Search

Abstract Background Despite the globally reducing hospitalization rates and the much lower risks of Covid-19 mortality, accurate diagnosis of the infection stage and prediction of outcomes are clinically of interest. Advanced current technology can facilitate automating the process and help identifying those who are at higher risks of developing severe illness. This work explores and represents deep-learning-based schemes for predicting clinical outcomes in Covid-19 infected patients, using Visual Transformer and Convolutional Neural Networks (CNNs), fed with 3D data fusion of CT scan images and patients’ clinical data. Methods We report on the efficiency of Video Swin Transformers and several CNN models fed with fusion datasets and CT scans only vs. a set of conventional classifiers fed with patients’ clinical data only. A relatively large clinical dataset from 380 Covid-19 diagnosed patients was used to train/test the models. Results Results show that the 3D Video Swin Transformers fed with the fusion datasets of 64 sectional CT scans + 67 clinical labels outperformed all other approaches for predicting outcomes in Covid-19-infected patients amongst all techniques (i.e., TPR = 0.95, FPR = 0.40, F0.5 score = 0.82, AUC = 0.77, Kappa = 0.6). Conclusions We demonstrate how the utility of our proposed novel 3D data fusion approach through concatenating CT scan images with patients’ clinical data can remarkably improve the performance of the models in predicting Covid-19 infection outcomes. Significance Findings indicate possibilities of predicting the severity of outcome using patients’ CT images and clinical data collected at the time of admission to hospital.

Key clinical massage Pseudoaneurysms and aneurysms of the visceral arteries are rare entities. To the best of our knowledge, rupture of a proximal parental artery during endovascular treatment of a visceral aneurism/pseudoaneurysm has not been reported and should be kept in mind as a tragic possibility immediately following an apparently successful management of them. Abstract A 55‐year‐old woman with a history of coronary artery disease was referred to our hospital with abdominal pain as her primary complaint. Early works revealed anemia, a small amount of free peritoneal fluid, and a possible large aneurysm or pseudoaneurysm by the greater curvature of the stomach. She underwent emergency angiography that showed a large aneurism/pseudoaneurysm of the gastroepiploic artery. Successful embolization of the lesion was performed using the isolation technique. Perforation of a side branch of the gastroduodenal artery was observed on the immediate postembolization control angiography. Therefore, parent artery coiling was done immediately with good results. She was symptom‐free and stable hemodynamically after the procedure, during the hospital course, and in the follow‐ups.

Background: Alzheimer’s disease (AD) is the most common brain failure for which no cure has yet been found. The disease starts with a disturbance in the brain structure and then it manifests itself clinically. Therefore, by timely and correct diagnosis of changes in the structure of the brain, the occurrence of this disease or at least its progression can be prevented. Due to the fact that magnetic resonance imaging (MRI) can be used to obtain very useful information from the brain, and also because it is non-invasive, this method has been considered by researchers. Materials and Methods: The data were obtained from an MRI database (MIRIAD) of 69 subjects including 46 AD patients and 23 healthy controls (HC). Individuals were categorized based on two criteria including NINCDS-ADRAD and MMSE, as the gold standard. In this paper, we used the support vector machine (SVM) and Bayesian SVM classifiers. Results: Using the SVM classifier with Gaussian radial basis function (RBF) kernel, we distinguished AD and HC with an accuracy of 88.34%. The most important regions of interest (ROIs) in this study included right para hippocampal gyrus, left para hippocampal gyrus, right hippocampus, and left hippocampus. Conclusion: This study showed that the SVM model with Gaussian RBF kernel can distinguish AD from HC with high accuracy. These studies are of great importance in medical science. Based on the results of this study, MRI centers and neurologists can perform AD screening tests in people over the age of 50 years.

Introduction: In human, the growth of the cerebellum may be less affected by Intrauterine Growth Restriction (IUGR). Then, measurement of transverse cerebellar diameter can be used as an appropriate measurement for determining the gestational age of fetus with IUGR. This study was performed with aim to evaluate the agreement of different biometric parameters such as transverse cerebellar diameter, femur length and biparietal diameter in sonographic determining of gestational age of fetuses with IUGR and comparing them. Methods: This cross sectional study was performed on 35 pregnant women with small abdomen suspected to IUGR fetuses who referred to sonography ward of Kermanshah Imam Reza Hospital in 2016. Gestational age was determined based on the LMP of patient. Ultrasound was performed for patients; gestational age was measured by Transverse Cerebellar Diameter (TCD) and other sonographic routine parameters, and then the results were compared. Data were analyzed by SPSS software (version 22) and paired t-test, Wilcoxon, Pearson and Spearman correlation coefficient. PResults: There was no significant difference between gestational age predicted by TCD parameter and LMP (P>0.05). But, gestational age determined by the parameters of BPD and Femur Length was significantly lower than gestational age based on LMP (P

