A bi-stage feature selection approach for COVID-19 prediction using chest CT images.

The rapid spread of coronavirus disease has become an example of the worst disruptive disasters of the century around the globe. To fight against the spread of this virus, clinical image analysis of chest CT (computed tomography) images can play an important role for an accurate diagnostic. In the present work, a bi-modular hybrid model is proposed to detect COVID-19 from the chest CT images. In the first module, we have used a Convolutional Neural Network (CNN) architecture to extract features from the chest CT images. In the second module, we have used a bi-stage feature selection (FS) approach to find out the most relevant features for the prediction of COVID and non-COVID cases from the chest CT images. At the first stage of FS, we have applied a guided FS methodology by employing two filter methods: Mutual Information (MI) and Relief-F, for the initial screening of the features obtained from the CNN model. In the second stage, Dragonfly algorithm (DA) has been used for the further selection of most relevant features. The final feature set has been used for the classification of the COVID-19 and non-COVID chest CT images using the Support Vector Machine (SVM) classifier. The proposed model has been tested on two open-access datasets: SARS-CoV-2 CT images and COVID-CT datasets and the model shows substantial prediction rates of 98.39% and 90.0% on the said datasets respectively. The proposed model has been compared with a few past works for the prediction of COVID-19 cases. The supporting codes are uploaded in the Github link: https://github.com/Soumyajit-Saha/A-Bi-Stage-Feature-Selection-on-Covid-19-Dataset.

Association of anticoagulation use during continuous kidney replacement therapy and 90-day outcomes: A multicentre study.

Introduction: Anticoagulation is recommended during continuous kidney replacement therapy (CKRT) to prolong the filter lifespan for optimal filter performance. We aimed to evaluate the effect of anticoagulation during CKRT on dialysis dependence and mortality within 90 days of intensive care unit (ICU) admission. Method: Our retrospective observational study evaluated the first CKRT session in critically ill adults with acute kidney injury (AKI) in Singapore from April to September 2017. The primary outcome was a composite of dialysis dependence or death within 90 days of ICU admission; the main exposure variable was anticoagulation use (regional citrate anticoagulation [RCA] or systemic heparin). Multivariable logistic regression was performed to adjust for possible confounders: age, female sex, Acute Physiology and Chronic Health Evaluation (APACHE II) score, liver dysfunction, coagulopathy (international normalised ratio[INR] >1.5) and platelet counts of less than 100,000/uL). Results: The study cohort included 276 patients from 14 participating adult ICUs, of whom 176 (63.8%) experienced dialysis dependence or death within 90 days of ICU admission (19 dialysis dependence, 157 death). Anticoagulation significantly reduced the odds of the primary outcome (adjusted odds ratio [AOR] 0.47, 95% confidence interval [CI] 0.27-0.83, P=0.009). Logistic regression analysis using anticoagulation as a 3-level indicator variable demonstrated that RCA was associated with mortality reduction (AOR 0.46, 95% CI 0.25-0.83, P=0.011), with heparin having a consistent trend (AOR 0.51, 95% CI 0.23-1.14, P=0.102). Conclusion: Among critically ill patients with AKI, anticoagulation use during CKRT was associated with reduced dialysis or death at 90 days post-ICU admission, which was statistically significant for regional citrate anticoagulation and trended in the same direction of benefit for systemic heparin anticoagulation. Anticoagulation during CKRT should be considered whenever possible.

Breast cancer detection from thermal images using a Grunwald-Letnikov-aided Dragonfly algorithm-based deep feature selection method.

Breast cancer is one of the deadliest diseases in women and its incidence is growing at an alarming rate. However, early detection of this disease can be life-saving. The rapid development of deep learning techniques has generated a great deal of interest in the medical imaging field. Researchers around the world are working on developing breast cancer detection methods using medical imaging. In the present work, we have proposed a two-stage model for breast cancer detection using thermographic images. Firstly, features are extracted from images using a deep learning model, called VGG16. To select the optimal subset of features, we use a meta-heuristic algorithm called the Dragonfly Algorithm (DA) in the second step. To improve the performance of the DA, a memory-based version of DA is proposed using the Grunwald-Letnikov (GL) method. The proposed two-stage framework has been evaluated on a publicly available standard dataset called DMR-IR. The proposed model efficiently filters out non-essential features and had 100% diagnostic accuracy on the standard dataset, with 82% fewer features compared to the VGG16 model.

