Search Results (123)

Background/Objectives: Cochlear implantation (CI) significantly enhances speech perception and quality of life in patients with severe-to-profound sensorineural hearing loss, yet outcomes vary substantially. Accurate preoperative prediction of CI outcomes remains challenging. This study aimed to develop and validate a machine learning model predicting postoperative speech recognition using a large, single-center dataset. Additionally, we compared model performance with expert clinical predictions to evaluate potential clinical utility. Methods: We retrospectively analyzed data from 2571 adult patients with postlingual hearing loss who received their cochlear implant between 2000 and 2022 at Hannover Medical School, Germany. A decision tree regression model was trained to predict monosyllabic (MS) word recognition score one to two years post-implantation using preoperative clinical variables (age, duration of deafness, preoperative MS score, pure tone average, onset type, and contralateral implantation status). Model evaluation was performed using a random data split (10%), a chronological future cohort (patients implanted after 2020), and a subset where experienced audiologists predicted outcomes for comparison. Results: The model achieved a mean absolute error (MAE) of 17.3% on the random test set and 17.8% on the chronological test set, demonstrating robust predictive performance over time. Compared to expert audiologist predictions, the model showed similar accuracy (MAE: 19.1% for the model vs. 18.9% for experts), suggesting comparable effectiveness. Conclusions: Our machine learning model reliably predicts postoperative speech outcomes and matches expert clinical predictions, highlighting its potential for supporting clinical decision-making. Future research should include external validation and prospective trials to further confirm clinical applicability. Full article

Background and Objectives: The transoral endoscopic thyroidectomy-vestibular approach (TOETVA) and parathyroidectomy-vestibular approach (TOEPVA) are scar-free alternatives to conventional surgery but are classified as clean-contaminated due to the oral incision, raising concerns about surgical site infections (SSIs). This study evaluates whether perioperative antibiotic prophylaxis (pABX) alone is sufficient compared to extended antibiotic prophylaxis (eABX) in preventing SSIs in TOET/PVA, particularly considering the surgical learning curve. Materials and Methods: A retrospective study analyzed 162 patients undergoing TOET/PVA at a single center from January 2018 to June 2024. Patients were divided into two groups: 82 received eABX (intravenous cefazolin preoperatively plus 7 days of oral amoxicillin/clavulanate), and 80 received pABX alone (intravenous cefazolin). The inclusion criteria included complete postoperative hemogram and C-reactive protein (CRP) records; exclusions comprised other surgical approaches or missing data. Outcomes included postoperative white blood cell (WBC) count, CRP levels, and complications (seroma, cellulitis, and flap perforation), defined using Centers for Disease Control and Prevention (CDC) guidelines. The statistical analysis comprised t-tests, chi-square tests, and logistic regression, adjusting for confounders like age and sex. Results: The postoperative WBC and CRP levels were significantly higher in the pABX group (p = 0.001), but all values remained within the laboratory normal limits. Complications were observed in 14 patients: seroma in 11, cellulitis in 2, and flap perforation in 1. Complications occurred more frequently in the eABX group but without statistical significance (p = 0.103). The duration of surgery was longer in the eABX group (117.93 ± 52.35 vs. 72.44 ± 22.54 min, p = 0.001) and was an independent predictor of complications (OR = 1.018, 95% CI: 1.006–1.031, p = 0.004). Conclusions: Perioperative antibiotic prophylaxis alone does not increase the risk of SSIs compared to extended prophylaxis in TOETVA. However, eABX may be prudent during the learning curve due to longer operative times and higher complication risks. Future prospective, randomized trials are needed to standardize prophylaxis regimens. Full article

The mirror-drawing task has been widely used in motor learning research to investigate procedural memory and movement control. However, studies have primarily focused on global performance measures such as movement time and the number of errors and lack insight into localized learning patterns. This case study aimed to analyze motor learning characteristics by combining traditional measures with entropy analysis, a method for capturing movement stability and complexity. Using a star-shaped figure divided into 12 segments, a single participant performed 100 trials of the mirror-drawing task. The movement coordinates were recorded at 60 Hz using a stylus on a mirrored tablet screen. The results showed that movement time decreased over the trials and entropy values showed an initial increase, followed by a decrease, suggesting exploratory behavior and subsequent stabilization. In particular, the interference side segments requiring complex visual–motor transformations showed prolonged instability and delayed control stabilization compared with the noninterference side segments. The integration of entropy analysis allowed a clearer visualization of the trial-and-error phases and movement instability, providing novel insights into the motor learning process. These findings, though limited to a single case, contribute to the understanding of adaptive movement control strategies and suggest that local learning properties should be considered in skill acquisition research. Full article

