Deep Learning-Based Automated Labeling of Coronary Segments for Structured Reporting of Coronary Computed Tomography Angiography in Accordance With Society of Cardiovascular Computed Tomography Guidelines.

PURPOSE: To evaluate a novel deep learning (DL)-based automated coronary labeling approach for structured reporting of coronary artery disease according to the guidelines of the Society of Cardiovascular Computed Tomography (CT) on coronary CT angiography (CCTA). PATIENTS AND METHODS: A retrospective cohort of 104 patients (60.3 ± 10.7 y, 61% males) who had undergone prospectively electrocardiogram-synchronized CCTA were included. Coronary centerlines were automatically extracted, labeled, and validated by 2 expert readers according to Society of Cardiovascular CT guidelines. The DL algorithm was trained on 706 radiologist-annotated cases for the task of automatically labeling coronary artery centerlines. The architecture leverages tree-structured long short-term memory recurrent neural networks to capture the full topological information of the coronary trees by using a two-step approach: a bottom-up encoding step, followed by a top-down decoding step. The first module encodes each sub-tree into fixed-sized vector representations. The decoding module then selectively attends to the aggregated global context to perform the local assignation of labels. To assess the performance of the software, percentage overlap was calculated between the labels of the algorithm and the expert readers. RESULTS: A total number of 1491 segments were identified. The artificial intelligence-based software approach yielded an average overlap of 94.4% compared with the expert readers' labels ranging from 87.1% for the posterior descending artery of the right coronary artery to 100% for the proximal segment of the right coronary artery. The average computational time was 0.5 seconds per case. The interreader overlap was 96.6%. CONCLUSIONS: The presented fully automated DL-based coronary artery labeling algorithm provides fast and precise labeling of the coronary artery segments bearing the potential to improve automated structured reporting for CCTA.

Improving robustness of automatic cardiac function quantification from cine magnetic resonance imaging using synthetic image data.

Although having been the subject of intense research over the years, cardiac function quantification from MRI is still not a fully automatic process in the clinical practice. This is partly due to the shortage of training data covering all relevant cardiovascular disease phenotypes. We propose to synthetically generate short axis CINE MRI using a generative adversarial model to expand the available data sets that consist of predominantly healthy subjects to include more cases with reduced ejection fraction. We introduce a deep learning convolutional neural network (CNN) to predict the end-diastolic volume, end-systolic volume, and implicitly the ejection fraction from cardiac MRI without explicit segmentation. The left ventricle volume predictions were compared to the ground truth values, showing superior accuracy compared to state-of-the-art segmentation methods. We show that using synthetic data generated for pre-training a CNN significantly improves the prediction compared to only using the limited amount of available data, when the training set is imbalanced.

Safety profile of adenosine stress cardiac MRI in a tertiary hospital in India.

BACKGROUND: Coronary artery disease (CAD) is a major cause of morbidity and mortality in India. Stress Cardiac Magnetic Resonance (CMR) using vasodilator agent is well established in assessing the functional significance of CAD. Adenosine is the preferred agent, but can have severe side effects including dyspnoea, chest pain, atrioventricular block or bronchospasm. The stress CMR examination is not routinely performed in many of the clinical imaging departments in India. OBJECTIVE: The aim of this study was to establish safety of adenosine as a pharmacological stressor agent for CMR in a tertiary care radiology department in India. METHODS: A review of all patients undergoing stress CMR in our institution from May 2018 to May 2019 was made. Records were reviewed to collect response parameters and documented adverse reactions. RESULTS: A total of 1057 patients underwent stress CMR during this period. No death, myocardial infarction or atrio-ventricular block related complications were seen. Transient hypotension was seen in 20 patients (1.8') with spontaneous recovery after stopping infusion. Chest pain and breathlessness severe enough to discontinue the scan were seen in 6 (0.5') and 10 (0.9') patients, respectively. All patients with breathlessness recovered on low flow oxygen therapy with three requiring bronchodilator. Out of six patients with chest pain, three had immediate relief with sublingual nitroglycerin, and three required hospital admission for unstable angina. Of the latter three, 1 underwent revascularization on the same day and other two later in the week. CONCLUSION: Stress CMR using adenosine in appropriately selected patients is a highly safe procedure with significant side effects seen in less than 1' of patients. Therefore, it is safe to perform stress CMR studies in a fully equipped and well-trained radiology department in India.

Quantification of Subcutaneous and Visceral Fat in Abdominal CT and the Correlation with Metabolic Parameters.

