Lundin, Sori Kim

Abstract Background Dynamically predicting patients at risk of Alzheimer's disease (AD) is crucial for timely treatment and care in precision medicine. Method We present a novel architecture, a Kolmogorov‐Arnold networks (KANs)‐based joint prediction model of longitudinal and survival data (JM‐KAN). Real‐life data of patients with Mild Cognitive Impairment (MCI) with no prior AD diagnosis from National Alzheimer's Coordinating Center (NACC) dataset was utilized for individualized predictions. We built joint models with landmark times every 1 year up to 4 years after the MCI diagnosis, and predicted survival outcomes one year after each landmark (i.e. 1, 2, 3, and 4 years). Integrated AUC (iAUC), and integrated Brier score (iBS) were measured for existing prediction models and our novel architecture. Result Out of 2711 unique patients with MCI, 821 patients progressed to develop AD. Mean time‐to‐event in years was 3.19 (standard deviation (SD): 2.29). The number of visits prior to AD diagnosis was 4.03 (SD: 2.41). With iAUC of 0.789, iBS of 0.118, JM‐KAN has both the highest iAUC and lowest iBS among joint models, which indicates superiority in discrimination and overall performance. For individualized predictions, the model successfully predicted both longitudinal risk factors and survival probability, as new information came in every year. Conclusion Using neural network‐based Artificial Intelligence (AI) model, we demonstrated that this architecture improves upon prior methods in dynamically updating predictions of AD as new longitudinal data becomes available.

Abstract Introduction COVID-19, a global health crisis, prompted rapid vaccine development. Bivalent vaccines by Pfizer-BioNTech and Moderna target both original and Omicron strains. Despite vaccination efforts, the persistence of adverse events underscores the importance of ongoing safety monitoring. Method We collected COVID-19 bivalent vaccine adverse event reports from the Vaccine Adverse Event Reporting System (VAERS) (September 15, 2022, to September 1, 2023). The Medical Dictionary for Regulatory Activities (MedDRA) categorized reported symptoms into corresponding concepts. Statistical analyses included zero-truncated Poisson regression for estimating symptom rates, logistic regression for identifying risk factors, and Spearman correlation for assessing non-parametric relationships between adverse events in different System Organ Classes (SOCs). Results Our analysis of COVID-19 bivalent vaccine adverse events identified significant correlations between SOCs. Notably, a positive correlation (ρ = 0.3) linked Infections (SOC 1) with Respiratory disorders (SOC 13), while a negative correlation (ρ=-0.3) connected Investigations (SOC 23) with Injury (SOC 24). Zero-truncated Poisson modeling estimated an average of 6.179 symptoms per individual, varying by age, gender, and vaccine. Gender and age significantly influence reported adverse events associated with COVID-19 bivalent vaccines. Females generally experienced higher odds for most SOCs compared to males, except for cardiac and product-related issues. Older adults showed a higher likelihood of symptoms across various SOCs, while children and adolescents had distinct susceptibility patterns. Additionally, Moderna and Pfizer vaccines were associated with different SOCs, with Moderna linked mainly to immune and skin disorders and Pfizer associated primarily with cardiac and metabolic issues. Conclusion In conclusion, our study aligned self-reported items with widely-accepted medical terminology and rigorously analyzed adverse events following COVID-19 bivalent vaccine using advanced statistical methods. The significant findings provide crucial insights, guiding evidence-based strategies and interventions for vaccine safety.