Multiple sclerosis (MS) is an autoimmune disease of the central nervous system affecting over 2.8 million people around the world. Artificial intelligence (AI) is becoming increasingly utilised in many areas, including patient care for MS. AI is revolutionising the diagnosis and treatment of MS by enhancing the accuracy and efficiency of both processes. AI algorithms, particularly those based on machine learning, are being used to analyse medical imaging data, such as MRI scans, to detect early signs of MS, monitor disease progression and assess patient treatment response with greater precision. AI can help identify subtle changes in the brain and spinal cord that may be missed by human clinicians, leading to earlier diagnosis and more personalised treatment plans. Additionally, AI is being employed to predict disease outcomes, which could allow clinicians to tailor therapies for individual patients based on their unique disease characteristics. In drug development, AI is accelerating the identification of potential therapeutic targets and the optimisation of clinical trial designs, potentially leading to faster development of new treatments for MS. AI is also playing a critical role in MS fundamental research by promoting efficient analysis of vast amounts of single-cell data. Through these advancements, AI could improve the overall management of MS, offering more timely interventions and better patient outcomes. In this review, we discuss these topics and whether the influence of AI on diagnosis, treatment and research of MS can change the future of this field. Full article

Conventional deep learning models rely heavily on the McCulloch–Pitts (MCP) neuron, limiting their interpretability and biological plausibility. The Dendritic Neuron Model (DNM) offers a more realistic alternative by simulating nonlinear and compartmentalized processing within dendritic branches, enabling efficient and transparent learning. While DNMs have shown strong performance in various tasks, their learning capacity at the single-neuron level remains underexplored. This paper proposes a Reinforced Dynamic-grouping Differential Evolution (RDE) algorithm to enhance synaptic plasticity within the DNM framework. RDE introduces a biologically inspired mutation-selection strategy and an adaptive grouping mechanism that promotes effective exploration and convergence. Experimental evaluations on benchmark classification tasks demonstrate that the proposed method outperforms conventional differential evolution and other evolutionary learning approaches in terms of accuracy, generalization, and convergence speed. Specifically, the RDE-DNM achieves up to 92.9% accuracy on the BreastEW dataset and 98.08% on the Moons dataset, with consistently low standard deviations across 30 trials, indicating strong robustness and generalization. Beyond technical performance, the proposed model supports societal applications requiring trustworthy AI, such as interpretable medical diagnostics, financial screening, and low-energy embedded systems. The results highlight the potential of RDE-driven DNMs as a compact and interpretable alternative to traditional deep models, offering new insights into biologically plausible single-neuron computation for next-generation AI. Full article

With the increasing demand for time-series analysis, driven by the proliferation of IoT devices and real-time data-driven systems, detecting change points in time series has become critical for accurate short-term prediction. The variability in patterns necessitates frequent analysis to sustain high performance by acquiring the hyperparameter. The Cumulative Sum (CUSUM) method, based on calculating the cumulative values within a time series, is commonly used for change detection due to its early detection of small drifts, simplicity, low computational cost, and robustness to noise. However, its effectiveness heavily depends on the hyperparameter configuration, as a single setup may not be universally suitable across the entire time series. Consequently, fine-tuning is often required to achieve optimal results, yet this selection process is traditionally performed through trial and error or prior expert knowledge, which introduces subjectivity and inefficiency. To address this challenge, several strategies have been proposed to facilitate hyperparameter optimizations, as traditional methods are impractical. Meta-learning-based techniques present viable alternatives for periodic hyperparameter optimization, enabling the selection of configurations that adapt to dynamic scenarios. This work introduces a meta-modeling scheme designed to automate the recommendation of hyperparameters for the CUSUM algorithm. Benchmark datasets from the literature were used to evaluate the proposed framework. The results indicate that this framework preserves high accuracy while significantly reducing time requirements compared to Grid Search and Genetic Algorithm optimization. Full article

