Personalized Risk Assessment Research Articles

Automotive Consumer Loans Risk Assessment Predictive Modeling Using Generative Adversarial Network and Stacked Autoencoder Neural Networks

In this study, we propose a predictive model for automotive consumer loans risk assessment, leveraging Generative Adversarial Networks (GANs) and Stacked Autoencoder Neural Networks (SAEs). We address issues of high dimensionality and sparsity inherent in big data environments and tackle the class imbalance problem using GANs. Feature selection is effectively carried out using SAEs. Experimental results prove the superiority of our approach over traditional neural networks and our model without GANs or SAEs. The proposed model shows significant potential for application in personal credit risk assessment within automotive finance and beyond. Future work is aimed at extending and improving our model and applying it to other domains.

Family history and genetic risk score combined to guide cancer risk stratification: A prospective cohort study.

Family history (FH) of cancer and polygenic risk scores (PRS) are pivotal for cancer risk assessment, yet their combined impact remains unclear. Participants in the UK Biobank (UKB) were recruited between 2006 and 2010, with complete follow-up data updated until February 2020 for Scotland and January 2021 for England and Wales. Using UKB data (N = 442,399), we constructed PRS and incidence-weighted overall cancer PRS (CPRS). FH was assessed through self-reported standardized questions. Among 202,801 men (34.6% with FH) and 239,598 women (42.0% with FH), Cox regression was used to examine the associations between FH, PRS, and cancer risk. We found a significant dose-response relationship between FH of cancer and corresponding cancer risk (Ptrend < .05), with over 10 significant pairs of cross-cancer effects of FH. FH and PRS are positively correlated and independent. Joint effects of FH of cancer (multiple cancers) and PRS (CPRS) on corresponding cancer risk were observed: for instance, compared with participants with no FH of cancer and low PRS, men with FH of cancer and high PRS had the highest risk of colorectal cancer (hazard ratio [HR]: 3.69, 95% confidence interval [CI]: 3.01-4.52). Additive interactions were observed in prostate and overall cancer risk for men and breast cancer for women, with the most significant result being a relative excess risk of interaction (RERI) of 2.98, accounting for ~34% of the prostate cancer risk. In conclusion, FH and PRS collectively contribute to cancer risk, supporting their combined application in personalized risk assessment and early intervention strategies.

Stevens-Johnson syndrome and toxic epidermal necrolysis-like eruptions in patients treated with immune checkpoint inhibitors: a systematic review.

Immune checkpoint inhibitors (ICIs) have transformed cancer treatment by targeting immune checkpoints such as PD-1, PDL-1, and CTLA-4, but concerns about severe immune-related adverse events persist. The scarcity of literature on dermatologic implications, especially severe reactions such as Stevens-Johnson syndrome (SJS) and toxic epidermal necrolysis (TEN), highlights the urgent need for investigation. Our systematic review aims to address the gap in relevant literature by extensively examining the epidemiologic risk factors and management of SJS/TEN-like illnesses in ICI-treated patients to provide insights for risk assessment and clinical care. We identified 158 case reports that detailed the incidence of SJS/TEN in patients being treated with ICIs, examining demographic patterns, type of malignancy, clinical characteristics, and treatments linked to onset. We assessed mortality rates, risk elements, and the effectiveness of interventions to help guide clinical care. Analysis of 158 case reports revealed that SJS/TEN in ICI users is typically seen on average at the age of 63 and is more common in males. PD1 inhibitors such as nivolumab and pembrolizumab are often associated with various mucocutaneous patterns and significant risks with ICI use, especially TEN, which is linked to high morbidity and mortality rates. Our study notes limitations due to the inclusion of case reports or case series, such as potential publication and reporting biases, leading to skewed findings. Additionally, because of the heterogeneous reporting standards, the retrospective nature limits phenotypic precision, control for confounding variables, and data completeness. Our study provides valuable insights into the epidemiology, clinical features, management strategies, and outcomes of ICI-induced SJS/TEN, underscoring the importance of vigilant monitoring and personalized risk assessment in oncology practice. Continued research efforts are essential to optimize patient outcomes and enhance the safety profile of ICIs in cancer therapy.

