Prediction Model Research Articles

Artificial intelligence-enhanced biosurveillance for antimicrobial resistance in sub-Saharan Africa.

Antimicrobial resistance (AMR) remains a critical global health threat, with significant impacts on individuals and healthcare systems, particularly in low-income countries. By 2019, AMR was responsible for >4.9 million fatalities globally, and projections suggest this could rise to 10 million annually by 2050 without effective interventions. Sub-Saharan Africa (SSA) faces considerable challenges in managing AMR due to insufficient surveillance systems, resulting in fragmented data. Technological advancements, notably artificial intelligence (AI), offer promising avenues to enhance AMR biosurveillance. AI can improve the detection, tracking and prediction of resistant strains through advanced machine learning and deep learning algorithms, which analyze large datasets to identify resistance patterns and develop predictive models. AI's role in genomic analysis can pinpoint genetic markers and AMR determinants, aiding in precise treatment strategies. Despite the potential, SSA's implementation of AI in AMR surveillance is hindered by data scarcity, infrastructural limitations and ethical concerns. This review explores what is known about the integration and applicability of AI-enhanced biosurveillance methodologies in SSA, emphasizing the need for comprehensive data collection, interdisciplinary collaboration and the establishment of ethical frameworks. By leveraging AI, SSA can significantly enhance its AMR surveillance capabilities, ultimately improving public health outcomes.

The Role of Age in Predictive Models for Colorectal Cancer Risk: Insights From a Community-Based Real-World Study

The Role of Age in Predictive Models for Colorectal Cancer Risk: Insights From a Community-Based Real-World Study

Machine Learning Algorithms to Predict the Risk of Rupture of Intracranial Aneurysms: aSystematic Review.

Subarachnoid haemorrhage is apotentially fatal consequence of intracranial aneurysm rupture, however, it is difficult to predict if aneurysms will rupture. Prophylactic treatment of an intracranial aneurysm also involves risk, hence identifying rupture-prone aneurysms is of substantial clinical importance. This systematic review aims to evaluate the performance of machine learning algorithms for predicting intracranial aneurysm rupture risk. MEDLINE, Embase, Cochrane Library and Web of Science were searched until December 2023. Studies incorporating any machine learning algorithm to predict the risk of rupture of an intracranial aneurysm were included. Risk of bias was assessed using the Prediction Model Risk of Bias Assessment Tool (PROBAST). PROSPERO registration: CRD42023452509. Out of 10,307 records screened, 20studies met the eligibility criteria for this review incorporating atotal of 20,286 aneurysm cases. The machine learning models gave a0.66-0.90 range for performance accuracy. The models were compared to current clinical standards in six studies and gave mixed results. Most studies posed high or unclear risks of bias and concerns for applicability, limiting the inferences that can be drawn from them. There was insufficient homogenous data for ameta-analysis. Machine learning can be applied to predict the risk of rupture for intracranial aneurysms. However, the evidence does not comprehensively demonstrate superiority to existing practice, limiting its role as aclinical adjunct. Further prospective multicentre studies of recent machine learning tools are needed to prove clinical validation before they are implemented in the clinic.

Artificial intelligence, medications, pharmacogenomics, and ethics.

Artificial Intelligence (AI) and Machine Learning (ML) are revolutionizing various scientific and clinical disciplines including pharmacogenomics (PGx) by enabling the analysis of complex datasets and the development of predictive models. The integration of AI and ML with PGx has the potential to provide more precise, data-driven insights into new drug targets, drug efficacy, drug selection, and risk of adverse events. While significant effort to develop and validate these tools remain, ongoing advancements in AI technologies, coupled with improvements in data quality and depth is anticipated to drive the transition of these tools into clinical practice and delivery of individualized treatments and improved patient outcomes. The successful development and integration of AI-assisted PGx tools will require careful consideration of ethical, legal, and social issues (ELSI) in research and clinical practice. This paper explores the intersection of PGx with AI, highlighting current research and potential clinical applications, and ELSI including privacy, oversight, patient and provider knowledge and acceptance, and the impact on patient-provider relationship and new roles.