Reinforcement learning is a machine learning framework that relies on a lot of trial-and-error processes to learn the best policy to maximize the cumulative reward through the interaction between the agent and the environment. In the actual use of this process, the computing resources possessed by a single user are limited so that the cooperation of multiple users are needed, but the joint learning of multiple users introduces the problem of privacy leakage. This research proposes a method to safely share the effort of multiple users in an encrypted state and perform the reinforcement learning with outsourcing service to reduce users calculations combined with the homomorphic properties of cryptographic algorithms and multi-key ciphertext fusion mechanism. The proposed scheme has provably security, and the experimental results show that it has an acceptable impact on performance while ensuring privacy protection. Full article

Ulster University’s School of Medicine embraces a problem-based learning (PBL) approach, yet crafting scenarios for this method poses challenges, requiring collaboration among medical and academic experts who are often difficult to convene. This obstacle can compromise scenario quality and ultimately impede students’ learning experiences. To address this issue, the school trialed the use of AI technology to develop a case scenario focusing on headaches caused by cerebral haemorrhage. The process involved a dialogue between a single “author” and ChatGPT, with their outputs combined into a complete clinical case adhering to the school’s standard template. Six experienced PBL tutors conducted quality checks on the scenario. The tutors did not immediately endorse its use, recommending further enhancements. Suggestions included updating terminology, names, spelling, and protocols to align with current best practices, providing additional explanations such as interventions and improvements post-initial stability, incorporating real scans instead of descriptions, reviewing symptoms and timelines for realism, and addressing comprehension issues by refraining from directly providing answers and including probing questions instead. From this trial, several valuable lessons were learned: AI can assist a single author in crafting medical scenarios, easing the challenges of organizing expert teams. However, the author’s role shifts to reviewing and enhancing depth, guided by a template, with clinician input crucial for authenticity. ChatGPT respects patient data privacy and confidentiality by abstaining from providing scanned images, and while AI can generate discussion questions for tutorials, it may require modification to enhance specificity and provoke critical thought. Furthermore, AI can generate multiple-choice questions and compile reading resources to support self-directed learning. Overall, adopting AI technology can improve efficiency in the case-writing process. Full article

Background/Objectives: T cell receptor fusion constructs (TRuCs), a next generation engineered T cell therapy, hold great promise. To accelerate the clinical development of these therapies, improving patient selection is a crucial pathway forward. Methods: We retrospectively analyzed 23 mesothelioma patients (85 target tumors) treated in a phase 1/2 single arm clinical trial (NCT03907852). Five imaging sites were involved, the settings for the evaluations were Blinded Independent Central Reviews (BICRs) with double reads. The reproducibility of 3416 radiomics and delta-radiomics (Δradiomics) was assessed. The univariate analysis evaluated correlations at the target tumor level with (1) tumor diameter response; (2) tumor volume response, according to the Quantitative Imaging Biomarker Alliance; and (3) the mean standard uptake value (SUV) response, as defined by the positron emission tomography response criteria in solid tumors (PERCISTs). A random forest model predicted the response of the target pleural tumors. Results: Tumor anatomical distribution was 55.3%, 17.6%, 14.1%, and 10.6% in the pleura, lymph nodes, peritoneum, and soft tissues, respectively. Radiomics/Δradiomics reproducibility differed across tumor localizations. Radiomics were more reproducible than Δradiomics. In the univariate analysis, none of the radiomics/Δradiomics correlated with any response criteria. With an accuracy ranging from 0.75 to 0.9, three radiomics/Δradiomics were able to predict the response of target pleural tumors. Pivotal studies will require a sample size of 250 to 400 tumors. Conclusions: The prediction of responding target pleural tumors can be achieved using a machine learning-based radiomics/Δradiomics analysis. Tumor-specific reproducibility and the average values indicated that using tumor models to create an effective patient model would require combining several target tumor models. Full article

