Search

Abstract Predicting postpartum hemorrhage (PPH) before delivery is crucial for enhancing patient outcomes, enabling timely transfer and implementation of prophylactic therapies. We attempted to utilize machine learning (ML) using basic pre-labor clinical data and laboratory measurements to predict postpartum Hemoglobin (Hb) in non-complicated singleton pregnancies. The local databases of two academic care centers on patient delivery were incorporated into the current study. Patients with preexisting coagulopathy, traumatic cases, and allogenic blood transfusion were excluded from all analyses. The association of pre-delivery variables with 24-h post-delivery hemoglobin level was evaluated using feature selection with Elastic Net regression and Random Forest algorithms. A suite of ML algorithms was employed to predict post-delivery Hb levels. Out of 2051 pregnant women, 1974 were included in the final analysis. After data pre-processing and redundant variable removal, the top predictors selected via feature selection for predicting post-delivery Hb were parity (B: 0.09 [0.05–0.12]), gestational age, pre-delivery hemoglobin (B:0.83 [0.80–0.85]) and fibrinogen levels (B:0.01 [0.01–0.01]), and pre-labor platelet count (B*1000: 0.77 [0.30–1.23]). Among the trained algorithms, artificial neural network provided the most accurate model (Root mean squared error: 0.62), which was subsequently deployed as a web-based calculator: https://predictivecalculators.shinyapps.io/ANN-HB . The current study shows that ML models could be utilized as accurate predictors of indirect measures of PPH and can be readily incorporated into healthcare systems. Further studies with heterogenous population-based samples may further improve the generalizability of these models.

Abstract With improved treatment of coronary artery disease, more patients are surviving until heart failure occurs. This leads to an increase in patients needing devices for struggling with heart failure. Ventricular assist devices are known as the mainstay of these devices. This study aimed to design a centrifugal pump as a ventricular assist device. In order to design the pump, firstly, the geometrical parameters of the pump, including the gap distance, blade height, and position of the outlet relative to the blade, were investigated. Finally, the selected configuration, which had all the appropriate characteristics, both hydraulically and physiologically, was used for the rest of the study. The study of the blade, as the main component in energy transfer to the blood, in a centrifugal pump, has been considered in the present study. In this regard, the point-to-point design method, which is used in industrial applications, was implemented. The designer chooses the relationship between the blade angles at each radius in the point-to-point method. The present study selected logarithmic and second-order relations for designing the blade’s profile. In total, 58 blades were examined in this study, which differed regarding blade inlet and outlet angles and the relationship between angle and radial position. ANSYS CFX 17.0 software was utilized to simulate blades’ performances, and a benchmark pump provided by the US Food and Drug Administration (FDA) was used to validate the numerical simulations. Then, the selected impeller from the numerical investigation was manufactured, and its performance was compared experimentally with the FDA benchmark pump. A hydraulic test rig was also developed for experimental studies. The results showed that among the blades designed in this study, the blade with an input angle of 45° and an output angle of 55°, which is designed to implement a logarithmic relationship, has the best performance. The selected impeller configuration can increase the total head (at least by 20%) at different flow rates compared to the FDA pump.

Abstract Glass fiber-reinforced polymer (GFRP) reinforcements are superior to traditional steel bars in concrete structures, particularly in vertical elements like columns, and offer significant advantages over conventional steel bars when subjected to axial and eccentric loadings. However, there is limited experimental and numerical research on the behavior of GFRP-reinforced concrete (RC) columns under eccentric loading having different spacing of stirrups. In this study, six specimens were cast under three different values of eccentricities (25 mm, 50 mm, and 75 mm) with two groups of stirrups spacing (50 mm and 100 mm). The experimental results showed that by increasing the eccentricity value, there was a reduction in the load-carrying capacity of the specimens. The finite element ABAQUS software was used for the numerical investigation of this study. The results from the finite element analysis (FEA) were close to the experimental results and within the acceptable range. The maximum difference between the experimental and FEA results was 3.61% for the axial load and 12.06% for the deformation.

