Multidimensional Deep Ultraviolet (DUV) Synapses Based on Organic/Perovskite Semiconductor Heterojunction Transistors for Antispoofing Facial Recognition Systems.

Reliably discerning real human faces from fake ones, known as antispoofing, is crucial for facial recognition systems. While neuromorphic systems offer integrated sensing-memory-processing functions, they still struggle with efficient antispoofing techniques. Here we introduce a neuromorphic facial recognition system incorporating multidimensional deep ultraviolet (DUV) optoelectronic synapses to address these challenges. To overcome the complexity and high cost of producing DUV synapses using traditional wide-bandgap semiconductors, we developed a low-temperature (≤70 °C) solution process for fabricating DUV synapses based on PEA2PbBr4/C8-BTBT heterojunction field-effect transistors. This method enables the large-scale (4-in.), uniform, and transparent production of DUV synapses. These devices respond to both DUV and visible light, showing multidimensional features. Leveraging the unique ability of the multidimensional DUV synapse (MDUVS) to discriminate real human skin from artificial materials, we have achieved robust neuromorphic facial recognition with antispoofing capability, successfully identifying genuine human faces with an accuracy exceeding 92%.

Enzymatic Galvanic Redox Potentiometry for In Vivo Biosensing.

Redox potentiometry has emerged as a new platform for in vivo sensing, with improved neuronal compatibility and strong tolerance against sensitivity variation caused by protein fouling. Although enzymes show great possibilities in the fabrication of selective redox potentiometry, the fabrication of an enzyme electrode to output open-circuit voltage (EOC) with fast response remains challenging. Herein, we report a concept of novel enzymatic galvanic redox potentiometry (GRP) with improved time response coupling the merits of the high selectivity of enzyme electrodes with the excellent biocompatibility and reliability of GRP sensors. With a glucose biosensor as an illustration, we use flavin adenine dinucleotide-dependent glucose dehydrogenase as the recognition element and carbon black as the potential relay station to improve the response time. We find that the enzymatic GRP biosensor rapidly responds to glucose with a good linear relationship between EOC and the logarithm of glucose concentration within a range from 100 µM to 2.65 mM. The GRP biosensor shows high selectivity over O2 and coexisting neurochemicals, good reversibility, and sensitivity and can in vivo monitor glucose dynamics in rat brain. We believe that this study will pave a new platform for the in vivo potentiometric biosensing of chemical events with high reliability.

Overexpression of OsACL5 triggers environmentally-dependent leaf rolling and reduces grain size in rice.

Major polyamines include putrescine, spermidine, spermine and thermospermine, which play vital roles in growth and adaptation against environmental changes in plants. Thermospermine (T-Spm) is synthetised by ACL5. The function of ACL5 in rice is still unknown. In this study, we used a reverse genetic strategy to investigate the biological function of OsACL5. We generated several knockout mutants by pYLCRISPR/Cas9 system and overexpressing (OE) lines of OsACL5. Interestingly, the OE plants exhibited environmentally-dependent leaf rolling, smaller grains, lighter 1000-grain weight and reduction in yield per plot. The area of metaxylem vessels of roots and leaves of OE plants were significantly smaller than those of WT, which possibly caused reduction in leaf water potential, resulting in leaf rolling with rise in the environmental temperature and light intensity and decrease in humidity. Additionally, the T-Spm contents were markedly increased by over ninefold whereas the ethylene evolution was reduced in OE plants, suggesting that T-Spm signalling pathway interacts with ethylene pathway to regulate multiple agronomic characters. Moreover, the osacl5 exhibited an increase in grain length, 1000-grain weight, and yield per plot. OsACL5 may affect grain size via mediating the expression of OsDEP1, OsGS3 and OsGW2. Furthermore, haplotypes analysis indicated that OsACL5 plays a conserved function on regulating T-Spm levels during the domestication of rice. Our data demonstrated that identification of OsACL5 provides a theoretical basis for understanding the physiological mechanism of T-Spm which may play roles in triggering environmentally dependent leaf rolling; OsACL5 will be an important gene resource for molecular breeding for higher yield.