AI-Driven Innovations in Earthquake Risk Mitigation: A Future-Focused Perspective

This study explores the transformative potential of artificial intelligence (AI) in revolutionizing earthquake risk mitigation across six key areas. Unlike traditional approaches, this paper examines how AI-driven innovations can uniquely enhance early warning systems, enabling real-time structural health monitoring, and providing dynamic, multi-hazard risk assessments that seamlessly integrate seismic data with other natural hazards such as tsunamis and landslides. It introduces groundbreaking applications of AI in earthquake-resilient design, where generative design algorithms and predictive analytics create structures that optimally balance safety, cost, and sustainability. The study also presents a novel discussion on the ethical implications of AI in this domain, stressing the critical need for transparency, accountability, and bias mitigation. Looking forward, the manuscript envisions the development of advanced AI platforms capable of delivering real-time, personalized risk assessments, immersive public training programs, and collaborative design tools that adapt to evolving seismic data. These innovations promise not only to significantly enhance current earthquake preparedness but also to pave the way toward a future where the societal impact of earthquakes is drastically reduced. This work underscores the potential of AI’s role in shaping a safer, more resilient future, emphasizing the importance of continued innovation, ethical governance, and collaborative efforts.

ERG-AI: enhancing occupational ergonomics with uncertainty-aware ML and LLM feedback

Workers, especially those involved in jobs requiring extended standing or repetitive movements, often face significant health challenges due to Musculoskeletal Disorders (MSDs). To mitigate MSD risks, enhancing workplace ergonomics is vital, which includes forecasting long-term employee postures, educating workers about related occupational health risks, and offering relevant recommendations. However, research gaps remain, such as the lack of a sustainable AI/ML pipeline that combines sensor-based, uncertainty-aware posture prediction with large language models for natural language communication of occupational health risks and recommendations. We introduce ERG-AI, a machine learning pipeline designed to predict extended worker postures using data from multiple wearable sensors. Alongside providing posture prediction and uncertainty estimates, ERG-AI also provides personalized health risk assessments and recommendations by generating prompts based on its performance and prompting Large Language Model (LLM) APIs, like GPT-4, to obtain user-friendly output. We used the Digital Worker Goldicare dataset to assess ERG-AI, which includes data from 114 home care workers who wore five tri-axial accelerometers in various bodily positions for a cumulative 2913 hours. The evaluation focused on the quality of posture prediction under uncertainty, energy consumption and carbon footprint of ERG-AI and the effectiveness of personalized recommendations rendered in easy-to-understand language.

Acceptability of de-intensified screening for women at low risk of breast cancer: a randomised online experimental survey

BackgroundRisk-stratified approaches to breast screening show promise for increasing benefits and reducing harms. But the successful implementation of such an approach will rely on public acceptability. To date, research suggests that while increased screening for women at high risk will be acceptable, any de-intensification of screening for low-risk groups may be met with less enthusiasm. We report findings from a population-based survey of women in England, approaching the age of eligibility for breast screening, to compare the acceptability of current age-based screening with two hypothetical risk-adapted approaches for women at low risk of breast cancer.MethodsAn online survey of 1,579 women aged 40–49 with no personal experience of breast cancer or mammography. Participants were recruited via a market research panel, using target quotas for educational attainment and ethnic group, and were randomised to view information about (1) standard NHS age-based screening; (2) a later screening start age for low-risk women; or (3) a longer screening interval for low-risk women. Primary outcomes were cognitive, emotional, and global acceptability. ANOVAs and multiple regression were used to compare acceptability between groups and explore demographic and psychosocial factors associated with acceptability.ResultsAll three screening approaches were judged to be acceptable on the single-item measure of global acceptability (mean score > 3 on a 5-point scale). Scores for all three measures of acceptability were significantly lower for the risk-adapted scenarios than for age-based screening. There were no differences between the two risk-adapted scenarios. In multivariable analysis, higher breast cancer knowledge was positively associated with cognitive and emotional acceptability of screening approach. Willingness to undergo personal risk assessment was not associated with experimental group.ConclusionWe found no difference in the acceptability of later start age vs. longer screening intervals for women at low risk of breast cancer in a large sample of women who were screening naïve. Although acceptability of both risk-adapted scenarios was lower than for standard age-based screening, overall acceptability was reasonable. The positive associations between knowledge and both cognitive and emotional acceptability suggests clear and reassuring communication about the rationale for de-intensified screening may enhance acceptability.