Abstract Therapeutic Alliance (TA) has been consistently reported as a robust predictor of therapy outcomes and is one of the most investigated therapy relational factors. Research on therapists' and clients’ contributions to the alliance development and the alliance-outcome relationship had shown mixed results. The relation of the therapist’s and client’s biological markers with the alliance is an important and under-investigated topic. Taking advantage of data mining techniques, this exploratory study aimed to investigate the role of different therapist and client factors, including heart rate (HR) and electrodermal activity (EDA), in relation to TA. Twenty-two dyads with 6 therapists and 22 clients participated in the study. The Working Alliance Inventory (WAI) was used to evaluate the client’s and therapist's perception of the alliance at the end of each session and through the therapy processes. The Cross-Industry Standard Process for Data Mining (CRISP-DM) was used to explore patterns that may contribute to TA. Machine Learning (ML) models have been employed to provide insights into the predictors and correlates of TA. Our results showed that Linear Regression (LR) was the best technique for predicting the therapist’s TA, with client “Diagnostic” and therapy “Termination” being identified as significant predictors of the therapist’s TA. In addition, for clients’ TA, the Random Forest (RF) was shown to have the best performance. The therapist’s TA and therapy “Outcome” were observed as the most influential predictors for the client’s TA. In addition, while the Heart Rate (therapist) was negatively associated with the therapist’s TA, EDA in the client was a physiological indicator related to the client’s TA. Overall, these findings can assist in identifying key factors that therapists should focus on to enhance the quality of therapeutic alliance. Results are discussed in terms of their consistency with empirical literature, innovative and interdisciplinary research on the therapeutic alliance field, and, in particular, the use of the Data Mining approach in a psychotherapy context.

Abstract We introduced Double Attention Res-U-Net architecture to address medical image segmentation problem in different medical imaging system. Accurate medical image segmentation suffers from some challenges including, difficulty of different interest object modeling, presence of noise, and signal dropout throughout the measurement. The base line image segmentation approaches are not sufficient for complex target segmentation throughout the various medical image types. To overcome the issues, a novel U-Net-based model proposed that consists of two consecutive networks with five and four encoding and decoding levels respectively. In each of networks, there are four residual blocks between the encoder-decoder path and skip connections that help the networks to tackle the vanishing gradient problem, followed by the multi-scale attention gates to generate richer contextual information. To evaluate our architecture, we investigated three distinct data-sets, (i.e., CVC-ClinicDB dataset, Multi-site MRI dataset, and a collected ultrasound dataset). The proposed algorithm achieved Dice and Jaccard coefficients of 95.79%, 91.62%, respectively for CRL, and 93.84% and 89.08% for fetal foot segmentation. Moreover, the proposed model outperformed the state-of-the-art U-Net based model on the external CVC-ClinicDB, and multi-site MRI datasets with Dice and Jaccard coefficients of 83%, 75.31% for CVC-ClinicDB, and 92.07% and 87.14% for multi-site MRI dataset, respectively.

Abstract Activated protein C (APC) is a serine protease with anticoagulant and cytoprotective activities which make it an attractive target for diagnostic and therapeutic applications. In this work, we present one-step activation of APC from a commercial source of protein C (PC, Ceprotin) followed by rapid and efficient purification using an APC-specific aptamer, HS02-52G, loaded on MyOne superparamagnetic beads. Due to the Ca2+-dependent binding of APC to HS02-52G, an efficient capturing of APC was applied in the presence of Ca2+ ions, while a gentle release of captured APC was achieved in the elution buffer containing low EDTA concentration (5 mM). The captured and eluted APC showed more than 95% purity according to SDS-PAGE gel analysis and an enzyme-linked fluorescent assay (VIDAS Protein C). The purification yield of 45% was calculated when 4.2 µg APC was used, however this yield reduced to 21% if the starting amount of APC increased to 28.5 µg. Altogether, this method is recommended for rapid and efficient PC activation and APC purification. The purified APC can be used directly for downstream processes where high concentration of pure and active APC is needed.

Abstract The levels of the gonadal hormones estradiol and progesterone vary throughout the menstrual cycle thereby affecting cognition, emotion, mood, and social behaviour. However, how these hormones modulate the balance of neural excitation and inhibition, which crucially regulate processing and plasticity, is not fully understood. We here used paired-pulse stimulation to investigate in healthy humans the action of low and high estradiol and progesterone on intracortical inhibition in somatosensory (SI) and visual cortex (V1). We found that paired-pulse suppression in both SI and VI depended on estradiol. During high estradiol levels, paired-pulse suppression was significantly reduced. No comparable effects were found for progesterone, presumably due to a confounding effect of estradiol. Also, no hormone level-depending effects were observed for single-pulse evoked SEPs (somatosensory evoked potentials) and VEPs (visual evoked potentials) indicating a specific hormonal action on intracortical processing. The results demonstrate that estradiol globally modulates the balance of excitation and inhibition of SI and VI cortex.