Application of machine learning model in predicting the likelihood of blood transfusion after hip fracture surgery.

OBJECTIVE: Anemia is one of the common adverse reactions after hip fracture surgery. The traditional method to solve anemia is allogeneic transfusion. However, the transfusion may lead to some complications such as septicemia and fever. So far, few studies have reported roles of machine learning in predicting whether blood transfusion is needed or not after hip fracture surgery. Therefore, the purpose of this study is to develop machine learning models to predict the likelihood of postoperative blood transfusion in patients undergoing hip fracture surgery. METHODS: This study enrolled 1355 patients who underwent hip fracture surgery at the Affiliated Hospital of Qingdao University from January 2016 to December 2021. Among all patients, 210 cases received postoperative blood transfusion. All patients were randomly divided into a training group and a testing group at a ratio of 7:3. In the training group, univariate and multivariate logistic regression analyses were used to determine independent risk factors for the postoperative transfusion. Then, based on these independent risk factors, tenfold cross-validation method was utilized to develop five machine learning models, including logistic, multilayer perceptron (MLP), extreme gradient boosting (XGBoost), random forest (RF), and support vector machine (SVM). The receiver operating characteristic (ROC) curve, area under ROC curve (AUC), and Matthews correlation coefficient (MCC) were generated to evaluate the performance of the models. Calibration plot and decision curve analysis (DCA) were used to test the performance, stability, and clinical applicability of the models. The models were validated using the testing group; and the ROC curve, MCC, calibration plot, and DCA curves were also generated to validate the performance, stability, and clinical applicability of the models. To further verify the robustness of the model, we randomly grabbed 70% of the samples in the testing set, performed 1000 iterations, and calculated the AUC and confidence interval of the five models. Finally, we used SHapley Additive exPlanations (SHAP) to explain these models. RESULTS: Multivariate logistic regression analysis showed that there were 8 independent risk factors, including age, blood transfusion history, albumin (ALB), globulin (GLO), total bilirubin (TBIL), indirect bilirubin (IBIL), hemoglobin (HB), and blood loss > 200 ml. We finally selected five independent risk factors including HB, GLO, age, IBIL, and blood loss > 200 ml. Based on these five independent risk factors, we generated six characteristic variables, namely HB, HB × HB, HB × blood loss, GLO × HB, age, age × IBIL, and established five machine learning models using a tenfold cross-validation method. In the training group, the AUC values of logistic, RF, MLP, SVM, and XGB were 0.9320, 0.8911, 0.9327, 0.9225, and 0.8825, respectively, and the average AUC was 0.9122 ± 0.0212. The MCC values were 0.65, 0.77, 0.65, 0.66, and 0.68, respectively, and the calibration plot and DCA performed well. In the testing group the AUC values of logistic, RF, MLP, SVM, and XGB were 0.8483, 0.7978, 0.8576, 0.8598, and 0.8216, respectively. The average AUC was 0.8370 ± 0.0238, and the MCC values were 0.41, 0.35, 0.40, 0.41, and 0.41, respectively. The calibration plot and DCA in the testing group also showed good performance. The AUC values and confidence intervals of the 1000-iteration model were: logistic (AUC, min confidence interval [CI]-max confidence interval [CI] 0.848, 0.804-0.903), RF (AUC, minCI-maxCI 0.797, 0.734-0.857), MLP (AUC, minCI-maxCI 0.858, 0.812-0.902), SVM (AUC, minCI-maxCI 0.859, 0.819-0.910), and XGB (AUC, minCI-maxCI 0.821, 0.764-0.894). The model performed well. Finally, according to SHAP, among all five models, HB played the most important role in model prediction and interpretation. CONCLUSION: The five models we developed all performed well in predicting the likelihood of blood transfusion after hip fracture surgery. Therefore, we believed that the prediction model based on machine learning had great application prospects in clinical practice, which could help clinicians better predict the risk of blood transfusion after hip fracture surgery.

An effector of Erysiphe necator translocates to chloroplasts and plasma membrane to suppress host immunity in grapevine.

The powdery mildew (Erysiphe necator) is a prevalent pathogen hampering grapevine growth in the vineyard. An arsenal of candidate secreted effector proteins (CSEPs) was encoded in the E. necator genome, but it is largely unclear what role CSEPs plays during the E. necator infection. In the present study, we identified a secreted effector CSEP080 of E. necator, which was located in plant chloroplasts and plasma membrane. Transient expressing CSEP080 promotes plant photosynthesis and inhibits INF1-induced cell death in tobacco leaves. We found that CSEP080 was a necessary effector for the E. necator pathogenicity, which interacted with grapevine chloroplast protein VviB6f (cytochrome b6-f complex iron-sulfur subunit), affecting plant photosynthesis. Transient silencing VviB6f increased the plant hydrogen peroxide production, and the plant resistance to powdery mildew. In addition, CSEP080 manipulated the VviPE (pectinesterase) to promote pectin degradation. Our results demonstrated the molecular mechanisms that an effector of E. necator translocates to host chloroplasts and plasma membrane, which suppresses with the grapevine immunity system by targeting the chloroplast protein VviB6f to suppress hydrogen peroxide accumulation and manipulating VviPE to promote pectin degradation.