A deep learning model for classifying human facial expressions from infrared thermal images.

The analysis of human facial expressions from the thermal images captured by the Infrared Thermal Imaging (IRTI) cameras has recently gained importance compared to images captured by the standard cameras using light having a wavelength in the visible spectrum. It is because infrared cameras work well in low-light conditions and also infrared spectrum captures thermal distribution that is very useful for building systems like Robot interaction systems, quantifying the cognitive responses from facial expressions, disease control, etc. In this paper, a deep learning model called IRFacExNet (InfraRed Facial Expression Network) has been proposed for facial expression recognition (FER) from infrared images. It utilizes two building blocks namely Residual unit and Transformation unit which extract dominant features from the input images specific to the expressions. The extracted features help to detect the emotion of the subjects in consideration accurately. The Snapshot ensemble technique is adopted with a Cosine annealing learning rate scheduler to improve the overall performance. The performance of the proposed model has been evaluated on a publicly available dataset, namely IRDatabase developed by RWTH Aachen University. The facial expressions present in the dataset are Fear, Anger, Contempt, Disgust, Happy, Neutral, Sad, and Surprise. The proposed model produces 88.43% recognition accuracy, better than some state-of-the-art methods considered here for comparison. Our model provides a robust framework for the detection of accurate expression in the absence of visible light.

In-vivo & in-vitro toxicity test of molecularly engineered PCMS: A potential drug for wireless remote controlled treatment.

PC, PCM, PCS, and PCMS are our designed & synthesized ∼8 nm PAMAM dendrimer (P) -based organic supramolecular systems, for example, PCMS has 32 molecular motors (M), 4 pH sensors (S) and 2 multi-level molecular electronic switches (C). We have reported earlier following a preliminary in-vitro test that the synthesized PCMS can selectively target cancer cell nucleotides if triggered wirelessly by an electromagnetic pulse. Here to further verify its drug potential, we have studied the preliminary efficacy, toxicity, and pharmacokinetics of P derivatives (PC, PCM, PCMS) in-vivo and in-vitro. We used ethanol-induced gastric inflammation model and cultured human gastric epithelial cells AGS to examine to the toxicity of PAMAM dendrimers cell permeability and toxicity, in (a) the cultured human gastric epithelium cells (AGS), and in (b) the gastric ulcer mice model. Here we report that the toxicity of PAMAM dendrimer (>G3.5) P can be reduced by adding C, M and S. Gastric ulcer is the primary stage of the manifestation of acute inflammation, even gastric epithelial cancer. Ethanol causes ulceration (ulcer index 30), thus upregulates both pro and active MMP-9. A 50 µl PCMS dose prior to ethanol administration reduces ulceration by ∼80% and downregulates MMP-9 and prevents oxidative damages of gastric tissue by ECM remodeling. Alcohol's inflammation of mouse stomach causes up-regulation of both pro and active MMP-9, resulting in oxidative damages of gastric tissue by ECM remodeling. PCMS in particular dose window reverses & alters ECM remodeling, thus, neutralizing alcohol-induced inflammation & generation of ROS.

Safety and tolerability of nonbronchoscopic lavage in ARDS.

