CLEAR item#18
“Definition of the reference standard (i.e., outcome variable). Describe the reference standard or outcome measure that the radiomic approach will predict (e.g., pathological grade, histopathological subtypes, genomic markers, local–regional control, survival, etc.). Provide the rationale for the choice of the reference standard (e.g., higher reproducibility rates). Clearly state the reproducibility concerns, potential biases, and limitations of the reference standard.” [1] (from the article by Kocak et al.; licensed under CC BY 4.0)
Reporting examples for CLEAR item#18
Example#1. “Methylation patterns in the CpG islands of MGMT were determined using pathological specimens after surgery by the chemical modification of unmethylated cytosines to uracils, followed by methylation-specific polymerase chain reaction (PCR) using primers specific for either methylated or modified, unmethylated DNA. […] The MGMTp methylation status was analyzed only on a part of the tumor and, therefore, these data could not describe the possible intra-tumoral heterogeneity.” [2] (from the article by Doniselli et al.; licensed under CC BY 4.0)
Example#2. “Data were labelled as malignant or benign based on histopathology using Murine Double Minute 2 (MDM2) gene amplification by fluorescence in situ hybridisation (FISH).” [3] (from the article by Fradet et al.; licensed under CC BY 4.0)
Example#3. “The reference standard was the pathology from surgical specimen. […] The pCR was classified according to Miller-Payne grade: grade 1 for no reduction, grade 2 for minor loss (≤ 30%), grade 3 for loss from 30 to 90%, grade 4 for marked loss (> 90%), and grade 5 for no residual invasive cancer. Patients with grades 1, 2, 3, or 4 were scored as non-pCR.” [4] (from the article by Fusco et al.; licensed under CC BY 4.0)
Explanation and elaboration of CLEAR item#18
The reference standard, or more specifically, the outcome variable that the radiomic approach predicts, should be carefully described in a radiomic article. Even in cases when the reference standard is widely accepted, it is necessary to justify the choice with an explanation based on pertinent citations from the literature. Furthermore, the paper should openly express worries about the repeatability of the reference standard, identifying and resolving any biases that might be present in its application. Moreover, the limitations related to using this reference standard inside the study’s framework need to be made clear. Such a thorough description guarantees not only a clear comprehension of the reference standard but also a critical assessment of its suitability and dependability within the framework of the radiomic analysis that is provided in the paper.
References
- Kocak B, Baessler B, Bakas S, et al (2023) CheckList for EvaluAtion of Radiomics research (CLEAR): a step-by-step reporting guideline for authors and reviewers endorsed by ESR and EuSoMII. Insights Imaging 14:75. https://doi.org/10.1186/s13244-023-01415-8
- Doniselli FM, Pascuzzo R, Agrò M, et al (2024) Development of A Radiomic Model for MGMT Promoter Methylation Detection in Glioblastoma Using Conventional MRI. Int J Mol Sci 25:138. https://doi.org/10.3390/ijms25010138
- Fradet G, Ayde R, Bottois H, et al (2022) Prediction of lipomatous soft tissue malignancy on MRI: comparison between machine learning applied to radiomics and deep learning. Eur Radiol Exp 6:41. https://doi.org/10.1186/s41747-022-00295-9
- Fusco R, Granata V, Maio F, et al (2020) Textural radiomic features and time-intensity curve data analysis by dynamic contrast-enhanced MRI for early prediction of breast cancer therapy response: preliminary data. Eur Radiol Exp 4:8. https://doi.org/10.1186/s41747-019-0141-2