Publications by authors named "E Garcia-Alvarez"
Article Synopsis
- Developing an automatic acne grading system using machine learning requires extensive data and labeling, but the authors created a model that performs well on low-resolution images from fewer patients with uneven acne severity distribution.
- The study utilized 1,374 paired images from 391 patients, expertly labeled, to train a deep learning model which achieved 66.67% accuracy on the test set.
- The results suggest that this model can be a viable tool for medical practitioners and provide standardized assessments for patients, highlighting its potential despite limited data.
Apolipoprotein D (Apo D) overexpression is a general finding across neurodegenerative conditions so the role of this apolipoprotein in various neuropathologies such as multiple sclerosis (MS) has aroused a great interest in last years. However, its mode of action, as a promising compound for the development of neuroprotective drugs, is unknown. The aim of this work was to address the potential of Apo D to prevent the action of cuprizone (CPZ), a toxin widely used for developing MS models, in oligodendroglial and neuroblastoma cell lines.
View Article and Find Full Text PDFUnlabelled: Physical examination is essential in diagnosing tendinous lesions. This is particularly true of the flexor digitorum superficialis of the little finger (FDS5), which is functionally absent in approximately 30% of the population. The objective of our study was to determine the diagnostic value of 3 clinical tests commonly used to assess the function of this tendon.
View Article and Find Full Text PDFEfficiency of pragmatic search strategies to update clinical guidelines recommendations.
BMC Med Res Methodol
July 2015
Background: A major challenge in updating clinical guidelines is to efficiently identify new, relevant evidence. We evaluated the efficiency and feasibility of two new approaches: the development of restrictive search strategies using PubMed Clinical Queries for MEDLINE and the use of the PLUS (McMaster Premium Literature Service) database.
Methods: We evaluated a random sample of recommendations from a national guideline development program and identified the references that would potentially trigger an update (key references) using an exhaustive approach.