Alleviated NCOA4-mediated ferritinophagy protected RA FLSs from ferroptosis in lipopolysaccharide-induced inflammation under hypoxia.

OBJECTIVE: Ferroptosis is a reactive oxygen species (ROS)- and iron-dependent form of non-apoptotic cell death process. Previous studies have demonstrated that ferroptosis participates in the development of inflammatory arthritis. However, the role of ferroptosis in rheumatoid arthritis (RA) inflammatory hypoxic joints remains unclear. This study sought to explore the underlying mechanism of ferroptosis on lipopolysaccharide (LPS)-induced RA fibroblast-like synoviocytes (FLSs). METHODS: FLSs, isolated from patients with RA, were treated with LPS and ferroptosis inducer (erastin and RSL-3), and ferroptosis inhibitor (Fer-1 and DFO), respectively. The cell viability was measured by CCK-8. The cell death was detected by flow cytometer. The proteins level were tested by Western blot. The cytosolic ROS and lipid peroxidation were determined using DCFH-DA and C11-BODIPY581/591 fluorescence probes, respectively. The small interfering RNA (siRNA) was used to knock down related proteins. The levels of malondialdehyde (MDA), 4-hydroxynonenal (4-HNE), iron, inflammatory cytokines (IL6 and IL8), and LDH were analyzed by commercial kits. RESULTS: Ferroptosis was activated by LPS in RA FLS with increased cellular damage, ROS and lipid peroxidation, intracellular Fe and IL8, which can be further amplified by ferroptosis inducer (erastin and RSL-3) and inhibited by ferroptosis inhibitor (Fer-1 and DFO). Mechanistically, LPS triggered ferroptosis via NCOA4-mediated ferritinophagy in RA FLSs, and knockdown of NCOA4 strikingly prevent the process of ferroptosis. Intriguingly, LPS-induced RA FLSs became insensitive to ferroptosis and NCOA4-mediated ferritinophagy under hypoxia compared with normoxia. Knockdown of HIF-1α reverted ferroptosis and ferritinophagy evoking by LPS-induced RA FLSs inflammation under hypoxia. In addition, low dose of auranofin (AUR) induced re-sensitization of ferroptosis and ferritinophagy through inhibiting the expression of HIF-1α under hypoxia. CONCLUSIONS: NCOA4-mediated ferritinophagy was a key driver of ferroptosis in inflammatory RA FLSs. The suppression of NCOA4-mediated ferritinophagy protected RA FLSs from ferroptosis in LPS-induced inflammation under hypoxia. Targeting HIF-1α/NCOA4 and ferroptosis could be an effective and valuable therapeutic strategy for synovium hyperplasia in the patients with RA.

Application of machine learning for mortality prediction in patients with candidemia: Feasibility verification and comparison with clinical severity scores.

BACKGROUND: Clinical severity scores, such as acute physiology, age, chronic health evaluation II (APACHE II), sequential organ failure assessment (SOFA), Pitt Bacteremia Score (PBS), and European Confederation of Medical Mycology Quality (EQUAL) score, may not reliably predict candidemia prognosis owing to their prespecified scorings that can limit their adaptability and applicability. OBJECTIVES: Unlike those fixed and prespecified scorings, we aim to develop and validate a machine learning (ML) approach that is able to learn predictive models adaptively from available patient data to increase adaptability and applicability. METHODS: Different ML algorithms follow different design philosophies and consequently, they carry different learning biases. We have designed an ensemble meta-learner based on stacked generalisation to integrate multiple learners as a team to work at its best in a synergy to improve predictive performances. RESULTS: In the multicenter retrospective study, we analysed 512 patients with candidemia from January 2014 to July 2019 and compared a stacked generalisation model (SGM) with APACHE II, SOFA, PBS and EQUAL score to predict the 14-day mortality. The cross-validation results showed that the SGM significantly outperformed APACHE II, SOFA, PBS, and EQUAL score across several metrics, including F1-score (0.68, p < .005), Matthews correlation coefficient (0.54, p < .05 vs. SOFA, p < .005 vs. the others) and the area under the curve (AUC; 0.87, p < .005). In addition, in an independent external test, the model effectively predicted patients' mortality in the external validation cohort, with an AUC of 0.77. CONCLUSIONS: ML models show potential for improving mortality prediction amongst patients with candidemia compared to clinical severity scores.

