Novelty in Public Health and Epidemiology Informatics.
| Autor: | Diallo G; Inria SISTM, Team AHeaD - INSERM Bordeaux Population Health Research Center, Univ. Bordeaux, Bordeaux, France., Bordea G; Team AHeaD - Inserm BPH Research Center & LaBRI UMR 5800, Univ. Bordeaux, Bordeaux, France. |
|---|---|
| Jazyk: | angličtina |
| Zdroj: | Yearbook of medical informatics [Yearb Med Inform] 2022 Aug; Vol. 31 (1), pp. 273-275. Date of Electronic Publication: 2022 Dec 04. |
| DOI: | 10.1055/s-0042-1742526 |
| Abstrakt: | Objectives: To highlight novelty studies and current trends in Public Health and Epidemiology Informatics (PHEI). Methods: Similar to last year's edition, a PubMed search of 2021 scientific publications on PHEI has been conducted. The resulting references were reviewed by the two section editors. Then, 11 candidate best papers were selected from the initial 782 references. These papers were then peer-reviewed by selected external reviewers. They included at least two senior researchers, to allow the Editorial Committee of the 2022 IMIA Yearbook edition to make an informed decision for selecting the best papers of the PHEI section. Results: Among the 782 references retrieved from PubMed, two were selected as the best papers. The first best paper reports a study which performed a comprehensive comparison of traditional statistical approaches (e.g., Cox Proportional Hazards models) vs. machine learning techniques in a large, real-world dataset for predicting breast cancer survival, with a focus on explainability. The second paper describes the engineering of deep learning models to establish associations between ocular features and major hepatobiliary diseases and to advance automated screening and identification of hepatobiliary diseases from ocular images. Conclusion: Overall, from this year edition, we observed that the number of studies related to PHEI has decreased. The findings of the two studies selected as best papers on the topic suggest that a significant effort is still being made by the community to compare traditional learning methods with deep learning methods. Using multimodality datasets (images, texts) could improve approaches for tackling public health issues. Competing Interests: Disclosure The authors report no conflicts of interest in this work. (IMIA and Thieme. This is an open access article published by Thieme under the terms of the Creative Commons Attribution-NonDerivative-NonCommercial License, permitting copying and reproduction so long as the original work is given appropriate credit. Contents may not be used for commercial purposes, or adapted, remixed, transformed or built upon. (https://creativecommons.org/licenses/by-nc-nd/4.0/).) |
| Databáze: | MEDLINE |
| Externí odkaz: |
|