Hospital Length-of-Stay Prediction Using Machine Learning Algorithms—A Literature Review

Predicting hospital length of stay is critical for efficient hospital management, enabling proactive resource allocation, the optimization of bed availability, and optimal patient care. This paper explores the potential of machine learning algorithms to revolutionize hospital length-of-stay predictions, contributing to healthcare efficiency and patient care. The main objective is to identify the most effective machine learning algorithm for building a predictive model capable of predicting hospital length of stay. The bibliographic search of the existing literature on machine learning algorithms applied to hospital length of stay predictions highlighted the most relevant papers within this area of research. The papers were analyzed in terms of model types and metrics that contributed to the considerable impact on healthcare decision making. We also discuss the challenges and limitations of machine learning algorithms for predicting length of stay, and the importance of data quality and ethical considerations.

Tropical Interbasin Interaction as Effective Predictors of Late-Spring Precipitation Variability in the Southern Great Plains

Abstract The southern Great Plains experience fluctuating precipitation extremes that significantly impact agriculture and water management. Despite ongoing efforts to enhance forecast accuracy, the underlying causes of these climatic phenomena remain inadequately understood. This study elucidates the relative influence of the tropical Pacific and Atlantic basins on April–May–June precipitation variability in this region. Our partial ocean assimilation experiments using the Community Earth System Model unveil the prominent role of interbasin interaction, with the Pacific and Atlantic contributing approximately 70% and 30%, respectively, to these interbasin contrasts. Our statistical analyses suggest that these tropical interbasin contrasts could serve as a more reliable indicator for late-spring precipitation anomalies than El Niño–Southern Oscillation. The conclusions are reinforced by analyses of seven climate forecasting systems within the North American Multi-Model Ensemble, offering an optimistic outlook for enhancing real-time forecasting of late-spring precipitation in the southern plains. However, the current predictive skills of the interbasin contrasts across the prediction systems are hindered by the lower predictability of the tropical Atlantic Ocean, pointing to the need for future research to refine climate prediction models further. Significance Statement Agriculture and infrastructure in the southern plains face challenges from severe late-spring precipitation extremes. Traditional predictors like El Niño–Southern Oscillation (ENSO) lose effectiveness during the critical spring-to-summer transition, creating a forecasting gap. This study introduces the concept of tropical interbasin interactions, known to enhance seasonal predictability for late-spring precipitation in the southern plains. Novel climate model experiments highlight contributions from the tropical Pacific and Atlantic, offering a promising predictability that potentially surpasses the limitations of ENSO-based predictions. These outcomes hold the potential for developing operational forecasts of late-spring precipitation anomalies in the southern plains, enabling proactive risk management.

Prognostic model development using novel genetic signature associated with adenosine metabolism and immune status for patients with hepatocellular carcinoma.

The high mortality rate of hepatocellular carcinoma (HCC) is partly due to advanced diagnosis, emphasizing the need for effective predictive tools in HCC treatment. The aim of this study is to propose a novel prognostic model for HCC based on adenosine metabolizing genes and explore the potential relationship between them. Regression analysis was performed to identify differentially expressed genes associated with adenosine metabolism in HCC patients using RNA sequencing data obtained from a public database. Adenosine metabolism-related risk score (AMrisk) was derived using the least absolute shrinkage and selection operator (LASSO) Cox regression and verified using another database. Changes in adenosine metabolism in HCC were analyzed using functional enrichment analysis and multiple immune scores. The gene expression levels in patient samples were validated using quantitative reverse transcription polymerase chain reaction. Thirty adenosine metabolism-related differentially expressed genes were identified in HCC, and six genes (ADA, P2RY4, P2RY6, RPIA, SLC6A3, and VEGFA) were used to calculate the AMrisk score; the higher the risk scores, the lower the overall survival. Moreover, immune infiltration activation and immune checkpoints were considerably higher in the high-risk group. Additional in vitro experiments validated the enhanced expression of these six genes in HCC. The established predictive model demonstrated that adenosine metabolism-related genes was significantly associated with prognosis in HCC patients.

