Publications by authors named "M Ozalp"
Article Synopsis
- The study investigates Elabela as a potential laboratory marker for predicting placenta previa and placenta accreta, two serious obstetric conditions that currently lack biochemical diagnostic tools.
- Conducted between 2020 and 2022 at two tertiary centers, the research analyzed Elabela levels in groups with plancental anomalies, finding significant differences in Elabela levels among control, previa, and accreta groups.
- Results indicated a promising cut-off value for Elabela in diagnosing both conditions, with sensitivity and specificity showing potential for Elabela to be used as a biochemical marker, encouraging further research in this area.
Objective: The aim was to find a cost-effective, more practical method to be used in the early gestational weeks as an alternative to the oral glucose tolerance test (OGTT) for predicting gestational diabetes mellitus (GDM). The method selected was adipose tissue measurements made in the first trimester.
Material And Methods: The study was designed as a prospective, cohort study.
Article Synopsis
- Recent advancements in machine learning, particularly with architectures like transformers and few-shot learning models, have improved text generation and image analysis tasks.
- The 'no-free lunch' theorem indicates that there's no one-size-fits-all model; different algorithms excel in varying circumstances based on dataset characteristics.
- The study identifies a "goldilocks zone" for model performance: few-shot learning models excel with very small datasets, transformers perform well with small-to-medium and diverse datasets, while classical models are best with larger, sufficiently sized datasets.
Article Synopsis
- Cancer diagnosis is on the rise globally, and access to cancer drugs is a key health policy issue in Türkiye; researchers used drug sales data to evaluate this access.
- The study analyzed sales data from IQVIA, adjusting for factors like population growth, cancer rates, and Euro exchange rates, revealing varying drug consumption trends based on cancer type.
- Findings indicated that pricing strategies, influenced by exchange rates and population factors, are barriers to accessing oncology medicines, prompting a need for policy review to improve access.
Butyrylcholinesterase (BChE) is a target of interest in late-stage Alzheimer's Disease (AD) where selective BChE inhibitors (BIs) may offer symptomatic treatment without the harsh side effects of acetylcholinesterase (AChE) inhibitors. In this study, we explore multiple machine learning strategies to identify BIs , optimizing for precision over all other metrics. We compare state-of-the-art supervised contrastive learning (CL) with deep learning (DL) and Random Forest (RF) machine learning, across single and sequential modeling configurations, to identify the best models for BChE selectivity.
View Article and Find Full Text PDF