MtPCDI: a machine learning-based prognostic model for prostate cancer recurrence.

This research seeks to formulate a prognostic model for forecasting prostate cancer recurrence by examining the interaction between mitochondrial function and programmed cell death (PCD). The research involved analyzing four gene expression datasets from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) using univariate Cox regression. These analyses identified genes linked with mitochondrial function and PCD that correlate with recurrence prognosis. Various machine learning algorithms were then employed to construct an optimal predictive model. A key outcome was the creation of a mitochondrial-related programmed cell death index (mtPCDI), which effectively predicts the prognosis of prostate cancer patients. It was observed that individuals with lower mtPCDI exhibited higher immune activity, correlating with better recurrence outcomes. The study demonstrates that mtPCDI can be used for personalized risk assessment and therapeutic decision-making, highlighting its clinical significance and providing insights into the biological processes affecting prostate cancer recurrence.

Assessing the quality of AI information from ChatGPT regarding oral surgery, preventive dentistry, and oral cancer: An exploration study

AimEvaluation of the quality of dental information produced by the ChatGPT artificial intelligence language model within the context of oral surgery, preventive dentistry, and oral cancer. MethodologyThis study adopted quantitative methods approach. The experts prepared 50 questions (including dimensions of, risk factors, preventive measures, diagnostic methods, and treatment options) that would be presented to ChatGPT, and its responses were rated for their accuracy, completeness, relevance, clarity or comprehensibility, and possible risks using a standardized rubric. To carry out the assessment of the responses by ChatGPT, a standardized scoring rubric was used. Evaluation process included feedback concerning the strengths, weaknesses, and potential areas of improvement in the responses provided by ChatGPT. ResultsWhile achieving the highest score for preventive dentistry at 4.3/5 and being able to communicate the complex information coherently, the tool showed lower accuracy for oral surgery and oral cancer, scoring 3.9/5 and 3.6/5, respectively, with several gaps for post-operative instructions, personalized risk assessments, and specialized diagnostic methods. Potential risks, such as a lack of individualized advice, were shown in 53% of the oral cancer and in 40% of the oral surgery. While showing promise in some domains, ChatGPT had important limitations in specialized areas that require nuanced expertise. ConclusionThe findings point to the need for professional supervision while using AI-generated information and ongoing evaluation as capabilities evolve, for the assurance of responsible implementation in the best interest of patient care.

Research on the Application of Alternative Data in Credit Risk Management

With the development of financial technology, traditional credit assessment models have gradually shown their limitations. Especially in assessing borrowers with no credit history or weak credit records. The rise of alternative data provides a new dimension for credit risk prediction, including but not limited to social media behavior, online transaction records, geographic location data, etc. This paper explores the current application status, challenges, and future development trends of alternative data in personal credit risk assessment, and explores the application and effects of various forms of alternative data through different classifications. This paper refers to the relevant literature on alternative data and credit risk management and finds that the application of alternative data can not only supplement part of the information reference to enhance the risk management model but also further provide certain credit credentials for groups that cannot obtain credit services with traditional credit data. It has potential contributions to improving credit risk management and promoting the development of inclusive finance.

365. METHYLATION BIOMARKERS TO STRATIFY BARRETT’S ESOPHAGUS PATIENTS PROGRESSING AND NON-PROGRESSING TO ESOPHAGEAL ADENOCARCINOMA