Using Explainable AI and Transfer Learning to Understand and Predict the Maintenance of Atlantic Blocking With Limited Observational Data

AbstractBlocking events are an important cause of extreme weather, especially long‐lasting blocking events that trap weather systems in place. The duration of blocking events is, however, underestimated in climate models. Explainable Artificial Intelligence are a class of data analysis methods that can help identify physical causes of prolonged blocking events and diagnose model deficiencies. We demonstrate this approach on an idealized quasigeostrophic (QG) model developed by Marshall and Molteni (1993), https://doi.org/10.1175/1520‐0469(1993)050<1792:taduop>2.0.co;2. We train a convolutional neural network (CNN), and subsequently, build a sparse predictive model for the persistence of Atlantic blocking, conditioned on an initial high‐pressure anomaly. Shapley Additive ExPlanation (SHAP) analysis reveals that high‐pressure anomalies in the American Southeast and North Atlantic, separated by a trough over Atlantic Canada, contribute significantly to prediction of sustained blocking events in the Atlantic region. This agrees with previous work that identified precursors in the same regions via wave train analysis. When we apply the same CNN to blockings in the ERA5 atmospheric reanalysis, there is insufficient data to accurately predict persistent blocks. We partially overcome this limitation by pre‐training the CNN on the plentiful data of the Marshall‐Molteni model, and then using Transfer learning (TL) to achieve better predictions than direct training. SHAP analysis before and after TL allows a comparison between the predictive features in the reanalysis and the QG model, quantifying dynamical biases in the idealized model. This work demonstrates the potential for machine learning methods to extract meaningful precursors of extreme weather events and achieve better prediction using limited observational data.

Multimodal MRI and 1H-MRS for Preoperative Stratification of High-Risk Molecular Subtype in Adult-Type Diffuse Gliomas

Isocitrate dehydrogenase (IDH) and O6-methylguanine-DNA methyltransferase (MGMT) genes are critical molecular markers in determining treatment options and predicting the prognosis of adult-type diffuse gliomas. Objectives: this study aimed to investigate whether multimodal MRI enables the differentiation of genotypes in adult-type diffuse gliomas. Methods: a total of 116 adult-type diffuse glioma patients (61 males, 51.5 (37, 62) years old) who underwent multimodal MRI before surgery were retrospectively analysed. Multimodal MRI included conventional MRI, proton magnetic resonance spectroscopy (1H-MRS), and diffusion tensor imaging (DTI). Conventional visual features, N-acetyl-aspartate (NAA)/Creatine (Cr), Choline (Cho)/Cr, Cho/NAA, fractional anisotropy (FA), mean diffusivity (MD), and diffusion histogram parameters were extracted on the whole tumour. Multimodal MRI parameters of IDH-mutant and IDH-wildtype gliomas were compared using the Mann–Whitney U test, Student’s t-test, or Pearson chi-square tests. Logistic regression was used to select the MRI parameters to predict IDH-mutant gliomas. Furthermore, multimodal MRI parameters were selected to establish models for predicting MGMT methylation in the IDH-wildtype gliomas. The performance of models was evaluated by the receiver operating characteristics curve. Results: a total of 56 patients with IDH-mutant gliomas and 60 patients with IDH-wildtype glioblastomas (GBM) (37 with methylated MGMT and 17 with unmethylated MGMT) were diagnosed by 2021 WHO classification criteria. The enhancement degree (OR = 4.298, p < 0.001), necrosis/cyst (OR = 5.381, p = 0.011), NAA/Cr (OR = 0.497, p = 0.037), FA-Skewness (OR = 0.497, p = 0.033), MD-Skewness (OR = 1.849, p = 0.035), FAmean (OR = 1.924, p = 0.049) were independent factors for the multimodal combined prediction model in predicting IDH-mutant gliomas. The combined modal based on conventional MRI, 1H-MRS, DTI parameters, and histogram performed best in predicting IDH-wildtype status (AUC = 0.890). However, only NAA/Cr (OR = 0.17, p = 0.043) and FA (OR = 0.38, p = 0.015) were associated with MGMT methylated in IDH-wildtype GBM. The combination of NAA/Cr and FA-Median is more accurate for predicting MGMT methylation levels than using these elements alone (AUC, 0.847 vs. 0.695/0.684). Conclusions: multimodal MRI based on conventional MRI, 1H-MRS, and DTI can provide compound imaging markers for stratified individual diagnosis of IDH mutant and MGMT promoter methylation in adult-type diffuse gliomas.