Under the background of dual carbon, the retrofitting of the equipment operation system of a refrigeration station and the optimization combination of its control system are significant for its efficient operation and energy saving. The single-direction variable flow technology is often used in the chilled water system in refrigeration stations nowadays. However, the single-direction variable flow technology cannot achieve both thermal balance and flow balance for the chiller system, which is unfavorable for improving energy efficiency and reliability. To improve the reliability and energy efficiency of the refrigeration station equipment, the bidirectional variable flow technology of primary and secondary chilled water pumps was presented. Meanwhile, the feasibility of fuzzy neural networks in bidirectional variable flow systems and their energy-saving effect were studied. Before the energy saving retrofit, the refrigeration station used traditional PID (proportional-integral-derivative) controllers, and the chilled water system used single-direction variable flow technology; After the energy-saving retrofit, the refrigeration station adopted a fuzzy neural network control algorithm to optimize the PID controller parameters, and at the same time, the chilled water system used bidirectional variable flow technology. Through a large number of trial calculations of the established neural network model, it was found that 2 hidden layers and 25 hidden layer nodes can achieve higher accuracy. Specifically, the controller of the central refrigeration station consists of a training neural network and a predictive neural network working in parallel. The task of training neural networks is to learn the relationship between different input parameters and the whole energy consumption. Then it serves as the excitation function of the prediction network. The function of the predictive neural network is to find the control parameters that minimize energy consumption. The application results showed that before and after the retrofit annual power consumption and energy-saving effects were very Significant. After the energy-saving retrofit of the refrigeration station, the energy saving is 422,775 KWh every year, the energy-saving rate is 11.67%, and the annual saving cost is about 0.3382 million yuan. The results demonstrated that bidirectional variable flow technology and its control methods were feasible, reasonable, and worthy of promotion. Full article

Background/Objectives: The diagnosis and treatment of endometriosis should be based on the best available evidence. Emphasising the risk of bias, the pyramid of evidence has the double-blind, randomised controlled trial and its meta-analyses on top. After the grading of all evidence by a group of experts, clinical guidelines are formulated using well-defined rules. Unfortunately, the impact of evidence-based medicine (EBM) on the management of endometriosis has been limited and, possibly, occasionally harmful. Methods: For this research, the inherent problems of diagnosis and treatment were discussed by a working group of endometriosis and EBM specialists, and the relevant literature was reviewed. Results: Most clinical decisions are multivariable, but randomized controlled trials (RCTs) cannot handle multivariability because adopting a factorial design would require prohibitively large cohorts and create randomization problems. Single-factor RCTs represent a simplification of the clinical reality. Heuristics and intuition are both important for training and decision-making in surgery; experience, Bayesian thinking, and learning from the past are seldom considered. Black swan events or severe complications and accidents are marginally discussed in EBM since trial evidence is limited for rare medical events. Conclusions: The limitations of EBM for managing endometriosis and the complementarity of multivariability, heuristics, Bayesian thinking, and experience should be recognized. Especially in surgery, the value of training and heuristics, as well as the importance of documenting the collective experience and of the prevention of complications, are fundamental. These additions to EBM and guidelines will be useful in changing the Wild West mentality of surgery resulting from the limited scope of EBM data because of the inherent multivariability, combined with the low number of similar interventions. Full article

Tumor heterogeneity is a significant factor influencing cancer treatment effectiveness and can arise from genetic, epigenetic, and phenotypic variations among cancer cells. Understanding how tumor heterogeneity impacts tumor evolution and therapy response can lead to more effective treatments and improved patient outcomes. Traditional bulk genomic approaches fail to provide insights into cellular-level events, whereas single-cell RNA sequencing (scRNA-seq) offers transcriptomic analysis at the individual cell level, advancing our understanding of tumor growth, progression, and drug response. However, implementing single-cell approaches in clinical trials involves challenges, such as obtaining high-quality cells, technical variability, and the need for complex computational analysis. Effective implementation of single-cell genomics in clinical trials requires a collaborative “Team Medicine” approach, leveraging shared resources, expertise, and workflows. Here, we describe key technical considerations in implementing the collection of research biopsies and lessons learned from integrating scRNA-seq into City of Hope’s clinical trial design, highlighting collaborative efforts between computational and clinical teams across breast, brain, and ovarian cancer studies to understand the composition, phenotypic state, and underlying resistance mechanisms within the tumor microenvironment. Full article