Subtle Structural Differences Affect the Inhibitory Potency of RGD-Containing Cyclic Peptide Inhibitors Targeting SPSB Proteins.

The SPRY domain-containing SOCS box proteins SPSB1, SPSB2, and SPSB4 utilize their SPRY/B30.2 domain to interact with a short region in the N-terminus of inducible nitric oxide synthase (iNOS), and recruit an E3 ubiquitin ligase complex to polyubiquitinate iNOS, resulting in the proteasomal degradation of iNOS. Inhibitors that can disrupt the endogenous SPSB-iNOS interactions could be used to augment cellular NO production, and may have antimicrobial and anticancer activities. We previously reported the rational design of a cyclic peptide inhibitor, cR8, cyclo(RGDINNNV), which bound to SPSB2 with moderate affinity. We, therefore, sought to develop SPSB inhibitors with higher affinity. Here, we show that cyclic peptides cR7, cyclo(RGDINNN), and cR9, cyclo(RGDINNNVE), have ~6.5-fold and ~2-fold, respectively, higher SPSB2-bindng affinities than cR8. We determined high-resolution crystal structures of the SPSB2-cR7 and SPSB2-cR9 complexes, which enabled a good understanding of the structure-activity relationships for these cyclic peptide inhibitors. Moreover, we show that these cyclic peptides displace full-length iNOS from SPSB2, SPSB1, and SPSB4, and that their inhibitory potencies correlate well with their SPSB2-binding affinities. The strongest inhibition was observed for cR7 against all three iNOS-binding SPSB proteins.

Approaching Sustainable Lithium-Ion Batteries through Voltage-Responsive Smart Prelithiation Separator with Surface-Engineered Sacrificial Lithium Agents.

The surge in lithium-ion batteries has heightened concerns regarding metal resource depletion and the environmental impact of spent batteries. Battery recycling has become paramount globally, but conventional techniques, while effective at extracting transition metals like cobalt and nickel from cathodes, often overlook widely used spent LiFePO4 due to its abundant and low-cost iron content. Direct regeneration, a promising approach for restoring deteriorated cathodes, is hindered by practicality and cost issues despite successful methods like solid-state sintering. Hence, a smart prelithiation separator based on surface-engineered sacrificial lithium agents is proposed. Benefiting from the synergistic anionic and cationic redox, the prelithiation separator can intelligently release or intake active lithium via voltage regulation. The staged lithium replenishment strategy was implemented, successfully restoring spent LiFePO4's capacity to 163.7 mAh g-1 and a doubled life. Simultaneously, the separator can absorb excess active lithium up to approximately 600 mAh g-1 below 2.5 V to prevent over-lithiation of the cathode This innovative, straightforward, and cost-effective strategy paves the way for the direct regeneration of spent batteries, expanding the possibilities in the realm of lithium-ion battery recycling.

Stability Enhancement of Galvanic Redox Potentiometry by Optimizing the Redox Couple in Counterpart Poles.