Saikosaponin B2 ameliorates depression-induced microglia activation by inhibiting ferroptosis-mediated neuroinflammation and ER stress.

ETHNOPHARMACOLOGICAL RELEVANCE: Saikosaponins B2 (SSB2) is one of the main active components isolated from Radix Bupleuri (Bupleurum chinense DC.), a herb widely used of traditional Chinese medicine. It has been used for the treatment of depression for more than two thousand years. However, the molecular mechanisms remain to be determined. AIM OF THE STUDY: In this study, we evaluated the anti-inflammatory effect and elucidated underlying molecular mechanisms of SSB2 in LPS-induced primary microglia and CUMS-induced mice model of depression. METHOD: The effects of SSB2 treatment were investigated both in vitro and in vivo. The chronic unpredictable mild stimulation (CUMS) procedure was applied to establish the animal model of depression. Behavioural tests were used to evaluate the depressive-like behaviors in CUMS-exposed mice, including sucrose preference test, open field test, tail suspension test, and forced swimming test. The GPX4 gene of microglia was silenced using shRNA, and inflammatory cytokines were determined by Western Blot and immunofluorescence analysis. Endoplasmic reticulum stress and ferroptosis-related markers were detected by qPCR, flow cytometry and confocal microscopy. RESULT: SSB2 reversed depressive-like behaviours in CUMS-exposed mice and relieved central neuroinflammation and ameliorated hippocampal neural damage. SSB2 alleviated LPS-induced activation of microglia through the TLR4/NF-κB pathway. LPS-induced ferroptosis, with increased levels of ROS, intracellular Fe2+, mitochondrial membrane potential, lipid peroxidation, GSH, SLC7A11, FTH, GPX4 and Nrf2, and decreased transcription levels of ACSL4 and TFR1, was attenuated with SSB2 treatment in primary microglia cells. GPX4 knockdown activated ferroptosis, induced endoplasmic reticulum (ER) stress, and abrogated the protective effects of SSB2. Further, SSB2 attenuated ER stress, balanced calcium homeostasis, reduced lipid peroxidation and intracellular Fe2+ content by regulating the level of intracellular Ca2+. CONCLUSIONS: Our study suggested that SSB2 treatment can inhibit ferroptosis, maintain calcium homeostasis, relieve endoplasmic reticulum stress and attenuate central neuroinflammation. SSB2 exhibited anti-ferroptosis and anti-neuroinflammatory effects through the TLR4/NF-κB pathway in a GPX4-dependent manner.

Multifunctional Ternary Hybrid Hydrogel Sensor Prepared via the Synergistic Stabilization Effect.

Since highly stretchable hydrogels have demonstrated their promising applications in flexible tactile sensors and wearable devices, the current challenge has been imposed on stretchable and multifunctional electronics. Here, we report a multifunctional sensor composed of a liquid metal (LM) nanodroplet-adhered self-assembled polymeric network, anionic carboxymethylcellulose (CMC), and cationic polyacrylamide (PAAm). The synergistic effect, zeta potential reduction, by CMC and macromolecules enveloped by LM contributes to the stabilization of the ternary system during preparation and, thus, the homogenization of the products. By engineering and optimizing the ternary hybrid hydrogels, excellent extensibility (tensile strain near 300%), readily reversible hysteresis loops, and accessible deformability (low modulus of 104 Pa) are afforded. The fabricated sensor exhibits a high tensile strain gauge factor of around 0.7 and a high compressive stress sensitivity of up to 0.12 kPa-1, a fast response time below 125 ms, and a high stability and precision in usage. In a series of practical scenarios, the assembled sensor displays distinguished abilities to monitor bodily motions, record electrocardiograms, authenticate handwriting, discern temperature, and infer materials, making them highly promising for multifunctional intelligent soft sensing.

Anisotropic, Flexible Wood Hydrogels and Wrinkled, Electrodeposited Film Electrodes for Highly Sensitive, Wide-Range Pressure Sensing.

Biological muscles generally possess well-aligned muscle fibers and thus excellent strength and toughness. Inspired by their microstructure, tough wood hydrogels with a preserved unique alignment of cellulose fibers, mechanical anisotropy, and desirable flexibility were developed by introducing chemically and ionically cross-linked poly(acrylic acid) (PAA) into the abundant pores of delignified wood. PAA chains well infiltrated the parallelly aligned cellulose fibers of wood and formed a layer-by-layer network structure, resulting in strong, elastic tangential, and radial wood hydrogel slices. The tangential slices had a high compressive strength of 1.73 MPa and a maximum strain at fracture of 69.4%, while those of the radial slices were 0.6 MPa and 47.0%. After sandwiching the radial and tangential hydrogel slices with reduced graphene oxide (rGO) film electrodes into capacitive pressure sensors (CPSs), the tangential CPS displayed the most desired, gradient sensitivity values in the whole stress range. Additionally, the wrinkling treatment of the rGO electrode greatly improved the capacitance responsiveness toward pressure. The real-time sensing stress values of our tangential CPS during monitoring practical human activities were calculated in the range of 0.1-1.3 MPa, demonstrating the achievement of ultrafast, highly sensitive, and wide-stress-range detection for potential applications in human-machine interfaces.