Conv‐ELSTM: An ensemble deep learning approach for predicting short‐term wind power

AbstractAccurate and reliable forecasting of wind power is essential for the stable integration of wind energy into the electrical grid. However, the chaotic nature of wind power presents a significant challenge in utilizing data for effective short‐term forecasting, such as 60‐min predictions. This article introduces a hybrid data‐driven framework that employs an ensemble deep learning model to provide highly precise short‐term wind power predictions. The framework leverages a data‐driven approach to identify the intrinsic components of wind power data, including high‐frequency and low‐frequency components. A convolutional layer‐based feature fusion network is then established to properly extract important information from irrelevant wind energy features. Subsequently, an ensemble of long short‐term memory (LSTM) networks is developed to forecast wind power using the fused features, thereby mitigating the disadvantage of a single prediction model. The numerical experiment is carried out based on two different real‐life datasets. The results demonstrate the effectiveness of the proposed method in forecasting short‐term wind power compared to five benchmarks.

Estimated diagnostic performance of prostate MRI performed with clinical suspicion of prostate cancer.

To assess the diagnostic performance of prostate MRI by estimating the proportion of clinically significant prostate cancer (csPCa) in patients without prostate pathology. This three-center retrospective study included prostate MRI examinations performed for clinical suspicion of csPCa (Grade group ≥ 2) between 2018 and 2022. Examinations were divided into two groups: pathological diagnosis within 1 year after the MRI (post-MRI pathology) is present and absent. Risk prediction models were developed using the extracted eleven common predictive variables from the patients with post-MRI pathology. Then, the csPCa proportion in the patients without post-MRI pathology was estimated by applying the model. The area under the receiver operating characteristic curve (AUC), sensitivity, specificity, and positive and negative predictive values (PPV/NPV) of prostate MRI in diagnosing csPCa were subsequently calculated for patients with and without post-MRI prostatepathology (estimated statistics) with a positive threshold of PI-RADS ≥ 3. Of 12,191 examinations enrolled (mean age, 65.7 years ± 8.4 [standard deviation]), PI-RADS 1-2 was most frequently assigned (55.4%) with the lowest pathological confirmation rate of 14.0-18.2%. Post-MRI prostate pathology was found in 5670 (46.5%) examinations. The estimated csPCa proportions across facilities were 12.6-15.3%, 18.4-31.4%, 45.7-69.9%, and 75.4-88.3% in PI-RADS scores of 1-2, 3, 4, and 5, respectively. The estimated (observed) performance statistics were as follows: AUC, 0.78-0.81 (0.76-0.79); sensitivity, 76.6-77.3%; specificity, 67.5-78.6%; PPV, 49.8-66.6% (52.0-67.7%); and NPV, 84.4-87.2% (82.4-86.6%). We proposed a method to estimate the probabilities harboring csPCa for patients who underwent prostateMRI examinations, which allows us to understand the PI-RADS diagnostic performance with several metrics. The reported estimated performance metrics are expected to aid in understanding the true diagnostic value of PI-RADS in the entire prostate MRI population performed with clinical suspicion of prostate cancer. Calculating performance metrics only from patients who underwent prostate biopsy may be biased due to biopsy selection criteria, especially in PI-RADS 1-2. The estimated area under the receiver operating characteristic curve of PI-RADS in the entire prostate MRI population ranged from 0.78 to 0.81 at three facilities. The estimated statistics are expected to help us understand the true PI-RADS performance and serve as a reference for future studies.

Attention 3D UNET for dose distribution prediction of high-dose-rate brachytherapy of cervical cancer: Intracavitary applicators.