Potentiometry based on the galvanic cell mechanism, i.e., galvanic redox potentiometry (GRP), has recently emerged as a new tool for in vivo neurochemical sensing with high neuronal compatibility and good sensing property. However, the stability of open circuit voltage (EOC) outputting remains to be further improved for in vivo sensing application. In this study, we find that the EOC stability could be enhanced by adjusting the sort and the concentration ratio of the redox couple in the counterpart pole (i.e., indicating electrode) of GRP. With dopamine (DA) as the sensing target, we construct a spontaneously powered single-electrode-based GRP sensor (GRP2.0) and investigate the correlation between the stability and the redox couple used in the counterpart pole. Theoretical consideration suggests that the EOC drift is minimum when the concentration ratio of the oxidized form (O1) to the reduced form (R1) of the redox species in the backfilled solution is 1:1. The experimental results demonstrate that, compared with other redox species (i.e., dissolved O2 at 3 M KCl, potassium ferricyanide (K3Fe(CN)6), and hexaammineruthenium(III) chloride (Ru(NH3)6Cl3)) used as the counterpart pole, potassium hexachloroiridate(IV) (K2IrCl6) exhibits better chemical stability and outputs more stable EOC. As a result, when IrCl62-/3- with the concentration ratio of 1:1 is used as the counterpart, GRP2.0 displays not only an excellent EOC stability (i.e., 3.8 mV drifting during 2200 s for in vivo recording) but also small electrode-to-electrode variation (i.e., the maximum EOC variation between four electrodes is 2.7 mV). Upon integration with the electrophysiology, GRP2.0 records a robust DA release, accompanied by a burst of neural firing, during the optical stimulation. This study paves a new avenue to stable neurochemical sensing in vivo.

A Methodology of Fabricating Novel Electrodes for Semiconductor Devices: Doping and Van der Waals Integrating Organic Semiconductor Films.

Electrodes are indispensable components in semiconductor devices, and now are mainly made from metals, which are convenient for use but not ideal for emerging technologies such as bioelectronics, flexible electronics, or transparent electronics. Here the methodology of fabricating novel electrodes for semiconductor devices using organic semiconductors (OSCs) is proposed and demonstrated. It is shown that polymer semiconductors can be heavily p- or n-doped to achieve sufficiently high conductivity for electrodes. In contrast with metals, the doped OSC films (DOSCFs) are solution-processable, mechanically flexible, and have interesting optoelectronic properties. By integrating the DOSCFs with semiconductors through van der Waals contacts different kinds of semiconductor devices can be constructed. Importantly, these devices exhibit higher performance than their counterparts with metal electrodes, and/or excellent mechanical or optical properties that are unavailable in metal-electrode devices, suggesting the superiority of DOSCF electrodes. Given the existing large amount of OSCs, the established methodology can provide abundant electrode choices to meet the demand of various emerging devices.

Purinergic Receptor P2Y6 Is a Negative Regulator of NK Cell Maturation and Function.

NK cells are critical innate immune cells that target the tumor cells and cancer-initiating cells and clear viruses by producing cytokines and cytotoxic granules. However, the role of the purinergic receptor P2Y6 in the NK cells remains largely unknown. In this study, we discovered that the expression of P2Y6 was decreased upon the activation of the NK cells. Moreover, in the P2Y6-deficient mice, we found that the deficiency of P2Y6 promoted the development of the NK precursor cells into immature NK and mature NK cells. We also found that the P2Y6 deficiency increased, but the P2Y6 receptor agonist UDP or UDP analog 5-OMe-UDP decreased the production of IFN-γ in the activated NK cells. Furthermore, we demonstrated that the P2Y6-deficient NK cells exhibited stronger cytotoxicity in vitro and antimetastatic effects in vivo. Mechanistically, P2Y6 deletion promoted the expression of T-bet (encoded by Tbx21), with or without the stimulation of IL-15. In the absence of P2Y6, the levels of phospho-serine/threonine kinase and pS6 in the NK cells were significantly increased upon the stimulation of IL-15. Collectively, we demonstrated that the P2Y6 receptor acted as a negative regulator of the NK cell function and inhibited the maturation and antitumor activities of the NK cells. Therefore, inhibition of the P2Y6 receptor increases the antitumor activities of the NK cells, which may aid in the design of innovative strategies to improve NK cell-based cancer therapy.

Interactions of Nanomaterials with Gut Microbiota and Their Applications in Cancer Therapy.

