ESR eHealth 21 - How to critically appraise artificial intelligence (AI)/radiomics papers
Lectures
1
Chairperson's introduction
05:00Constance De Margerie-Mellon, Paris / FR
2
On metrics and methods: perils and pitfalls in statistical analyses
15:00Nathaniel Mercaldo, Boston / US
3
Read avidly but think critically: must-have qualities to look for in radiomics research
15:00Renato Cuocolo, Napoli / IT
4
AI reporting guidelines: how to focus on quality and reproducibility
15:00Leonor Cerda Alberich, Valencia / ES
5
Panel discussion: What can we learn from the literature so far? Where do we go next?
10:00Panel discussion: What can we learn from the literature so far? Where do we go next?
Every day, many scientific manuscripts regarding artificial intelligence and radiomics are published. To readers, it is not always clear how to best approach these papers, especially when many technical details are provided. This session will discuss strategies to critically appraise such manuscripts, what qualities to look for and how to identify potential issues in the statistical analyses. Currently, available reporting guidelines will be highlighted to provide readers, as well as potential authors and reviewers, with help objectively assessing a manuscript's reporting quality.
5 min
Chairperson's introduction
Constance De Margerie-Mellon, Paris / France
15 min
On metrics and methods: perils and pitfalls in statistical analyses
Nathaniel Mercaldo, Boston / United States
15 min
Read avidly but think critically: must-have qualities to look for in radiomics research
Renato Cuocolo, Napoli / Italy
15 min
AI reporting guidelines: how to focus on quality and reproducibility
Leonor Cerda Alberich, Valencia / Spain
10 min
Panel discussion: What can we learn from the literature so far? Where do we go next?
PEP Subscription Required
This course is only accessible for ESR Premium Education Package subscribers.