METhodological RadiomICs Score (METRICS)

Last revised by Henry Knipe on 6 May 2024

Citation, DOI, disclosures and article data

Citation:

Kocak B, Knipe H, METhodological RadiomICs Score (METRICS). Reference article, Radiopaedia.org (Accessed on 04 Jun 2024) https://doi.org/10.53347/rID-188491

DOI:

https://doi.org/10.53347/rID-188491

Permalink:

https://radiopaedia.org/articles/188491

rID:

188491

Article created:

6 May 2024, Burak Kocak

Disclosures:

At the time the article was created Burak Kocak had no financial relationships to ineligible companies to disclose.

View Burak Kocak's current disclosures

Last revised:

6 May 2024, Henry Knipe ◉

Disclosures:

At the time the article was last revised Henry Knipe had the following disclosures:

Integral Diagnostics, Shareholder (ongoing)
Micro-X Ltd, Shareholder (ongoing)

These were assessed during peer review and were determined to not be relevant to the changes that were made.

View Henry Knipe's current disclosures

Revisions:

4 times, by 2 contributors - see full revision history and disclosures

Sections:

Artificial Intelligence

Tags:

radiomics, artificial intelligence, machine learning, texture analysis, reporting, refs

The METhodological RadiomICs Score (METRICS) is a 30-item quality evaluation tool for artificial intelligence (AI) and radiomics papers ¹. It aims to assess and improve the quality of radiomics research. METRICS is endorsed by the European Society of Medical Imaging Informatics (EuSoMII).

developed by a modified Delphi method
has a conditional format to cover various forms of radiomic pipeline alternatives
covers both hand-crafted and deep learning-based radiomics
the METRICS tool is accessible online (see external links), making calculating the total score easier ¹
METRICS does not primarily aim to improve and assess the reporting quality - CheckList for Evaluation of Radiomics Research (CLEAR), a reporting guideline endorsed by ESR and EuSoMII is highly recommended for this purpose ²