Cancer treatment is a challenge by its incredible complexity. As a key driver and player of cancer, gut microbiota influences the efficacy of cancer treatment. Modalities to manipulate gut microbiota have been reported to enhance antitumor efficacy in some cases. Nanomaterials (NMs) have been comprehensively applied in cancer diagnosis, imaging, and theranostics due to their unique and excellent properties, and their effectiveness is also influenced by gut microbiota. Nanotechnology is capable of targeting and manipulating gut microbiota, which offers massive opportunities to potentiate cancer treatment. Given the complexity of gut microbiota-host interactions, understanding NMs-gut interactions and NMs-gut microbiota interactions are important for applying nanotechnologies towards manipulating gut microbiota in cancer prevention and treatment. In this review, we provide an overview of NMs-gut interactions and NMs-gut microbiota interactions and highlight the influences of gut microbiota on the diagnosis and treatment effects of NMs, further illustrating the potential of nanotechnologies in cancer therapy. Investigation of the influences of NMs on cancer from the perspective of gut microbiota will boost the prospect of nanotechnology intervention of gut microbiota for cancer therapy.

Potential Controllable Redox Couple for Mild and Efficient Lithium Recovery from Spent Batteries.

The prosperity of the lithium-ion battery market is dialectically accompanied by the depletion of corresponding resources and the accumulation of spent batteries. It is an urgent priority to develop green and efficient battery recycling strategies for helping ease resources and environmental pressures at the current stage. Here, we propose a mild and efficient lithium extracting strategy based on potential controllable redox couples. Active lithium in the spent battery without discharging is extracted using a series of tailored aprotic solutions comprised of polycyclic aromatic hydrocarbons and ethers. This ensures a safe yet efficient recycling process with nearly ≈100 % lithium recovery. We further investigate the Li+ -electron concerted redox reactions and the effect of solvation structure on kinetics during the extraction, and broaden the applicability of the Li-PAHs solution. This work can stimulate new inspiration for designing novel solutions to meet efficient and sustainable demands in recycling batteries.

Mitigating Electron Leakage of Solid Electrolyte Interface for Stable Sodium-Ion Batteries.

The interfacial stability is highly responsible for the longevity and safety of sodium ion batteries (SIBs). However, the continuous solid-electrolyte interphase(SEI) growth would deteriorate its stability. Essentially, the SEI growth is associated with the electron leakage behavior, yet few efforts have tried to suppress the SEI growth, from the perspective of mitigating electron leakage. Herein, we built two kinds of SEI layers with distinct growth behaviors, via the additive strategy. The SEI physicochemical features (morphology and componential information) and SEI electronic properties (LUMO level, band gap, electron work function) were investigated elaborately. Experimental and calculational analyses showed that, the SEI layer with suppressed growth delivers both the low electron driving force and the high electron insulation ability. Thus, the electron leakage is mitigated, which restrains the continuous SEI growth, and favors the interface stability with enhanced electrochemical performance.

Double-Barreled Micropipette Enables Neuron-Compatible In Vivo Analysis.

The sensors based on ion transport in a confined nano-/microchannel (i.e., iontronic sensors) have brought new opportunities for in vivo neurochemical assay, especially for electroinactive molecules. However, the interference on spontaneous neuronal activity induced by the electric field around the sensors has not been addressed. Here, the electric field distribution with a double-barreled micropipette was shrunk and quantified by finite element simulation, which can explain and minimize the influence on spontaneous neuronal activity. The parameters affecting the electric field distribution, including the pore size, applied voltage, and angle degree, were studied to balance the sensitivity and interference on spontaneous neuronal activity. The double-barreled micropipette, as a pH sensor with high selectivity and sensitivity, has been successfully applied to real-time pH sensing in rat brain. This study offers a new way for in vivo monitoring neurochemical dynamics with neuron-compatibility.

P2Y6 Deficiency Enhances Dendritic Cell-Mediated Th1/Th17 Differentiation and Aggravates Experimental Autoimmune Encephalomyelitis.

