Improving the Robustness and Quality of Biomedical CNN Models through Adaptive Hyperparameter Tuning

Deep learning is an obvious method for the detection of disease, analyzing medical images and many researchers have looked into it. However, the performance of deep learning algorithms is frequently influenced by hyperparameter selection, the question of which combination of hyperparameters are best emerges. To address this challenge, we proposed a novel algorithm for Adaptive Hyperparameter Tuning (AHT) that automates the selection of optimal hyperparameters for Convolutional Neural Network (CNN) training. All of the optimal hyperparameters for the CNN models were instantaneously selected and allocated using a novel proposed algorithm Adaptive Hyperparameter Tuning (AHT). Using AHT, enables CNN models to be highly autonomous to choose optimal hyperparameters for classifying medical images into various classifications. The CNN model (Deep-Hist) categorizes medical images into basic classes: malignant and benign, with an accuracy of 95.71%. The most dominant CNN models such as ResNet, DenseNet, and MobileNetV2 are all compared to the already proposed CNN model (Deep-Hist). Plausible classification results were obtained using large, publicly available clinical datasets such as BreakHis, BraTS, NIH-Xray and COVID-19 X-ray. Medical practitioners and clinicians can utilize the CNN model to corroborate their first malignant and benign classification assessment. The recommended Adaptive high F1 score and precision, as well as its excellent generalization and accuracy, imply that it might be used to build a pathologist's aid tool.

Biochemical Parameters as Prognostic Markers in Severely Ill COVID-19 Patients

Background Prognostication plays a pivotal role in critical care medicine. Its importance is indisputable in the management of coronavirus disease 2019 (COVID-19), as the presentation of this disease may vary from docile, self-limiting symptoms to lethal conditions. Amid the COVID-19 pandemic, much emphasis was initially placed on molecular and serological testing. However, it was realized later that routine laboratory tests also provide key information in terms of the severity of the disease and thus could be used to predict the outcome of these patients. Methodology The aim of our study was to evaluate the biochemical parameters as prognostic markers in severely ill COVID-19 patients. We carried out a retrospective, case-control study. The study population was comprised of all severely ill COVID-19 patients admitted between October 2020 and January 2021 at our level 3 COVID hospital. Cases were defined as the patients who expired despite treatment and all resuscitative measures as per the standard operating procedures (SOPs) of our COVID intensive care unit (ICU) while controls were defined as the patients that were transferred out of the COVID ICU for further recovery. The detailed history, findings of physical examination, vitals recorded by point of care testing (POCT) devices at our ICU, clinical diagnosis, and the results of the biochemical analysis were recorded in a specially designed pro forma. The biochemical parameters recorded at the time of admission were compared between the groups of controls and cases in order to evaluate their role as predictors of mortality using appropriate statistical methods. P-values less than 0.05 were considered statistically significant. For all the parameters that showed a statistically significant difference, receiver operating characteristics (ROC) analysis was done to assess the utility of biochemical parameters as predictors of mortality or survival. Areas under the curve (AUCs) of 0.6 to 0.7, 0.7 to 0.8, 0.8 to 0.9, and >0.9 were considered acceptable, fair, good, and excellent for discrimination, respectively. Results Of the 178 severely ill COVID-19 patients enrolled in the study, 86 were controls and 92 were cases (52% mortality). Serum urea (p<0.0001), creatinine (p=0.0019), aspartate transaminase (AST) (p=0.0104), lactate dehydrogenase (LDH) (p=0.0001), procalcitonin (PCT) (p=0.0344), and interleukin 6 (IL-6) (p=0.0311) levels were significantly higher (p<0.05), while total protein (p=0.0086), albumin (p<0.0001), and indirect bilirubin (p=0.0147) levels were significantly lower (p<0.05) in cases as compared to controls. The difference was statistically insignificant (p>0.05) for serum sodium, potassium, total and direct bilirubin, globulin, alanine transaminase (ALT), alkaline phosphatase (ALP), D-dimer, and ferritin. On ROC analysis, urea was fair (AUC=0.721), creatinine (AUC=0.698) and IL-6 (AUC=0.698) were acceptable predictors of mortality, while albumin (AUC=0.698) was an acceptable predictor of survival in severely ill COVID-19 patients during their intensive care stay. Conclusion Understanding the pathophysiological changes associated with the severity of COVID-19 in terms of an alteration of biochemical parameters is a pressing priority. Our study highlights the importance of routine laboratory tests in predicting outcomes in severely ill COVID-19 patients.