The objective of this study was to determine the association of Computed tomography quantified visceral (VAT) and subcutaneous adipose tissue (SAT) with estimated glomerular filtration rate (eGFR) using CKD-EPI formula, diabetes mellitus, body mass index (BMI), proteinuria, lipid profile, and hypertension. Increased adipose tissue is the primary phenotypic characteristic of obesity. The amount and distribution of adipose tissue is associated with many adverse consequences, as hypertension, type 2 diabetes, cardiovascular disease etc. Abdominal CT with semi-automated software can quantify adipose tissue and predict the risk for metabolic diseases. This cross-sectional study was carried out of 101 individuals (71 Males and 30 Females) with the mean age of 59.4 ± 5.2 years. Axial sections of non-contrast CT abdomen between L4-5 intervertebral disc (10 mm) were selected to quantify VAT and SAT utilizing GE advanced workstation software. We took 18.5-22.9 kg/ m2 as the normal BMI for Indian population is. Hypertension was defined as per JNC 8 guidelines. The mean eGFR among hypertensives (no of ppl) was 78.1 ± 38.1, compared to 106 ± 29.8 in normotensives (p<0.001). A trend was observed towards lower eGFR with higher VAT (r=-0.194, p= 0.052). The mean eGFR among diabetics (82.1 ± 38 ml/min) was lower compared with non-diabetics (104.7 ±31.9ml/min, p= 0.002). There is a significant increase in VAT with increasing age (p=0.003). In a sample of 72 individuals' with high TGL and low HDL, showed trend towards increased adipose tissue in subcutaneous and visceral compartment. BMI showed positive association with VAT (p<0.001) and SAT(p<0.001). CT quantification of adipose tissue can be used as a predictive tool to assess risk for metabolic diseases and decline in renal function. Further, it helps in early implementation of pharmacological or life-style based modification for better survival.

Exostosis Bursata - Multimodality Imaging Approach.

Osteochondromas or exostosis are common benign bone tumours, commonly arising from the metaphyseal region of long bones (femur, humerus, tibia). Osteochondroma of the scapula are rare and cause mechanical irritation leading to bursal formation. We hereby report a case of 30-year-old man who presented with painful chest wall swelling and its multimodality approach to establish the diagnosis.

CT and MR of recurrent primary cutaneous adenoid cystic carcinoma with multiple metastases.

Primary cutaneous adenoid cystic carcinoma is a rare slow-growing neoplasm, with limited literature reporting the involvement of the scalp. It has a tendency to recur locally; however, lymph node, distant pulmonary and bony metastases are exceptionally rare. We highlight the case of a 65-year-old female with primary cutaneous adenoid cystic carcinoma with distant pulmonary and bony metastases and the importance of imaging in diagnosing distant metastasis and perineural spread.

Dacron graft aneurysm with dissection.

Dacron grafts have been used as a conduit for large caliber arteries for many years successfully. However, these grafts can undergo complications such as aneurysm formation, rupture, and failure. Evaluation of these complications are of paramount importance because of its tendency to rupture and cause death. Imaging plays an important role in identifying and monitoring of these complications, and also provides a road map to the vascular surgeons for early intervention and revascularization.

Is there a genetic predisposition to new-onset diabetes after kidney transplantation?

Kidney transplant recipients may develop new-onset diabetes after transplantation (NODAT) and transplant-associated hyperglycemia (TAH) (NODAT or new-onset impaired glucose tolerance-IGT). We studied 251 consecutive renal transplant South Asian recipients for incidence of NODAT and its risk factors between June 2004 and January 2009. Pre-transplant glucose tolerance test (GTT) identified non-diabetics (n = 102, IGT-24, NGT-78) for analysis. Baseline immunosuppression along with either cyclosporine (CsA) (n = 70) or tacrolimus (Tac) (n = 32) was given. Patients underwent GTT 20 days (mean) post-transplant to identify NODAT, normal (N) or IGT. TAH was observed in 40.2% of the patients (40% in CsA and 40.6% in Tac) (P = 0.5). NODAT developed in 13.7% of the patients (12.9% in CsA and 15.6% in Tac) (P = 0.5). Overall, Hepatitis C (P = 0.007), human leukocyte antigen (HLA) B52 (P = 0.03) and lack of HLA A28 (A68/69) (P = 0.03) were associated with TAH. In the Tac group, higher Day 1 dosage (P <0.001), HLA A1 (P = 0.04), B13 (P = 0.03) and lack of DR2 (P = 0.004) increased the risk of TAH. In the CsA group, HLA A10 (P = 0.03), failure of triglyceride (P = 0.001) or low-density lipoprotein (LDL) (P = 0.03) to lower or high-density lipoprotein to rise (P = 0.001), and higher post-transplant LDL (P <0.001) and cholesterol levels (P = 0.02) were associated with NODAT or TAH. Post-transplant fasting plasma glucose on Day 1 had sensitivity-54.5%, specificity-50.1%, positive predictive value-18.1% and negative predictive value-84.8% for detecting NODAT. In conclusion, there is a genetic predisposition to NODAT and TAH in South Asia as seen by the HLA associations, and a predisposition exists to the individual diabetogenic effects of Tac and CsA based on HLA type. This could lead to more careful selection of calcineurin inhibitors based on HLA types in the South Asian population.

Renal transplantation in the elderly: South Indian experience.