Dendritic cells (DCs) are essential APCs and play a crucial role in initiating and regulating the adaptive immune response. In this study, we have reported that P2Y6, a member of G protein-coupled receptors, inhibits the maturation and activation of DCs via suppressing the activation of the transcription factor NF-κB. Furthermore, loss of P2Y6 does not impact T cells homeostasis in the steady-state. However, in vitro studies show that P2Y6 signaling inhibits the production of IL-12 and IL-23 and the polarization of Th1 and Th17 subsets mediated by DCs. In addition, we find that mice lacking P2Y6 develop more severe experimental autoimmune encephalomyelitis compared with wild-type mice. Our results indicate that P2Y6 functions as a pivotal regulator on DC maturation, and the loss of P2Y6 results in the aggravated experimental autoimmune encephalomyelitis, which suggests that P2Y6 may play a pivotal role in the pathogenesis of autoimmune diseases.

A modular and self-adjuvanted multivalent vaccine platform based on porcine circovirus virus-like nanoparticles.

BACKGROUND: Virus-like particles (VLPs) are supramolecular structures composed of multiple protein subunits and resemble natural virus particles in structure and size, making them highly immunogenic materials for the development of next-generation subunit vaccines. The orderly and repetitive display of antigenic epitopes on particle surface allows efficient recognition and cross-link by B cell receptors (BCRs), thereby inducing higher levels of neutralizing antibodies and cellular immune responses than regular subunit vaccines. Here, we present a novel multiple antigen delivery system using SpyCatcher/Spytag strategy and self-assembled VLPs formed by porcine circovirus type 2 (PCV2) Cap, a widely used swine vaccine in solo. RESULTS: Cap-SC, recombinant Cap with a truncated SpyCatcher polypeptide at its C-terminal, self-assembled into 26-nm VLPs. Based on isopeptide bonds formed between SpyCatcher and SpyTag, classical swine fever virus (CSFV) E2, the antigen of interest, was linked to SpyTag and readily surface-displayed on SpyCatcher decorated Cap-SC via in vitro covalent conjugation. E2-conjugated Cap VLPs (Cap-E2 NPs) could be preferentially captured by antigen presenting cells (APCs) and effectively stimulate APC maturation and cytokine production. In vivo studies confirmed that Cap-E2 NPs elicited an enhanced E2 specific IgG response, which was significantly higher than soluble E2, or the admixture of Cap VLPs and E2. Moreover, E2 displayed on the surface did not mask the immunodominant epitopes of Cap-SC VLPs, and Cap-E2 NPs induced Cap-specific antibody levels and neutralizing antibody levels comparable to native Cap VLPs. CONCLUSION: These results demonstrate that this modularly assembled Cap-E2 NPs retains the immune potential of Cap VLP backbone, while the surface-displayed antigen significantly elevated E2-induced immune potency. This immune strategy provides distinctly improved efficacy than conventional vaccine combination. It can be further applied to the development of dual or multiple nanoparticle vaccines to prevent co-infection of PCV2 and other swine pathogens.

Dysbiosis of gut microbiota and intestinal damage in mice induced by a single intravenous exposure to CdTe quantum dots at low concentration.

Although quantum dots (QDs) have shown great potential for various biomedical applications, their potential toxicity still needs to be comprehensively investigated. Previous studies showed that intravenous exposure of CdTe QDs at low concentration did not lead to obvious in vivo toxicity in the long term. However, the influence of CdTe QDs on the gut microbiota and the intestine is still unknown. Here, we explored whether single intravenous injection of CdTe QDs at low concentration can affect the gut microbiota and intestine of mice in short term. The results showed that CdTe QDs caused an imbalance of gut microbiota, especially the rapid increase in Lactobacillus on day 1 post-treatment. Meanwhile, the intestine exhibited the promotion of oxidative stress, inflammatory response, and hemorrhaging on days 5 and 15. These results demonstrate that the gut microbiota and the intestine are very sensitive to the toxicity of low-concentration CdTe QDs. This study provides further insight and method for the biosafety evaluation of nanomaterials.