Formulating a clinically acceptable plan within the time-constrained clinical setting of brachytherapy poses challenges to clinicians. Deep learning based dose prediction methods have shown favorable solutions for enhancing efficiency, but development has primarily been on external beam radiation therapy. Thus, there is a need for translation to brachytherapy. This study proposes a dose prediction model utilizing an attention-gating mechanism and a 3D UNET for cervical cancer high-dose-rate intracavitary brachytherapy treatment planning with tandem-and-ovoid/ring applicators. A multi-institutional data set consisting of 77 retrospective clinical brachytherapy plans was utilized in this study. The data were preprocessed and augmented to increase the number of plans to 252. A 3D UNET architecture with attention gates was constructed and trained for mapping the contour information to dose distribution. The trained model was evaluated on a testing data set using various metrics, including dose statistics and dose-volume indices. We also trained a baseline UNET model for a fair comparison. The attention-gated 3D UNET model exhibited competitive accuracy in predicting dose distributions similar to the ground truth. The average values of the mean absolute errors were 0.46±11.71Gy (vs. 0.47±9.16Gy for a baseline UNET) in CTVHR, 0.55±0.67Gy (vs. 0.70±1.54Gy for a baseline UNET) in bladder, 0.42±0.46Gy (vs. 0.49±1.34Gy for a baseline UNET) in rectum, and 0.31±0.65Gy (vs. 0.20±3.76Gy for a baseline UNET) in sigmoid. Our results showed that the mean individual differences in ΔD2cc for bladder, rectum, and sigmoid were 0.38±1.19 (p=0.50), 0.43±0.71 (p=0.41), and -0.47±0.79 (p=0.30) Gy, respectively. Similarly, the mean individual differences in ΔD1cc for bladder, rectum, and sigmoid were 0.09±1.21 (p=0.36), 0.20±0.95 (p=0.24), and -0.21±0.59 (p=0.30) Gy. The mean individual differences for ΔD90, ΔV100%, ΔV150%, and ΔV200% of the CTVHR were -0.45±2.42 (p=0.26) Gy, 0.55±9.42% (p=0.78), 0.82±4.21% (p=0.81), and -0.80±10.48% (p=0.36), respectively. The model requires less than 5 s to predict a full 3D dose distribution for a new patient plan. Attention-gated 3D UNET revealed a promising capability in predicting voxel-wise dose distributions compared to 3D UNET. This model could be deployed for clinical use to predict 3D dose distributions for near real-time decision-making before planning, quality assurance, and guiding future automated planning, making the current workflow more efficient.

Textbook oncological outcome of locally advanced gastric cancer patients with preoperative sarcopenia: a multicenter clinical study.

The impact of postoperative sarcopenia on the Textbook Oncological Outcome (TOO) in locally advanced gastric cancer (LAGC) remains uncertain. This study investigates the relationship between sarcopenia and TOO, explores its long-term prognostic value, and develops a prognostic model incorporating sarcopenia and TOO for survival prediction. We performed a retrospective analysis of clinical and pathological data from patients with LAGC who underwent radical surgery at two Chinese tertiary referral hospitals. Sarcopenia was defined as an SMI < 36.4 cm2/m2 in males and < 28.4 cm2/m2 in females. TOO was defined as the addition of perioperative chemotherapy to the textbook outcomes (TO). A nomogram was developed to predict postoperative overall survival (OS) and recurrence-free survival (RFS) in LAGC patients. The study included 972 patients with LAGC. The overall TOO achievement rate was 67.1%. The TOO achievement rate was significantly higher in patients non-sarcopenia compared to those with sarcopenia (68.9% vs. 61.1%, P = 0.031). Logistic regression revealed that age ≥ 65, high ASA score, and sarcopenia were independent risk factors for TOO failure. Cox regression analysis identified TOO, sarcopenia, tumor size, differentiation, vascular invasion, pT stage, and pN stage as independent predictors of OS and RFS. Nomogram models based on sarcopenia and TOO accurately predicted the 3-year and 5-year OS and RFS. Preoperative sarcopenia was an independent predictor of TOO implementation. A prognostic prediction model that integrates preoperative sarcopenia and TOO, which outperforms the current staging system, can aid clinicians in effectively assessing the prognosis of patients with LAGC.

Using Long Short-Term Memory (LSTM) recurrent neural networks to classify unprocessed EEG for seizure prediction

ObjectiveSeizure prediction could improve quality of life for patients through removing uncertainty and providing an opportunity for acute treatments. Most seizure prediction models use feature engineering to process the EEG recordings. Long-Short Term Memory (LSTM) neural networks are a recurrent neural network architecture that can display temporal dynamics and, therefore, potentially analyze EEG signals without performing feature engineering. In this study, we tested if LSTMs could classify unprocessed EEG recordings to make seizure predictions.MethodsLong-term intracranial EEG data was used from 10 patients. 10-s segments of EEG were input to LSTM models that were trained to classify the EEG signal. The final seizure prediction was generated from 5 outputs of the LSTM model over 50 s and combined with time information to account for seizure cycles.ResultsThe LSTM models could make predictions significantly better than a random predictor. When compared to other publications using the same dataset, our model performed better than several others and was comparable to the best models published to date. Furthermore, this framework could still produce predictions significantly better than chance when the experimental paradigm design was altered, without the need to reperform feature engineering.SignificanceRemoving the need to perform feature engineering is an advancement on previously published models. This framework can be applied to many different patients’ needs and a variety of acute interventions. Also, it opens the possibility of personalized seizure predictions that can be altered to meet daily needs.