In developing countries, renal transplantation is offered to young end-stage renal disease (ESRD) patients, while the older ones face limitations due to higher mortality risk. We retrospectively analyzed 225 patients who underwent renal transplantation from living donors, aged 40-60 years (Group A) and >60 years (Group B), focusing on their survival outcome. Group A (n = 181) had mean creatinine (mg/dL) 1.41 ± 0.84, 1.30 ± 0.65 and 1.40 ± 0.60 and mean eGFR (mL/min/1.73 m(2)) of 65.32 ± 23.03, 69.14 ± 32.65 and 59.21 ± 22.79 at 0, 3 and 6 months post-transplantation. Death-censored graft survival was 93.1% in first year followed by 91.2% in subsequent 4 years. Patient survival was 92.5% in first year, 90.7% in the next 2 years, and 89.2% in 4th year. Highest cumulative graft survival was 86.7% in the first year with 83.4%, 82.7% and 82.4% during the subsequent 3 years. Group B (n = 44) had mean creatinine (mg/dL) of 1.46 ± 1.02, 1.29 ± 0.23 and 1.2 ± 0.29 with a mean eGFR (mL/min/1.73 m(2)) of 67.90 ± 23.48, 67.02 ± 12.76 and 75.23 ± 15.19 at 0, 3 and 6 months. Highest death-censored graft survival was 97.4% in the first year with 94.7% in next 3 years. Patient survival was 88.1% throughout 4 years post-transplantation. Cumulative graft survival was 84.1% during 4 years. Biopsy-proven acute rejection rate was 28.7% in group A and 15.9% in group B (P = 0.058). There was higher mortality rate in group B with death mainly due to infections and cardiovascular complication. Cardiovascular risk assessment, pre-transplant cancer screening and judicious use of immunosuppressive agents should help minimize adverse events, balanced with an inherently reduced risk of acute rejection, hence the graft survival advantage and is the way forward to maximize patient and renal allograft survival in elderly patients.

Urinary tract infections in the era of newer immunosuppressant agents: a tertiary care center study.

We studied the incidence and the risk factors predisposing to post transplantation urinary tract infection (UTI) and the association with use of different immunosuppressive regimens. We performed a retrospective analysis of 152 recipients of renal transplantation over a period of two years. Seventy one (46.71%) patients had culture positive UTI, Escherichia coli (45.1%) being the commonest. Thirty four (22.39%) patients had acute rejection and 14.4% of those had suffered UTI in the early post transplant period. Immunosuppression included induction with various antibodies and maintenance on antirejection medications. Trimethoprim-sulphamethoxazole was given as prophylaxis throughout the period. The UTI was treated according to microbiological sensitivity. 2.8% died due to urosepsis. In our retrospective analysis renal transplant recipients under the age of 45, female gender and diabetics suffered more UTI. Combination therapy with micro-emulsion form of cyclosporine A, prednisolone and azathioprine developed more UTI (P= 0.0418).

Can serial eGFR, body mass index and smoking predict renal allograft survival in south Asian patients.

Limited data exist regarding long-term allograft survival in South Asian patients in the era of modern immunosuppressive therapy. This retrospective cohort study was undertaken to see the graft survival based on serial eGFR, immunosuppressive therapy, BMI and other confounding factors including smoking in patients who have undergone renal transplantation in a tertiary care center in south India. Three hundred and three kidney transplant recipients including live and cadaveric transplantation performed between 2001 and 2006 were included in this study. The mean graft survival after transplantation was 6.38 +/- 0.11 years, graft survival at one, two, three and five years were 95.7%, 92.72%, 91.72% and 89.21%, respectively. The mean serum creatinine and eGFR in the biopsy proven acute rejection (BPAR) group were 1.74 +/- 0.94 mg/dL and 43.73 +/- 13.65 mL/min com-pared with 1.24 +/- 0.59 mg/ dL and 61.50 +/- 17.40 mL/min in the non-BPAR group (P < 0.001 and P= 0.0159) respectively. The mean BMI in the BPAR group at one year was 26.59 +/- 3.18 kg/m 2 compared with 21.63 +/- 2.29 kg/m 2 in the non-BPAR group (P < 0.05). The mean graft survival in patients who were smokers at the time of pretransplant evaluation was 89.3% compared with 92.5% in the non-smokers (P=0.347). This retrospective cohort study found that serial eGFR, body mass index and smoking were significant predictors of graft survival following renal transplantation in South Asian patients.

How to make peritoneal dialysis affordable in developing countries.

Peritoneal dialysis (PD) is an underutilized renal replacement therapy in the developing world. It offers advantages of simplicity, reduced need of training, lack of dependence on infrastructure and location. The population is extremely underserved by healthcare and means to achieve it. PD is unavailable in many African nations. We explore the logistics of PD, domestic manufacture of PD fluid and accessories and ways to sustain it. Realization of local factors, ways to reduce peritonitis, reduced dosage in patients with residual renal function and use of generics to treat anemia that help improve the logistics. The role of national government especially in countries where dialysis is rationed and its lack of involvement leaving the billions to fetch for themselves is discussed. Innovative schemes by private insurers have improved PD outcome locally. These include subsidized once-in-a-lifetime PD treatment payment and industry sponsored nurse and technician visits to patients. Finally, the factors preventing nephrologists in delivering PD such as lack of training, reimbursement, infrastructure and affordability are discussed.