Dynamics of Middle East Respiratory Syndrome Coronavirus (MERS-CoV) involving fractional derivative with Mittag-Leffler kernel.

Since 2012, the Middle East has seen a steady rise in the Middle East Respiratory Syndrome Coronavirus (MERS-CoV). A fractional derivative of the non-singular Mittag-Leffler type is used in this research to conduct a mathematical analysis of the dynamics of MERS-CoV infection transmission. The dynamics of such a disease with an additional degree of freedom and non-singular behavior are discovered through the use of the aforementioned fractional operator, and this is one of the important components of our prepared paper. Using the concept of fixed point theory, the existence and uniqueness of solutions are demonstrated. The stability analysis is also tested with the help of the Ulam-Hyers approach, respectively. The numerical solution has been conducted by using the fractional Adams-Bashforth scheme. In the numerical simulation, all classes are demonstrated through the graphical presentation regarding the changing values of fractional-order at time t. The results at various fractional-order laying between (0,1] are drawn with the help of Matlab. We also provide a comparison of the proposed approach with that of the Caputo operator. The outcomes that were achieved illustrate that the considered scheme is highly methodical and very efficient compared to the Caputo fractional operator.

Study of Utility of Basic Arterial Blood Gas Parameters and Lactate as Prognostic Markers in Patients With Severe Dengue.

Background The importance of prognostication in critical care cannot be over-emphasized, especially in the context of diseases like dengue, as their presentation may vary from mild fever to critical life-threatening illness. With the help of prognostic markers, it is possible to identify patients at higher risk and thus improve their outcome with timely intervention. Basic arterial blood gas (ABG) parameters, i.e., potential of hydrogen (pH), partial pressure of oxygen (PO2), partial pressure of carbon dioxide (PCO2) and bicarbonate are useful parameters, especially in critical care medicine as they are known to vary with the severity of illness. Hyperlactatemia is often referred to as a "powerful predictor of mortality". Basic ABG parameters and lactate have been used as an essential prognostic modality in critically ill patients for decades; however, the evidence remains limited for their role as prognostic markers in patients with severe dengue. Method We carried out an observational retrospective cohort study comprising 163 patients with severe dengue, admitted between July 2021 and November 2021 at Medical Intensive Care Unit (MICU) of Shri Ram Murti Smarak Institute of Medical Sciences (SRMS IMS), Bareilly, Uttar Pradesh, India. Basic ABG parameters and lactate levels at the time of admission to MICU were compared between survivor and non-survivor groups of patients with severe dengue in order to evaluate their prognostic utility as predictors of mortality. Results pH (p<0.0001), PO2 (p=0.01) and bicarbonate (<0.0001) levels were significantly lower, while PCO2 (p=0.002) and lactate (p<0.0001) levels were significantly higher in non-survivor group as compared to survivor group. Lactate was found to be the best prognostic marker with Area Under the Curve (AUC) of 88.7% on Receiver Operating Characteristics (ROC) analysis. Conclusion Basic arterial blood gas parameters and lactate can be used as feasible prognostic markers in patients with severe dengue.

Phytochemistry, biological activities and in silico molecular docking studies of Oxalis pes-caprae L. compounds against SARS-CoV-2.

