Effectiveness of Transformer-Based Large Language Models in Identifying Adverse Drug Reaction Relations from Unstructured Discharge Summaries in Singapore.

Introduction: Transformer-based large language models (LLMs) have transformed the field of natural language processing and led to significant advancements in various text processing tasks. However, the applicability of these LLMs in identifying related drug-adverse event (AE) pairs within clinical context may be limited by the prevalent use of non-standard sentence structures and grammar.

Method: Nine transformer-based LLMs pre-trained on biomedical domain corpora are fine-tuned on annotated data (n = 5088) to classify drug-AE pairs in unstructured discharge summaries as causally related or unrelated.

Guideline-directed Medical Therapy in Nonagenarians and Centenarians (≥ 90 Years Old) After First-onset Myocardial Infarction---a National Registry Study.

Background: Guideline-directed medical therapies (GDMTs), such as beta-blockers, antiplatelet drugs, lipid-lowering drugs, and renin-angiotensin system agents, have been associated with reduced risk of mortality after acute myocardial infarction (AMI). However, this survival benefit conferred by GDMTs in nonagenarians and centenarians (≥ 90 years old) is not well-defined.

Methods: We investigated restricted mean survival times of patients ≥ 90 years of age with first-onset AMI treated with GDMTs from 2007 to 2020 in the Singapore Myocardial Infarction Registry.

Dynamic simulation of breast behaviour during different activities based on finite element modelling of multiple components of breast.

This study presents an advanced dynamic finite element (FE) model of multiple components of the breast to examine the biomechanical impact of different types of physical activities and activity intensity on the breast tissues. Using 4D scanning and motion capture technologies, dynamic data are collected during different activities. The accuracy of the FE model is verified based on relative mean absolute error (RMAE), and optimal material parameters are identified by using a validated stepwise grid search method.

Impact of a Patient Portal-Based Telehealth Outreach Program on Recall of Patients with Diabetic Retinopathy.

To evaluate the effectiveness of a patient portal telehealth outreach program to return patients with diabetic retinopathy (DR) lost to follow-up (LTFU) for eye care. Patients with DR receiving intravitreal injection (IVI) therapy who were >90 days beyond recommended return were deemed LTFU. Outreach messages were sent via a patient portal, when available, or through the U.

Right Ventricular Restrictive Physiology Is Associated With Right Ventricular Direct Flow From 4D Flow CMR.

Background: Right ventricular restrictive physiology (RVRP) is a common occurrence in repaired tetralogy of Fallot (rTOF). The relationship of RVRP with biventricular blood flow components and kinetic energy (KE) from 4-dimensional (4D) flow cardiovascular magnetic resonance (CMR) is unclear.

Objectives: The purpose of this study was to investigate the association of 4D flow CMR parameters with RVRP in rTOF patients.