COVID-19 chest X-ray detection through blending ensemble of CNN snapshots.

The novel COVID-19 pandemic, has effectively turned out to be one of the deadliest events in modern history, with unprecedented loss of human life, major economic and financial setbacks and has set the entire world back quite a few decades. However, detection of the COVID-19 virus has become increasingly difficult due to the mutating nature of the virus, and the rise in asymptomatic cases. To counteract this and contribute to the research efforts for a more accurate screening of COVID-19, we have planned this work. Here, we have proposed an ensemble methodology for deep learning models to solve the task of COVID-19 detection from chest X-rays (CXRs) to assist Computer-Aided Detection (CADe) for medical practitioners. We leverage the strategy of transfer learning for Convolutional Neural Networks (CNNs), widely adopted in recent literature, and further propose an efficient ensemble network for their combination. The DenseNet-201 architecture has been trained only once to generate multiple snapshots, offering diverse information about the extracted features from CXRs. We follow the strategy of decision-level fusion to combine the decision scores using the blending algorithm through a Random Forest (RF) meta-learner. Experimental results confirm the efficacy of the proposed ensemble method, as shown through impressive results upon two open access COVID-19 CXR datasets - the largest COVID-X dataset, as well as a smaller scale dataset. On the large COVID-X dataset, the proposed model has achieved an accuracy score of 94.55% and on the smaller dataset by Chowdhury et al., the proposed model has achieved a 98.13% accuracy score.

COFE-Net: An ensemble strategy for Computer-Aided Detection for COVID-19.

Biomedical images contain a large volume of sensor measurements, which can reveal the descriptors of the disease under investigation. Computer-based analysis of such measurements helps detect the disease, and thereby swiftly aid medical professionals to choose adequate therapy. In this paper, we propose a robust deep learning ensemble framework known as COVID Fuzzy Ensemble Network, or COFE-Net. This strategy is proposed for the task of COVID-19 screening from chest X-rays (CXR) and CT Scans, as a part of Computer-Aided Detection (CADe) for medical practitioners. We leverage the strategy of Transfer Learning for Convolutional Neural Networks (CNNs) widely adopted in recent literature, and further propose an efficient ensemble network for their combination. The principles of fuzzy logic have been leveraged to combine the measured decision scores generated by three state-of-the-art CNNs - Inception V3, Inception ResNet V2 and DenseNet 201 - through the Choquet fuzzy integral. Experimental results support the efficacy of our approach over empirical ensembling, as the fuzzy ensembling strategy for biomedical measurement consists of dynamic refactoring of the classifier ensemble weights on the fly, based upon the confidence scores for coalitions of inputs. This is the chief advantage of our biomedical measurement strategy over others as other methods do not adjust to the multiple generated measurements dynamically unlike ours.Impressive results on multiple datasets demonstrate the effectiveness of the proposed method. The source code of our proposed method is made available at: https://github.com/theavicaster/covid-cade-ensemble.

Causes of hospitalization in patients on maintenance hemodialysis with arteriovenous fistula in a tertiary care hospital in West Bengal, India.

Patients with renal failure and on maintenance hemodialysis (HD) have a higher propensity toward cardiovascular and infectious diseases. The aim of this study was to find the causes for hospital admission in patients suffering from kidney failure and on maintenance HD. This cross-sectional, observational study was conducted in a tertiary care hospital in West Bengal, India, from January to December, 2015. Patients with chronic kidney disease stage 5 for more than one year and on HD with arteriovenous fistula admitted for other than HD were included in the study. Days of hospital stay and current diagnosis were stored for further analysis. Data were expressed in mean, standard deviation, percentage, and frequency. All the statistical tests were carried out in GraphPad prism 6.01. Data of total 49 (30 male, 19 female) patients with mean age 55.8 ± 10.98 years (range 27-75 years) were analyzed. Eighteen (36.73%) and 48 (97.96%) patients were suffering from type 2 diabetes mellitus and hypertension (HTN), respectively. Average stay in hospital was 10.31 ± 6.07 days (range: 5-43 days). Most common causes for hospitalization were left ventricular failure (LVF) (59.18%) followed by respiratory tract infection (RTI) (14.29%). In patients with renal failure receiving maintenance HD, LVF is the most common cause for hospital admission followed by RTI. Hence, the management of HTN and preventive measures for RTI should be stressed in HD patients.

A study on cognitive decline with respect to metabolic syndrome and inflammation in elderly Indians.

BACKGROUND: This study was undertaken to find out if metabolic syndrome (MetS) in the elderly was associated with cognitive decline and also if this association was modified by the presence of inflammation. MATERIALS AND METHODS: 100 patients more than 60 years of age were divided into 2 groups of 50 each and were age and sex matched. Group 1 and 2 had patients with and without MetS, respectively. The individual components of MetS were measured in each patient. Cognitive decline was measured by Modified Mini-Mental Score (3MS) of Teng. Inflammation was measured by high-sensitivity C-reactive protein (hs-CRP). RESULTS: Fasting hyperglycemia was the most common component of MetS (60% of group 1). The mean serum hs-CRP in patients of group 1 was 6.56 ± 9.72 while that in the patients of group 2 was 1.95 ± 1.93. In the group-1, 36% (n = 18) patients were having a decreased 3MS, whereas in group-2, 22% (n = 11) were having a decreased 3MS. MetS was associated with an odd's ratio of 1.99 for developing cognitive decline. 3MS had a negative correlation with hs-CRP values. Regression analysis showed a significant association of hs-CRP and MetS with cognitive decline in the elderly population. CONCLUSION: Cognitive decline in the elderly is associated with the presence of inflammation and MetS. Hence, early identification of the high-risk groups may offer benefit by disease course modification and better caregiving.

A study on pulmonary complications of systemic sclerosis in eastern India.

AIM: This study was undertaken to find out the characteristics of clinical, radiological and functional changes affecting the respiratory system in patients with systemic sclerosis (SSc) from eastern India, and the association of these characteristics with pulmonary hypertension. METHODS: This was a cross-sectional, observational study involving 46 patients. Other than the routine tests, anti-nuclear antibody (ANA), spirometry, diffusing capacity of lung for carbon monoxide (DLCO) measurement, chest radiograph, high-resolution computed tomography (HRCT) of thorax, 6-minute walk test and echocardiography were done. RESULTS: Out of a total of 46 patients, 27 patients had diffuse cutaneous SSc (dcSSc) and 19 had limited cuteaneous SSc (lcSSc). Eleven patients had pulmonary hypertension. The HRCT revealed diffuse parenchymal lung disease (DPLD) in 32 (65%) cases. The ANA was positive in 83% cases. Anti-Scl70 was found in 41% of patients with dcSSc and anti-centromere antibody was found in 47% of patients with lcSSc. Spirometry revealed restrictive pattern in 30 patients; 9 had obstruction; and the rest were normal. The DLCO was abnormal in 38 patients. A strong correlation was found between reduction in DLCO and pulmonary artery systolic pressure (PASP). Also, a strong association was observed between a drop of > 4% in oxygen saturation on 6-minute walk test and presence of pulmonary arterial hypertension (PAH). CONCLUSIONS: Majority of the patients with SSc had restrictive lung disease with abnormal DLCO and features resembling non-specific interstitial pneumonia. Nucleolar ANA was predominantly found in patients having PAH. Presence of DPLD had a negative association with presence of anti-centromere antibody. Reduction in DLCO and a fall of > 4% in oxygen saturation on 6-minute walk test may be used as predictors of PAH in asymptomatic individuals.