STUDY OBJECTIVE: This study compared the safety profiles of bronchoscopic lavage with nonbronchoscopic lavage in the investigation of patients with acute lung injury (ALI) or ARDS. DESIGN: Single-center, randomized, cross-over study. SETTING: General ICU in the United Kingdom. PARTICIPANTS: Fourteen patients with ALI or ARDS. INTERVENTIONS: Bronchoscopic BAL and nonbronchoscopic BAL 1 h apart. MEASUREMENTS AND RESULTS: Hemodynamic and ventilatory parameters were recorded during and for 1 h following each procedure. On average, bronchoscopic lavage took longer to perform than nonbronchoscopic lavage (7 min and 6 s vs 2 min and 28 s, p < 0.001). During the procedures, bronchoscopic lavage increased heart rate and systolic BP more than nonbronchoscopic lavage (23% vs 10% [p < 0.01] and 18% vs 7% [p < 0.01]). Three patients had ST-segment depression during bronchoscopic, and one patient had ST-segment depression during nonbronchoscopic lavage (p = 0.298). Bronchoscopic lavage reduced minute ventilation by 63 +/- 17.3%, while nonbronchoscopic lavage only reduced it by 36 +/- 21.9% (p < 0.001). Paco(2) rose more after bronchoscopic lavage than after nonbronchoscopic lavage. CONCLUSION: Nonbronchoscopic lavage is associated with less marked physiologic derangements than bronchoscopic lavage. Further studies are required to validate the hypothesis that nonbronchoscopic lavage may be safer in patients with unstable coronary heart disease or head injury/raised intracranial pressure who are at risk from unpredictable fluctuations in hemodynamic and ventilatory profiles.

Resonant oscillation language of a futuristic nano-machine-module: eliminating cancer cells & Alzheimer aß plaques.

Nano-machine-module is designed and synthesized as a futuristic drug (PCMS) for cancer and Alzheimers by doping 2 Nile Red molecules in the cavity of a 5(th) generation PAM AM dendrimer P, and attaching 32 molecular rotors M, 4 pH sensors S on its surface. Molecular rotors and sensors enable the dendritic box surface to target specific sites, minimizing termination of healthy cells, e.g. cancer cells, nuclei acids (DNA) & spirals of Abeta Amyloid are disintegrated. Combined Excitation Emission Spectroscopy (CEES) shows directed energy transfer along M↔C↔S, this energy transmission path is itself an oscillation, and we image live resonant oscillation of the PCMS and the target molecular system. PCMS engages into resonant oscillations with spiral molecular structures. PCMS is designed to sense microsatellite instability & spirals with resonance frequencies in the kHz range. PCM is toxic, but the toxicity disappears as S is added to derive PCMS. PCMS does not even affect the dynamic instability of microtubule, a basic operator of living cells.

Three-Dimensional Molecular Modeling of a Diverse Range of SC Clan Serine Proteases.

Serine proteases are involved in a variety of biological processes and are classified into clans sharing structural homology. Although various three-dimensional structures of SC clan proteases have been experimentally determined, they are mostly bacterial and animal proteases, with some from archaea, plants, and fungi, and as yet no structures have been determined for protozoa. To bridge this gap, we have used molecular modeling techniques to investigate the structural properties of different SC clan serine proteases from a diverse range of taxa. Either SWISS-MODEL was used for homology-based structure prediction or the LOOPP server was used for threading-based structure prediction. The predicted models were refined using Insight II and SCRWL and validated against experimental structures. Investigation of secondary structures and electrostatic surface potential was performed using MOLMOL. The structural geometry of the catalytic core shows clear deviations between taxa, but the relative positions of the catalytic triad residues were conserved. Evolutionary divergence was also exhibited by large variation in secondary structure features outside the core, differences in overall amino acid distribution, and unique surface electrostatic potential patterns between species. Encompassing a wide range of taxa, our structural analysis provides an evolutionary perspective on SC clan serine proteases.

Applications of 3D CISS sequence for problem solving in neuroimaging.

Three-dimensional (3D) constructive interference in steady state (CISS) is a gradient-echo MRI sequence that is used to investigate a wide range of pathologies when routine MRI sequences do not provide the desired anatomic information. The increased sensitivity of the 3D CISS sequence is an outcome of the accentuation of the T2 values between cerebrospinal fluid (CSF) and pathological structures. Apart from its well-recognized applications in the evaluation of the cranial nerves, CSF rhinorrhea and aqueduct stenosis, we have found the CISS sequence to be useful for the cisternal spaces, cavernous sinuses and the ventricular system, where it is useful for detecting subtle CSF-intensity lesions that may be missed on routine spin-echo sequences. This information helps in the management of these conditions. After a brief overview of the physics behind this sequence, we illustrate its clinical applications with representative cases and discuss its potential role in imaging protocols.