Filtration Behaviour of Ostrea edulis: Diurnal Rhythmicity Influenced by Light Cycles, Body Size and Water Temperature

AbstractDespite the existing studies on oyster physiology, it is a concern that, by ignoring the diurnal rhythmicity of feeding activities, the current knowledge on oyster filtration might be unrepresentative of their real filtration capacity. In the present study, the diurnal and nocturnal fluctuations of Ostrea edulis clearance rate were investigated for the first time under laboratory conditions, also considering the combined effects of water temperature and oyster size. A positive correlation was found between the mean clearance rate (CRm) and body size for both length and wet weight, with an average of 0.5 L h−1 and 0.1 L h−1 of water filtered by large and small adult oysters respectively during the 24-h experiment. A positive significant correlation between mean clearance rate and water temperature was found only for large oysters, suggesting further studies utilising temperature ranges closer to O. edulis thermal optimum are required to clarify the influence of temperature on oyster filtration activities across different size classes. The results of this study also showed a discontinuous ingestion of food by O. edulis throughout 24-h intervals, with higher values of clearance rate observed during hours of darkness at most temperatures for both size classes. This suggests photoperiod may have a potential influence on O. edulis filtration. Providing a better understanding of O. edulis filtration behaviour, these findings have implications for ecological restoration, aquaculture, and comparative studies on oyster bioenergetics and physiology, with direct applications to the development of predictive models and the assessment of ecosystem services provided by O. edulis habitat.

Determination of Sugar Concentrations in Aqueous Solution Using Multivariate Predictions Based on 1H‐NMR Spectroscopy

AbstractRenewable chemicals from carbohydrate‐rich wastes, like furfural and 5‐hydroxymethylfurfural (HMF), are gaining prominence as alternatives to petroleum‐based resources. Assessing the suitability of biomass as feedstock for furfural and HMF production requires knowledge of its composition. This study focuses on developing and validating predictive models for individual sugar concentrations in hydrolysates using quantitative 1H NMR data. Utilizing partial least square (PLS) regression, the dataset includes 137 NMR spectra of multi‐component sugar standards (arabinose, fructose, galactose, glucose, mannose, maltose, sucrose, and xylose). The best‐performing model achieved an R2 of 0.987–0.999 and RMSECV of 0.37–1.56 mM and is based on the non‐overlapping area of the NMR spectrum. Real‐world samples were used for validation, resulting in predicted sugar concentrations with a mean standard deviation of 0.5 mM. This high accuracy and streamlined analysis process make these models practical for quantifying large sample sets, showcasing the reliability and accessibility of extracting statistical information from H‐NMR data.

A predictive and prognostic model for metastasis risk and prognostic factors in gastrointestinal signet ring cell carcinoma.

This study aimed to predict metastasis risk and identify prognostic factors of gastrointestinal signet ring cell carcinoma (SRCC) using data from the SEER database, the largest cancer dataset in North America. Data were obtained from the SEER database, covering 17 cancer registries from 2004 to 2020. Demographic and clinical data included sex, age, race, tumor location, size, pathological grade, stage, overall survival time, and treatment modalities. Statistical analyses were conducted using SPSS and R software. Propensity Score Matching (PSM) ensured comparable baseline characteristics between gastric cancer (GC) and colorectal cancer (CRC) groups. LASSO regression analysis identified predictors of metastasis, leading to the construction of predictive models using the lrm function in R. Nomograms were visualized with the "rms" package and assessed via ROC curves, calibration curves, and decision curve analysis (DCA). Cox regression analyses identified prognostic indicators for overall survival (OS), and Kaplan-Meier curves compared OS between high-risk and low-risk groups. From 2004 to 2020, 7680 SRCC patients were identified, including 4980 GC and 2700 CRC patients. CRC patients were older and had larger tumors, higher staging, and worse differentiation. Nomograms demonstrated good discriminative ability, with AUCs of 0.704 and 0.694 for GC, and 0.694 and 0.701 for CRC in training and validation cohorts, respectively. The DCA curve indicates that this predictive model has a high gain in predicting metastasis and OS. The nomograms effectively predicted metastasis risk and OS in metastatic SRCC patients, offering clinical utility in stratifying patients and guiding treatment decisions, thereby enhancing personalized treatment approaches.