Phytochemicals are directly involved in therapeutic treatment or precursors to synthesize useful drugs. The current study was aimed to evaluate the phytocompounds and their biopotentials using methanolic and n-hexane extracts of various parts of Oxalis pes-caprae. For the phytochemical analysis, standard procedures were used, whereas Aluminum Chloride reagent and Follin-ciocalteau reagent methods were used to determine total flavonoid and phenolic contents. Radical scavenging DPPH, phosphomolybdenum reduction, and reducing power assays were used to assess antioxidative potentials. Antibacterial potential was determined by applying disc diffusion method while cytotoxicity was determined employing brine shrimp assay. FT-IR (Fourier-transform infrared) analysis was utilized to gather spectral information, while molecular docking tools were employed to look at how O. pes-caprae plant-based ligands interact with the target protein COVID-19 3CLPro (PDB:6LU7). Phenols, flavonoids, alkaloids and saponins were tested positive in preliminary phytochemical studies. TPC and TFC in different extracts ranging from (38.55 ± 1.72) to (65.68 ± 0.88) mg/g GAE/g and (24.75 ± 1.80) to (14.83 ± 0.92) mg/g QUE/g were used respectively. IC50 value (24.75 ± 0.76 g/mL) by OXFH, total antioxidant capacity (55.89 ± 1.75) mg/g by OXLM, reducing potential (34.98 ± 1.089) mg/g by OXSM, maximum zone of inhibition against B. subtilis (24 ± 0.65 mm) by OXLM and maximum cytotoxicity 96% with LD50 19.66 (µg/mL) by OXSM were the best calculated values among all extracts. Using molecular docking, it was found that Caeruleanone A, 2',4'-Dihydroxy-2″-(1-hydroxy-1-methylethyl) dihydrofuro [2,3-h] flavanone and Vadimezan demonstrated best affinity with the investigated SARS CoV-2 Mpro protein. This work provide justification about this plant as a source of effective phytochemicals and their potential against microbes could lead to development of biosafe drugs for the welfare of human being. In future, different in vitro and in vivo biological studies can be performed to further investigate its biomedical potentials.

Impact of an Orthopedic-Delivered 'One-Stop' Clinic for Ambulatory Trauma on Emergency Department and Fracture Clinic Pressures During the COVID-19 Pandemic.

INTRODUCTION:  In March 2020, the World Health Organization declared the severe acute respiratory syndrome (SARS) (coronavirus disease, COVID-19) outbreak as a pandemic. In response to the rising number of coronavirus cases in the United Kingdom, the British Orthopaedic Association (BOA) issued a series of emergency guidelines for the management of trauma and orthopedic patients during the pandemic. In line with this guidance, the orthopedic team at the Worcestershire Royal Hospital set up a 'one-stop-shop' minor injuries unit (MIU). This seven-day service provided a direct pathway to the trauma clinic for ambulatory patients who would usually be managed in the emergency department (ED), intending to reduce both the pressure on the ED and the need for further follow-up appointments. The aim of this project was to evaluate the service provided to patients and to inform our practice during the next stages of the pandemic and beyond. MATERIALS AND METHODS:  Data were collected retrospectively from a clinic database, dictated letters, and scanned patient notes. The data collection period was over six weeks from April 6, 2020 to May 18, 2020. Data collected included patient age and gender, time of arrival and departure, grade of reviewing clinician, diagnosis, treatment, and outcome of clinic attendance, including the timing of follow-up. RESULTS:  Some 700 patients were seen in the MIU over six weeks. Some 98% of patients were seen by an orthopedic registrar (resident) or a consultant (attending) and 85% were seen and treated within an hour. Some 71% of patients were discharged after their initial appointment, and only nine patients (1%) required a fracture clinic appointment within 72 hours. A total of 15 patients (2%) re-attended with concerns, and just four of these required additional interventions. CONCLUSIONS:  We delivered a seven-day minor injuries service in which the majority of patients received definitive management at first attendance, reducing the demand for fracture clinic appointments. Some 700 patients who would have been treated in the ED were seen in the MIU instead, relieving pressure on the ED. The lessons learned allowed us to plan for the 'second peak' in COVID-19 cases and will inform ongoing practice as we work to recover elective services.

Identification of novel mutations in SARS-COV-2 isolates from Turkey.

SARS-CoV-2 was first detected in the city of Wuhan, Hubei Province, China. In this study, we identified 11 unique mutations in viral SARS-COV-2 isolates from Turkey. Nine of them cause structural alterations in the S protein, nsp2, nsp3, nsp4 and nsp12 regions. The mutations identified here might have significant functional implications that need to be addressed in future studies in the context of vaccine engineering and therapeutic interventions. Moreover, transmission and phylogenetic analysis revealed multiple independent sources of introductions of SARS-CoV-2 into Turkey and a close relationship to the isolates from Saudi Arabia.