Abstract Background Over the last forty years, North America and Europe have witnessed a notable surge in cases of esophageal adenocarcinoma (EAC). This malignancy typically arises from Barrett esophagus (BE), a condition characterized by the replacement of normal squamous epithelium with specialized metaplastic cells. Our study aims to identify specific methylation markers capable of distinguishing between BE patients that will and will not progress to EAC. Methods We conducted differential expression analysis on RNASeq public dataset GSE210647 to distinguish BE progressors from non-progressors. Reads were aligned to hg38 genome using STAR aligner, normalized and analyzed with DESeq2. Significant genes were selected by |log2 fold change| ≥ 0.2 and padj &amp;lt; 0.01. GO, pathway enrichment, and transcription factor (TF) network analyses were performed to uncover biological disparities. Integration with previous methylation data (N.R. Dehkordi et al., 292. Specific methylation markers associated with changes in gene expression distinguish EAC and sub-types of BE patients, ISDE 2023) allowed the identification of overlapping biomarkers, aiding in better discrimination between the two groups. Results Differential expression analysis of GSE210647 identified 2261 significative DEGs between BE progressors and non-progressors. Pathway enrichment analysis in BE progressors revealed significant associations with cellular morphogenesis and reprogramming, alongside the induction of the bile acid pathway, known for its implications in invasiveness and cancer stem cell expansion. TF network analysis identified key regulators such as TBX20, FOXS1, and ZNF454, shedding light on their roles in esophageal carcinogenesis. Among the 505 CpG markers acquired in prior analyses, 127 genomic regions were linked with overexpressed genes either in progressors or in non-progressors BE patients. Leveraging this marker panel, BE samples were stratified into discrete subtypes, aligning either with normal esophageal tissue or EAC, and suggesting a potential stratification by disease progression. Conclusions The integration of expression and methylation analyses allows the identification of key genomic and epigenomic loci able to distinguish BE patients progressing or not-progressing towards EAC. The identified biomarkers are pivotal for personalized risk assessment and for the design of targeted interventions, holding promises for enhanced surveillance protocols, refined early EAC detection approaches, and improved therapeutic strategies aimed at mitigating the risk of EAC development in BE population.

A real-world pharmacovigilance study of cardiac adverse events induced by sugammadex in the FDA adverse event reporting system

ABSTRACT Background Sugammadex is a novel agent that reverses neuromuscular blockade during general anesthesia. Recent case reports have raised concerns regarding potential cardiac adverse events (CAEs). However, no large-scale real-world studies have yet evaluated the potential link between sugammadex and CAEs. Research design and methods Data from the FDA Adverse Event Reporting System were obtained. The association between sugammadex and CAE was evaluated using reporting odds ratio, proportional reporting ratio, Bayesian confidence propagation neural network, and multi-item gamma Poisson shrinker methods. Serious outcomes resulting from sugammadex-related CAEs were assessed, and complications associated with CAEs were evaluated. Results Nineteen CAEs were identified and classified into two categories: cardiac arrhythmias and coronary artery disorders. The most frequent CAEs were bradycardia (n = 202), cardiac arrest (n = 119), tachycardia (n = 30), and Kounis syndrome (n = 22). Subgroup analysis based on age, sex, and weight revealed parallel findings. The CAEs most likely to result in serious consequences were pulseless electrical activity and cardiac arrest. The most common concurrent adverse effects with CAEs were hypotension (n = 51), anaphylactic reactions (n = 46), and anaphylactic shock (n = 23). Conclusion This study suggests a potential link between sugammadex and CAEs, highlighting the need for careful monitoring and personalized risk assessment, especially in patients with cardiovascular risk factors.

Association of ACE and AGTR1 variants with retinopathy of prematurity: a case-control study and meta-analysis.