Factors Inducing Cutaneous Adverse Reactions in Cancer Patients Treated with PD-1 and PD-L1 Inhibitors: A Machine-Learning Algorithm Approach

Immune checkpoint inhibitors (ICIs) show promise in cancer treatment but can lead to immune-related adverse events (irAEs), notably affecting the skin. Understanding the factors behind these skin reactions is crucial for effective management during treatment. Hence, the aim of this study was to uncover associations between patient characteristics and cutaneous adverse reactions among cancer patients undergoing ICI treatment. The study involved 209 cancer patients receiving ICIs. Statistical methods, including the chi-square test, Fisher's exact test, and multivariable logistic regression, were employed to analyze variables such as hypertension, antihistamine use, cancer metastasis, diabetes, and opioid usage. Additionally, machine learning techniques, including logistic regression, elastic net, random forest, and support vector machines (SVM), were utilized to develop predictive models anticipating skin-related adverse events. Results highlighted significant associations between specific patient attributes and the incidence of skin reactions post-ICI treatment. Notably, patients using antihistamines or with cancer metastasis exhibited higher rates of skin adverse reactions, while those with diabetes or using opioids displayed lower incidence rates. Robust performance in forecasting these adverse events was observed, particularly in the predictive models employing logistic regression and elastic net. This pioneering study contributes crucial insights into predictive modeling for ICI-induced skin reactions, emphasizing the importance of personalized treatment strategies. By identifying risk factors and utilizing tailored predictive models, healthcare providers can proactively manage adverse events, optimizing the benefits of ICIs while mitigating potential side effects.

Exergy Analysis of a District Heating System with a Predictive Model Based on Neural Network Algorithms

The article presents an exergy analysis of a district heating system (DHS) that incorporates a predictive model based on neural network algorithms. The study focuses on evaluating the thermodynamic efficiency of the DHS when these methods are applied, enabling the prediction of temperature changes and rapid adjustment of system parameters to enhance efficiency. The impact of increasing the number of connected consumers on the exergy efficiency of the DHS with the predictive model (DHPM) is examined. The use of neural network algorithms, such as LSTM, significantly enhances the system's ability to predict external temperature changes and respond promptly, resulting in optimized energy consumption and improved exergy efficiency, particularly at the beginning and end of the heating season. A comparative analysis of the exergy efficiency of traditional DHS and the system with the predictive model is included. The exergy analysis is based on modeling the system’s performance with varying numbers of consumers. The results indicate that an increase in the number of consumers leads to a rise in exergy efficiency due to better energy distribution management. The potential for significant thermal energy savings and reduced operational costs through the use of predictive methods is discussed. The study confirms that applying neural network algorithms in predictive models of DHS can substantially improve efficiency and reliability, leading to a more rational use of energy resources, cost reduction, and minimized environmental impact. The findings support the adoption of these methods in large-scale district heating systems for optimization and enhanced energy efficiency.

Prediction model and technical and tactical decision analysis of women's badminton singles based on machine learning.

In the Paris Olympic cycle, South Korean women's athlete An Se-young rose to the top of the 2023 BWF Olympic points with a win rate of 89.5%. With An Se-young as the subject, this paper aims to carry out technical and tactical analysis of women's badminton singles and formulate a prediction model based on machine learning. Firstly, An's technical and tactical statistics are analyzed and presented in a proposed "three-stage" data classification method. Secondly, we improve our "three-stage" machine learning dataset using video analysis of 10 matches (21 point games) where An Se-young faced off against four other players ranked in the top five of the World Badminton Federation (BWF) in week 44 of 2023. Finally, we establish a prediction model for the scoring and losing of points in the women's badminton singles based on the 'Decision tree', 'Random forest', 'XGBoost', 'Support vector' and 'K-proximity' algorithms, and analyze the effectiveness of this model. The results show that the improved data classification is reasonable and can be used to predict the final score of a match. When the support vector machine uses the RBF function kernel, the accuracy reaches its highest at 87.5%, and the consistency of this prediction model is strong. An's playstyle is sustained and unified; she does not seek continuous pressure, but rather exploits and maximizes her aggression following any mistake made by her opponents, immediately utilizing assault methods such as kills or dives, often resulting in the conversion of points during the subsequent 2-3 strikes.