Electrical synapses between mushroom body neurons are critical for consolidated memory retrieval in Drosophila.

Electrical synapses between neurons, also known as gap junctions, are direct cell membrane channels between adjacent neurons. Gap junctions play a role in the synchronization of neuronal network activity; however, their involvement in cognition has not been well characterized. Three-hour olfactory associative memory in Drosophila has two components: consolidated anesthesia-resistant memory (ARM) and labile anesthesia-sensitive memory (ASM). Here, we show that knockdown of the gap junction gene innexin5 (inx5) in mushroom body (MB) neurons disrupted ARM, while leaving ASM intact. Whole-mount brain immunohistochemistry indicated that INX5 protein was preferentially expressed in the somas, calyxes, and lobes regions of the MB neurons. Adult-stage-specific knockdown of inx5 in αß neurons disrupted ARM, suggesting a specific requirement of INX5 in αß neurons for ARM formation. Hyperpolarization of αß neurons during memory retrieval by expressing an engineered halorhodopsin (eNpHR) also disrupted ARM. Administration of the gap junction blocker carbenoxolone (CBX) reduced the proportion of odor responsive αß neurons to the training odor 3 hours after training. Finally, the α-branch-specific 3-hour ARM-specific memory trace was also diminished with CBX treatment and in inx5 knockdown flies. Altogether, our results suggest INX5 gap junction channels in αß neurons for ARM retrieval and also provide a more detailed neuronal mechanism for consolidated memory in Drosophila.

Periprosthetic Joint Infection Prediction via Machine Learning: Comprehensible Personalized Decision Support for Diagnosis.

BACKGROUND: The criteria outlined in the International Consensus Meeting (ICM) in 2018, which were prespecified and fixed, have been commonly practiced by clinicians to diagnose periprosthetic joint infection (PJI). We developed a machine learning (ML) system for PJI diagnosis and compared it with the ICM scoring system to verify the feasibility of ML. METHODS: We designed an ensemble meta-learner, which combined 5 learning algorithms to achieve superior performance by optimizing their synergy. To increase the comprehensibility of ML, we developed an explanation generator that produces understandable explanations of individual predictions. We performed stratified 5-fold cross-validation on a cohort of 323 patients to compare the ML meta-learner with the ICM scoring system. RESULTS: Cross-validation demonstrated ML's superior predictive performance to that of the ICM scoring system for various metrics, including accuracy, precision, recall, F1 score, Matthews correlation coefficient, and area under receiver operating characteristic curve. Moreover, the case study showed that ML was capable of identifying personalized important features missing from ICM and providing interpretable decision support for individual diagnosis. CONCLUSION: Unlike ICM, ML could construct adaptive diagnostic models from the available patient data instead of making diagnoses based on prespecified criteria. The experimental results suggest that ML is feasible and competitive for PJI diagnosis compared with the current widely used ICM scoring criteria. The adaptive ML models can serve as an auxiliary system to ICM for diagnosing PJI.

Random RotBoost: An Ensemble Classification Method Based on Rotation Forest and AdaBoost in Random Subsets and Its Application to Clinical Decision Support.

In the era of bathing in big data, it is common to see enormous amounts of data generated daily. As for the medical industry, not only could we collect a large amount of data, but also see each data set with a great number of features. When the number of features is ramping up, a common dilemma is adding computational cost during inferring. To address this concern, the data rotational method by PCA in tree-based methods shows a path. This work tries to enhance this path by proposing an ensemble classification method with an AdaBoost mechanism in random, automatically generating rotation subsets termed Random RotBoost. The random rotation process has replaced the manual pre-defined number of subset features (free pre-defined process). Therefore, with the ensemble of the multiple AdaBoost-based classifier, overfitting problems can be avoided, thus reinforcing the robustness. In our experiments with real-world medical data sets, Random RotBoost reaches better classification performance when compared with existing methods. Thus, with the help from our proposed method, the quality of clinical decisions can potentially be enhanced and supported in medical tasks.