Retinopathy of prematurity (ROP) is a major cause of childhood blindness worldwide, linked to gene variants in the renin-angiotensin-aldosterone system, including angiotensin-converting enzyme (ACE) and angiotensin II receptor type 1 (AGTR1). This study aims to evaluate the association between ACE insertion/deletion (I/D) and AGTR1 rs5186A > C variants with the occurrence and progression of ROP in a Polish cohort. A total of 377 premature infants were enrolled in the study. The ACE variant was evaluated using PCR, and AGTR1 was assessed using TaqMan probes. Clinical characteristics, including risk factors and comorbidities, were documented. A meta-analysis of the effects of the studied variants on ROP was also conducted. The AGTR1 rs5186C allele was significantly associated with both the progression of ROP and treatment outcomes. Homozygotes exhibited a 2.47-fold increased risk of developing proliferative ROP and a 4.82-fold increased risk of treatment failure. The impact of this allele increased at low birth weight. A meta-analysis, including 191 cases and 1661 controls, indicated an overall risk of 1.7 (95%CI 1.02-2.84) for the recessive effect of the rs5186C allele. The ACE variant did not show a significant association with ROP in our population; however, a meta-analysis of 996 cases and 2787 controls suggested a recessive effect of the insertion allele (an odds ratio of 1.21 (95%CI 1.00-1.60)). These results indicate that gain-of-function AGTR1 variants may play a crucial role in the development of ROP, potentially by promoting angiogenesis and pro-inflammatory effects. Screening for these variants could facilitate the development of personalized risk assessment and treatment strategies for ROP.

Artificial intelligence enabled interpretation of ECG images to predict hematopoietic cell transplantation toxicity

AbstractArtificial intelligence (AI)–enabled interpretation of electrocardiogram (ECG) images (AI-ECGs) can identify patterns predictive of future adverse cardiac events. We hypothesized that such an approach would provide prognostic information for the risk of cardiac complications and mortality in patients undergoing hematopoietic cell transplantation (HCT). We retrospectively subjected ECGs obtained before HCT to an externally trained, deep-learning model designed to predict the risk of atrial fibrillation (AF). Included were 1377 patients (849 autologous [auto] HCT and 528 allogeneic [allo] HCT recipients). The median follow-up was 2.9 years. The 3-year cumulative incidence of AF was 9% (95% confidence interval [CI], 7-12) in patients who underwent auto HCT and 13% (10%-16%) in patients who underwent allo HCT. In the entire cohort, pre-HCT AI-ECG estimate of AF risk correlated highly with the development of clinical AF (hazard ratio [HR], 7.37; 95% CI, 3.53-15.4; P < .001), inferior survival (HR, 2.4; 95% CI, 1.3-4.5; P = .004), and greater risk of nonrelapse mortality (NRM; HR, 95% CI, 3.36; 1.39-8.13; P = .007), without increased risk of relapse. Association with mortality was only noted in allo HCT recipients, where the risk of NRM was greater. The use of cyclophosphamide after transplantation resulted in greater 90-day incidence of AF (13% vs 5%; P = .01) compared to calcineurin inhibitor–based graft-versus-host disease prophylaxis, corresponding to temporal changes in AI-ECG AF prediction after HCT. In summary, AI-ECG can inform risk of posttransplant cardiac outcomes and survival in HCT patients and represents a novel strategy for personalized risk assessment.

To Leave or Stay During the COVID-19 Crisis: Exploring the Impacts of CSR Motives on Employee Distrust and Turnover Intentions

ABSTRACT In response to the increasing prominence of corporate social responsibility (CSR) during the COVID-19 pandemic, this study investigates the interplay between employees’ perceived CSR motives, distrust in their employer, perceived crisis controllability, and turnover intention. Using a structural equation model and data from a survey of 740 U.S. employees, our findings indicate that intrinsic CSR heightens employees’ distrust in their organization, whereas extrinsic CSR reduces it. This distrust intensifies employees’ turnover intentions. Employees’ perceived crisis controllability, a personal risk assessment of the situation, mediates the relationship between employee distrust and turnover intention. The theoretical and practical implications are also discussed.

Exploring the potential of IL-10 for risk assessment and early intervention in pediatric ALL

Acute lymphoblastic leukemia (ALL), a leading cause of childhood cancer, targets immune system B and T cells. While understanding its causes is crucial, predicting susceptibility holds immense power for early diagnosis and intervention. This study explored the potential of interleukin 10 (IL-10), a key immune regulator, as a predictive tool in Egyptian children. Investigating 100 ALL patients and 100 healthy controls, we analyzed the IL10 gene polymorphism (-1082 A/G) and serum levels. Strikingly, both the G allele and higher serum IL-10 levels were significantly associated with increased ALL risk (p < 0.05, OR > 1). Moreover, IL-10 emerged as a remarkably accurate predictor, boasting an AUC of 0.995, with a sensitivity of 97% and specificity of 96%. These findings unveil the potential of IL-10 as a powerful predictive tool for pediatric ALL in the studied Egyptian population. Identifying individuals with the GG/AG haplotype and elevated IL-10 levels could enable early intervention and potentially improve outcomes. While further validation in larger and more diverse populations is needed, this study paves the way for personalized risk assessment and potentially revolutionizes how we combat this childhood killer.

