden, Caroline Illingworth, James Shambrook

Authors: Samavia Raza, Dalal Paras, Evangelos Skondras

A retrospective study of 545 breast magnetic resonance imaging: extramammary incidental findings Authors: Michaella Cameron, David Howlett

Purpose: Some cardiac abnormalities contain fat. Identification of this and its precise location can permit an accurate diagnosis to be made and appropriate patient management to be recommended. Cardiac magnetic resonance (CMR) provides anatomical, functional and accurate tissue characterisation of these fat containing pathologies. Some of these abnormalities are discrete lesions such as lipomas. Other pathologies show a different form of fat distribution, such as intramyocardial lipid accumulation in Fabry’s disease and recently epicardial fat has been proposed as an indicator of risk for adverse cardiac events. Methods and materials: A literature review and search of our radiology information system (RIS) for fat containing cardiac abnormalities identified on CMR was performed. Results: Our review identified typical imaging examples of lipomas, lipomatous hypertrophy of the interatrial septum, fatty replacement of infarcts and arrhythmogenic right ventricular cardiomyopathy (ARVC). We also identified a case of Fabry’s disease identified on CMR and examples of patients with significant epicardial fat. Conclusion: The full range of CMR imaging sequences, including the recently developed sequences of T1 mapping, permits an accurate tissue characterisation of these abnormalities without the need for invasive diagnostic procedures. Deep blue meets scarred lung: computer- based analysis of a fibrosing lung disease Authors: Joseph Jacob, Brian Bartholmai, Ronald Karwoski, Maria Kokosi, Arjun Nair, Srinivasan Rajagopalan, Simon Walsh, Athol Wells, David Hansell

Purpose: Infections remains a significant cause of morbidity and mortality among lung transplant recipients and initially may present with equivocal clinical findings and non-specific impaired pulmonary function tests. High-resolution computed tomography (HRCT) scan is a sensitive test for early detection of infective lung process. Methods and materials: Out of 113 lung transplant follow up patients seven had virology proven infection with HRCT scan within five days. Patients were separated by the specific virus detected and their CT scan appearances were reviewed retrospectively for presence or absence of various features such as ground glass opacities, centrilobular nodules, consolidation, septal and peribronchial thickening, and pleural effusions to see if a specific appearance could be attributed to a specific virus in these patients. Results: Lower lobe predominance was common to every viral infection. Features of mucus plugging or tree in bud opacities were the second most common feature seen in all viruses except metapneumovirus infection. Ground glass opacities were absent in cytomegalovirus infection. Features of consolidation were seen in respiratory syncytial virus (RSV), metapneumovirus and in one of the cases of coronavirus. Lympadenopathy was only seen in one of the cases of coronavirus. While the number of patients per virus are too few to form a definitive correlation of type of virus and radiological features, subtle differences are present which warrant further study for a larger cohort of patients to clarify results. Conclusion: All proven cases of viral infection showed HRCT findings, early scan may be useful to suggest the diagnosis and guide the management of these patients. Diagnostic strategies should be broad, attempting to screen for all known viruses.