Background and Objectives: Assessing pain deception is challenging due to its subjective nature. The main goal of this study was to evaluate the diagnostic value of pain deception using machine learning (ML) analysis with the Minnesota Multiphasic Personality Inventory-2 (MMPI-2) scales, considering accuracy, precision, recall, and f1-score as diagnostic parameters. Materials and Methods: This study was a single-blinded, randomized controlled trial. Subjects were randomly allocated into a non-deception (ND) group and a deception (D) group. Pain deception, as a form of psychological intervention, was taught to subjects in the D group to deceive the physician. MMPI-2, Waddell’s sign, and salivary alpha-amylase (SAA) were also measured. For analyzing the MMPI-2, the XGBoost ML algorithm was applied. Results: Of a total of 96 participants, 50 and 46 were assigned to the ND group and the D group, respectively. In the logistic regression analysis, pain and MMPI-2 did not show diagnostic value. However, in the ML analysis, values of the selected MMPI-2 (sMMPI-2) scales related to pain deception showed an accuracy of 0.724, a precision of 0.692, a recall of 0.692, and an f1-score of 0.692. Conclusions: Using MMPI-2 test results, ML can diagnose pain deception better than the conventional logistic regression analysis method by considering different scales and patterns together. Full article

Background/Objectives: Populations in Mediterranean countries are abandoning the traditional Mediterranean diet (MD) and lifestyle, shifting towards unhealthier habits due to profound cultural and socioeconomic changes. The SWITCHtoHEALTHY project aims to demonstrate the effectiveness of a multi-component nutritional intervention to improve the adherence of families to the MD in three Mediterranean countries, thus prompting a dietary behavior change. Methods: A parallel, randomized, single-blinded, and controlled multicentric nutritional intervention study will be conducted over 3 months in 480 families with children and adolescents aged 3–17 years from Spain, Morocco, and Turkey. The multi-component intervention will combine digital interactive tools, hands-on educational materials, and easy-to-eat healthy snacks developed for this study. Through the developed SWITCHtoHEALTHY app, families will receive personalized weekly meal plans, which also consider what children eat at school. The engagement of all family members will be prompted by using a life simulation game. In addition, a set of activities and educational materials for adolescents based on a learning-through-playing approach will be codesigned. Innovative and sustainable plant-based snacks will be developed and introduced into the children’s dietary plan as healthy alternatives for between meals. By using a full-factorial design, families will be randomized into eight groups (one control and seven interventions) to test the independent and combined effects of each component (application and/or educational materials and/or snacks). The impact of the intervention on diet quality, economy, and the environment, as well as on classical anthropometric parameters and vital signs, will be assessed in three different visits. The COM-B behavioral model will be used to assess essential factors driving the behavior change. The main outcome will be adherence to the MD assessed through MEDAS in adults and KIDMED in children and adolescents. Conclusions: SWITCHtoHEALTHY will provide new insights into the use of sustained models for inducing dietary and lifestyle behavior changes in the family setting. It will facilitate generating, boosting, and maintaining the switch to a healthier MD dietary pattern across the Mediterranean area. Registered Trial, National Institutes of Health, ClinicalTrials.gov (NCT06057324). Full article

(1) Background: Alzheimer’s disease (AD) is a progressive neurodegenerative disorder marked by cognitive decline and functional impairment. The NeuronUP platform is a computer program whose main function is cognitive stimulation through three types of activities that change so that the user does not manage to learn it. This program provides opportunities to work on various domains, including activities of daily living (ADLs), social skills, and cognitive functions. The main objective of this randomized clinical trial was to assess the impact of integrating the NeuronUP platform with conventional occupational therapy to enhance or maintain cognitive, perceptual, and quality of life (QoL) abilities in people with AD compared to a control group. (2) Methods: A randomized, single-blind clinical trial was conducted. The sample was randomized using a software program, OxMar, which allowed the separation of the sample into a control group (CG) that received their conventional occupational therapy sessions and an experimental group (EG) that received therapy with NeuronUP, in addition to their conventional occupational therapy sessions. An eighteen-week intervention was conducted. (3) Results: The study included 20 participants, and significant differences were observed in most variables analyzed, indicating improvements after the intervention, particularly in measures of QoL and cognitive status. (4) Conclusions: Our findings demonstrate that an eighteen-week experimental protocol, incorporating the NeuronUP platform alongside conventional occupational therapy, led to improvements in cognitive status and QoL in older adults with AD. Thus, integrating the NeuronUP platform as a complementary tool to occupational therapy can be a valuable resource for enhancing the QoL of individuals with AD. However, due to the small sample size, further studies are needed to corroborate these findings. Full article