Ultrasound-based multiregional radiomics analysis to differentiate breast masses

We have read the article titled “Multiregional radiomic model for breast cancer diagnosis: value of ultrasound-based peritumoral and parenchymal radiomics” by Guo et al. (1) with great interest. The study aimed to classify breast lesions using a multiregional (intratumoral, peritumoral, and parenchymal) radiomics model based on ultrasound imaging. The authors highlighted that the multiregional model demonstrated superior performance to the intratumoral model in distinguishing malignant from benign breast lesions. We commend the authors for their excellent work in exploring the potential clinical application value of the multiregional radiomics model in classifying benign and malignant breast lesions. However, to strengthen the study, some technical issues are suggested for consideration.

Firstly, the authors employed receiver operating characteristic (ROC) curves to determine discrimination performance, calibration curves to appraise diagnostic accuracy, and integrated discrimination improvement (IDI) to assess the performance enhancement resulting from the added diagnostic features. Nonetheless, decision curve analysis (DCA), which is one of the evaluation systems for radiomics models, should also be considered (2). By employing DCA, the practical utility of established models can be evaluated, allowing for quantification of the net benefit to patients across varying risk probabilities. Even with superior diagnostic performance, a diagnostic model will be ineffective if it offers low clinical benefit and lacks practical clinical application. DCA has been widely applied in evaluating the clinical benefits of radiomics models from multiple dimensions. Secondly, the radiomics model proposed in the present study did not incorporate medical information. Numerous studies have demonstrated the effective differentiation of benign from malignant breast nodules using specific clinical parameters (3,4). Hence, adding these parameters is worthwhile to construct a more comprehensive model. Thirdly, although the authors proposed a radiomics model, the present study did not explore its transformation into a practical diagnostic tool. Development of an online web-based calculator to assist clinicians in applying the developed model to their actual clinical practice should be considered (5). Lastly, despite using the random forest algorithm for model construction, the authors did not explain its benefits. It may be beneficial for the authors to compare the diagnostic effectiveness of various machine learning algorithms (6).

During the analysis of the radiomics model, it is critical to take into account not only diagnostic performance but also clinical utility. In order to construct a more accurate diagnostic model, it is essential to incorporate relevant information about multiple clinical variables and develop and compare multiple algorithmic models.

Supplementary file: Response to “Ultrasound-based multiregional radiomics analysis to differentiate breast masses”.

Acknowledgments

Funding: None.

Footnote

Provenance and Peer Review: This article was a standard submission to the journal. The article did not undergo external peer review.

Conflicts of Interest: Both authors have completed the ICMJE uniform disclosure form (available at https://qims.amegroups.com/article/view/10.21037/qims-23-675/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Guo S, Huang X, Xu C, Yu M, Li Y, Wu Z, Zhou A, Xu P. Multiregional radiomic model for breast cancer diagnosis: value of ultrasound-based peritumoral and parenchymal radiomics. Quant Imaging Med Surg 2023;13:3127-39. [Crossref] [PubMed]
2. Vickers AJ, Cronin AM, Elkin EB, Gonen M. Extensions to decision curve analysis, a novel method for evaluating diagnostic tests, prediction models and molecular markers. BMC Med Inform Decis Mak 2008;8:53. [Crossref] [PubMed]
3. Niu Z, Tian JW, Ran HT, Ren WD, Chang C, Yuan JJ, et al. Risk-predicted dual nomograms consisting of clinical and ultrasound factors for downgrading BI-RADS category 4a breast lesions - A multiple centre study. J Cancer 2021;12:292-304. [Crossref] [PubMed]
4. Pfob A, Sidey-Gibbons C, Barr RG, Duda V, Alwafai Z, Balleyguier C, et al. The importance of multi-modal imaging and clinical information for humans and AI-based algorithms to classify breast masses (INSPiRED 003): an international, multicenter analysis. Eur Radiol 2022;32:4101-15. [Crossref] [PubMed]
5. Chen Z, Chen J, Ying TC, Chen H, Wu C, Chen X, Huang Y, Su Z. Development and Deployment of a Novel Diagnostic Tool Based on Conventional Ultrasound for Fibrosis Assessment in Chronic Kidney Disease. Acad Radiol 2023;S1076-6332(23)00091-0. Epub ahead of print. [Crossref]
6. Zou Y, Shi Y, Liu J, Cui G, Yang Z, Liu M, Sun F. A Comparative Analysis of Six Machine Learning Models Based on Ultrasound to Distinguish the Possibility of Central Cervical Lymph Node Metastasis in Patients With Papillary Thyroid Carcinoma. Front Oncol 2021;11:656127. [Crossref] [PubMed]