Carbapenem-resistant Enterobacterales sepsis following endoscopic retrograde cholangiopancreatography: risk factors for 30-day all-cause mortality and the development of a nomogram based on a retrospective cohort

BackgroundEndoscopic retrograde cholangiopancreatography (ERCP) has become a routine endoscopic procedure that is essential for diagnosing and managing various conditions, including gallstone extraction and the treatment of bile duct and pancreatic tumors. Despite its efficacy, post-ERCP infections – particularly those caused by carbapenem-resistant Enterobacterales (CRE) – present significant risks. These risks highlight the need for accurate predictive models to enhance postprocedural care, reduce the mortality risk associated with post-ERCP CRE sepsis, and improve patient outcomes in the context of increasing antibiotic resistance.ObjectiveThis study aimed to examine the risk factors for 30-day mortality in patients with CRE sepsis following ERCP and to develop a nomogram for accurately predicting 30-day mortality risk.MethodsData from 195 patients who experienced post-ERCP CRE sepsis between January 2010 and December 2022 were analyzed. Variable selection was optimized via the least absolute shrinkage and selection operator (LASSO) regression model. Multivariate logistic regression analysis was then employed to develop a predictive model, which was evaluated in terms of discrimination, calibration, and clinical utility. Internal validation was achieved through bootstrapping.ResultsThe nomogram included the following predictors: age > 80 years (hazard ratio [HR] 2.61), intensive care unit (ICU) admission within 90 days prior to ERCP (HR 2.64), hypoproteinemia (HR 4.55), quick Pitt bacteremia score ≥ 2 (HR 2.61), post-ERCP pancreatitis (HR 2.52), inappropriate empirical therapy (HR 3.48), delayed definitive therapy (HR 2.64), and short treatment duration (< 10 days) (HR 5.03). The model demonstrated strong discrimination and calibration.ConclusionsThis study identified significant risk factors associated with 30-day mortality in patients with post-ERCP CRE sepsis and developed a nomogram to accurately predict this risk. This tool enables healthcare practitioners to provide personalized risk assessments and promptly administer appropriate therapies against CRE, thereby reducing mortality rates.

Screening Strategies and Methodologies.

While glaucoma is a leading cause of irreversible vision loss, it presents technical challenges in the design and implementation of screening. New technologies such as PRS and AI offer potential improvements in our ability to identify people at high risk of sight loss from glaucoma and may improve the viability of screening for this important disease. To review the current evidence and concepts around screening for glaucoma. A group of glaucoma-focused clinician scientists drew on knowledge and experience around glaucoma, its etiology, and the options for screening. Glaucoma is a chronic progressive optic neuropathy affecting around 76 million individuals worldwide and is the leading cause of irreversible blindness globally. Early stages of the disease are asymptomatic meaning a substantial proportion of cases remain undiagnosed. Early detection and timely intervention reduce the risk of glaucoma-related visual morbidity. However, imperfect tests and a relatively low prevalence currently limit the viability of population-based screening approaches. The diagnostic yield of opportunistic screening strategies, relying on the identification of disease during unrelated health care encounters, such as cataract clinics and diabetic retinopathy screening programs, focusing on older people and/or those with a family history, are hindered by a large number of false-positive and false-negative results. Polygenic risk scores (PRS) offer personalized risk assessment for adult-onset glaucoma. In addition, artificial intelligence (AI) algorithms have shown impressive performance, comparable to expert humans, in discriminating between potentially glaucomatous and non-glaucomatous eyes. These emerging technologies may offer a meaningful improvement in diagnostic yield in glaucoma screening. While glaucoma is a leading cause of irreversible vision loss, it presents technical challenges in the design and implementation of screening. New technologies such as PRS and AI offer potential improvements in our ability to identify people at high risk of sight loss from glaucoma and may improve the viability of screening for this important disease.