Background and Aim: This study aimed at analyzing the demographic characteristics of the patients with non-traumatic intracranial hemorrhage, their angiographic results and the correlation between computed tomography (CT) scan and angiography findings. Methods and Materials/Patients: In a descriptive study, we retrospectively reviewed the records of the patients with non-traumatic intracranial hemorrhage based on their brain CT or lumbar puncture findings from 2011 to 2017. For all patients, four vessel catheter angiography via the femoral approach was performed in the Medical Imaging Center of Kermanshah University of Medical Sciences, Iran. Results: We investigated 143 cases with non-traumatic intracranial hemorrhage which was indicated in CT findings (91.61%) or lumbar puncture (8.39%). Of 143 patients, 64 (44.8%) were men and 79 (55.2%) women (mean age 53.7±12.1 years old). Moreover, 104(72.7%) patients had Subarachnoid Hemorrhage (SAH), 19(13.3%) of them had Intracranial Hemorrhage (ICH), 8(5.6%) ones had Intraventricular Hemorrhage (IVH) and 12(8.4%) patients had normal brain CT. There were 100 cases of aneurysm (69.93%), 13 cases of Arteriovenous Malformations (AVM) (9.09%), 28 cases with negative angiograms (19.58%), and 2 cases with other pathologies (1.4%). Seven (4.89%) cases of multiple aneurysms were also recorded. Of 13 patients with AVM, 8(61.5%) patients had AVM in parieto-occipital region, 2(15.4%) in temporal, 1(7.7%) in frontal region and 2(15.4%) had deep AVM. Two (1.4%) patients with SAH had dural AVF. Twenty-eight (19.6%) patients had negative angiogram, 22 cases of them had second angiography after one to two weeks. The most common positive finding in the second angiogram was Acom aneurysm (18.18%). Conclusion: Aneurysms and AVMs are the most common causes of non-traumatic intracranial hemorrhage. The most common site of intracranial aneurysms is the circle of Willis. A second angiogram after one to two weeks is necessary in most patients with negative results in first angiography after non-traumatic intracranial hemorrhage.

We aimed to determine the diagnostic accuracy of maternal renal vasculature Doppler ultrasound indices in the prediction of preeclampsia. A total of 40 pregnant women with a gestational age of more than 20 weeks were included and followed. The pregnant women underwent a Doppler ultrasound examination to measure the resistance index (RI) of the interlobar arteries of right and left kidneys and the renal interlobar vein impedance index (RIVI). Of the included women, 15 patients developed preeclampsia based on clinical and laboratory criteria. The renal vascular Doppler ultrasound indices were compared between the two groups. Then, using the Receiver Operating Characteristic (ROC) analysis, the diagnostic accuracy of interlobar artery RI and RIVI were investigated in the prediction of preeclampsia occurrence. RIVI values of right and left kidneys were significantly higher in the preeclampsia group compared to the normal pregnancy group. The left kidney interlobar artery RI at a cut-point of 0.59 had a sensitivity of 100% and a specificity of 40% (area under curve=0.7, P-0.03) in the prediction of preeclampsia. The RIVI of the left kidney (adjusted odds ratio=17.14, 95% CI = 3.46 to 47.28) was statistically significant in predicting preeclampsia (P-0.006). We found that, besides other routine methods, using Doppler ultrasound and measuring RIVI can be reliable in the prediction of preeclampsia. [ABSTRACT FROM AUTHOR]

Introduction: Congenital dislocation of the knee (CDK) is rare, and clinical semiology at birth is not always suitably analyzed. Existing classifications fail to guide treatment. The aim of the present study was to develop a CDK classification for the neonatal period., Hypothesis: A classification based on neonatal severity of clinical signs is easy to implement on simple criteria., Material and Methods: Fifty-one CDKs (40 patients) seen neonatally were included. Three types could be distinguished in terms of reduction and stability: type I, easily reducible CDK, with reduction snap when the femoral condyles pass in flexion, remaining stable in flexion; type II, "recalcitrant" dislocation, reducible by posteroanterior "piston" but unstable, with iterative dislocation once posteroanterior pressure on the condyles is relaxed; and type III, irreducible. The number of anterior skin grooves, global range of motion, flexion deficit and reduction stability were noted for each type., Results: Mean age at first consultation was 5.6 days (range: 0-30). CDK was type I, II and III in respectively 28, 16 and 7 cases. Number of skin grooves, flexion and baseline range of motion were greater in type I than types II and III., Conclusion: The present neonatal clinical classification is original, logical and simple. It may be useful for prognosis and guiding treatment., Level of Evidence: IV, single-center retrospective series., (Copyright © 2016 Elsevier Masson SAS. All rights reserved.)

Introduction: An original classification of congenital dislocation of the knee (CDK) was drawn up, based on neonatal semiology. The objective of the present study was to assess impact on treatment decision-making and prognosis., Material and Methods: Fifty-one CDKs in 40 patients were classified neonatally into 3 types: I, reducible (n=28); II, recalcitrant (n=16); and III, irreducible (n=7). Number of anterior skin grooves, range of motion (RoM), flexion deficit and reduction stability were recorded. Depending on reducibility, treatment comprised: physiotherapy with splints, traction with cast immobilization, or surgery. At follow-up, knees were assessed in terms of RoM and stability., Results: Mean age at first consultation was 5.6 days (range: 0-30). Mean age at follow-up was 9 years (range: 1-26). Physiotherapy with splinting achieved stable reduction in all type-I knees. Five type-II knees (31%) required traction, none of which needed surgery. Four type-III knees (57%) required surgery. Outcome was good or excellent in 82% of type-I knees, good in 68% of type II and poor in all type-III knees., Conclusion: The study confirmed the relevance of the present neonatal classification to treatment, with increasing rates of surgical indication and decreasing rates of satisfactory outcome from types I to III. Therapeutic attitude can be graded according to severity of CDK., Level of Evidence: IV, single-center retrospective series., (Copyright © 2016 Elsevier Masson SAS. All rights reserved.)