AI Medical Compendium

Sparse Memory Finetuning as a Low-Forgetting Alternative to LoRA and Full Finetuning

arXiv May 4, 2026

Adapting a pretrained language model to a new task often hurts the general capabilities it already had, a problem known as catastrophic forgetting. Sparse Memory Finetuning (SMF) tries to avoid this by adding key-value memory layers to the model and,... read more

Computation and Language cs.LG

Decoding the neural mechanisms of salty peptide perception via electroencephalography and machine learning.

Food chemistry May 3, 2026

Excessive sodium intake poses major public health risks, driving the search for salt substitutes that preserve desirable flavor. Salty peptides have emerged as promising low-sodium alternatives, yet their neural perception, distinct from the ion-depe... read more

Temperature robustness of sulfur and hydrogen autotrophic reduction systems for antimonate detoxification: Experiment and machine learning-driven evaluation.

Journal of hazardous materials May 3, 2026

This study, for the first time, revealed the temperature robustness of elemental sulfur (S0) (SAR) and hydrogen (H2) (HAR), and a combination of S0 and H2 autotrophic reduction (SHAR) systems for Sb(V) removal from water. SHAR system demonstrated sup... read more

Support vector Regression-Optimized Ultrasonic-Assisted alkaline extraction of Chroogomphus rutilus Polysaccharides: Enhanced yield and preservation of helical conformations.

Ultrasonics sonochemistry May 3, 2026

Polysaccharides from Chroogomphus rutilus (CRPs) exhibit promising bioactivities; however, conventional alkaline extraction (CAE) often results in low yields and structural disruption. In this study, a support vector regression (SVR)-optimized ultras... read more

How improper dataset split hinders model generalizability: a systematic comparison in Human activity recognition and exercise evaluation tasks.

International journal of medical informatics May 3, 2026

BACKGROUND: Human Activity Recognition (HAR) and exercise assessment models are increasingly used in healthcare to support clinical evaluation, rehabilitation, and remote monitoring. However, their real-world applicability critically depends on the a... read more

Artificial intelligence predictions of knee kinematics, kinetics, and internal biomechanics during walking in people with knee osteoarthritis: A systematic review and meta-analysis.

Clinical biomechanics (Bristol, Avon) May 3, 2026

This review synthesizes current literature on the use of artificial intelligence (AI) to predict knee biomechanics during walking in people with knee osteoarthritis (OA). Four databases were searched from inception to 22/01/2025. Risk of bias was ass... read more

Integrating human and nonhuman intelligence to reduce tobacco's public-health burden.

Current opinion in psychiatry May 3, 2026

PURPOSE OF REVIEW: Tobacco use remains the leading preventable cause of death worldwide, while the rise of electronic nicotine products has sparked a new wave of initiation. The urgent need for scalable, multilevel tobacco-control interventions conve... read more

Imaging Biomarkers for Early Differentiation of Candida and Aspergillus Endogenous Fungal Endophthalmitis: Multicenter Optical Coherence Tomography-Based Analysis.

American journal of ophthalmology May 3, 2026

PURPOSE: Endogenous fungal endophthalmitis (EFE) is a rare, sight-threatening intraocular infection with heterogeneous clinical presentations. Candida and Aspergillus species are the most common etiologic agents and may manifest with retinochoroiditi... read more

Development of a four-gene host signature for paucibacillary TB among symptomatic individuals with sputum Xpert MTB/RIF Ultra very low and trace results.

International journal of infectious diseases : IJID : official publication of the International Society for Infectious Diseases May 3, 2026

OBJECTIVES: The Xpert MTB/RIF Ultra (Ultra) assay is widely used to diagnose pulmonary tuberculosis (TB), but "very low" or "trace" results may reflect either paucibacillary TB or TB-negative disease, complicating clinical decision-making. We evaluat... read more

Explainable artificial intelligence in endodontics: Concepts, clinical applications, and future directions.

Journal of dentistry May 3, 2026

OBJECTIVES: An inherent `black box` nature of artificial intelligence (AI) remains a significant barrier to clinical adoption. Explainable artificial intelligence (XAI) has emerged as an approach to enhance transparency and interpretability. This nar... read more

Artificial Intelligence Medical Compendium