Heart Disease Prediction Using Machine Learning

Heart disease remains a leading cause of mortality worldwide, necessitating improved methods for early detection and risk assessment. This paper reviews and analyzes the application of machine learning techniques in heart disease prediction, focusing on five primary algorithms: Naïve Bayes, k-Nearest Neighbor (KNN), Decision Tree, Artificial Neural Network (ANN), and Random Forest. By examining existing studies and datasets, we evaluate the effectiveness of these algorithms in predicting heart disease risk. Our analysis demonstrates that machine learning models can significantly enhance the accuracy and efficiency of heart disease prediction compared to traditional diagnostic methods. The Random Forest algorithm exhibited the highest overall performance, with studies reporting accuracy rates up to 95% in identifying potential heart disease cases. This review highlights the potential of machine learning in revolutionizing cardiovascular healthcare by enabling more personalized risk assessments and facilitating early intervention strategies. The integration of these advanced predictive models into clinical practice could substantially improve patient outcomes and reduce the global burden of heart disease. Keywords: Cardiovascular Risk Prediction, Machine Learning Algorithms, Electronic Health Records, Random Forest, Artificial Neural Networks, Feature Importance, Clinical Decision Support, Personalized Medicine, Predictive Analytics in Healthcare, Early Disease Detection.

Predicting progression-free survival in patients with epithelial ovarian cancer using an interpretable random forest model

Prognostic models play a crucial role in providing personalised risk assessment, guiding treatment decisions, and facilitating the counselling of patients with cancer. However, previous imaging-based artificial intelligence models of epithelial ovarian cancer lacked interpretability. In this study, we aimed to develop an interpretable machine-learning model to predict progression-free survival in patients with epithelial ovarian cancer using clinical variables and radiomics features. A total of 102 patients with epithelial ovarian cancer who underwent contrast-enhanced computed tomography scans were enrolled in this retrospective study. Pre-surgery clinical data, including age, performance status, body mass index, tumour stage, venous blood cancer antigen-125 (CA125) level, white blood cell count, neutrophil count, red blood cell count, haemoglobin level, and platelet count, were obtained from medical records. The volume of interest for each tumour was manually delineated slice-by-slice along the boundary. A total of 2074 radiomic features were extracted from the pre- and post-contrast computed tomography images. Optimal radiomic features were selected using the Least Absolute Shrinkage and Selection Operator logistic regression. Multivariate Cox analysis was performed to identify independent predictors of three-year progression-free survival. The random forest algorithm developed radiomic and combined models using four-fold cross-validation. Finally, the Shapley additive explanation algorithm was applied to interpret the predictions of the combined model. Multivariate Cox analysis identified CA-125 levels (P = 0.015), tumour stage (P = 0.019), and Radscore (P < 0.001) as independent predictors of progression-free survival. The combined model based on these factors achieved an area under the curve of 0.812 (95 % confidence interval: 0.802–0.822) in the training cohort and 0.772 (95 % confidence interval: 0.727–0.817) in the validation cohort. The most impactful features on the model output were Radscore, followed by tumour stage and CA-125. In conclusion, the Shapley additive explanation-based interpretation of the prognostic model enables clinicians to understand the reasoning behind predictions better.

Applications of Artificial Intelligence in Psychiatric Nursing: A Scope Review.

Rapid advances in artificial intelligence (AI) have reshaped healthcare, including psychiatric nursing, to address the limitations of traditional approaches and meet escalating mental health challenges. A scoping review analyzed 48 articles examining the application of AI in psychiatric nursing across different technologies and topics, noting trends in publications and countries involved. The articles covered different aspects of mental health using AI technologies such as machine learning and robotics, and primarily explored AI applications in mental health, specifically dementia, autism and schizophrenia. These studies highlighted the role of AI in personalized care plans, symptom monitoring and risk assessment. AI is promising, but faces challenges such as data bias and ethical concerns. Future research needs to focus on long-term studies, diverse populations, patient interaction and personalized treatments for practical integration into psychiatric nursing.