Oncotarget: Radiomics for patients with breast cancer using ultrasound


Oncotarget published "Radiomics in predicting recurrence for patients with locally advanced breast cancer using quantitative ultrasound" which reported that the purpose of the study was to investigate the role of pre-treatment quantitative ultrasound -radiomics in predicting recurrence for patients with locally advanced breast cancer.

Patients were determined to have recurrence or no recurrence based on clinical outcomes. With a median follow up of 69 months, 28 patients had disease recurrence.

Using the SVM model for the predicted non-recurrence and recurrence groups, the estimated 5-year recurrence-free survival was 83% and 54%, and the predicted 5-year overall survival was 85% and 74%, respectively.

A QUS-radiomics model using higher-order texture derivatives can identify patients with LABC at higher risk of disease recurrence before starting treatment.

A QUS-radiomics model using higher-order texture derivatives can identify patients with LABC at higher risk of disease recurrence before starting treatment.

Dr. Gregory J. Czarnota said, "Breast cancer is one of the most common cancer in women and is accountable for the leading cause of death."

More popularly known as “radiomics,” advanced imaging analysis has generated promise in disease stratification and predicting clinical outcomes.

Texture analysis from spectral images using gray level co-occurrence matrix can extract second-order imaging features like contrast, correlation, energy, and homogeneity, which can provide insights into different aspects of tumour heterogeneity.

Studies have demonstrated the clinical efficacy of QUS in predicting response to neoadjuvant chemotherapy in LABC, and in patients with head-neck malignancies treated with radiotherapy.

Figure 4:

Figure 4: Predicted survival plots using support vector machine classifier predicted groups (predicted recurrence vs. predicted non-recurrence)-recurrence-free survival (A) and overall survival (OS) (B).

In this study, the authors investigated the role of QUS obtained before the start of treatment in predicting the risk of tumour recurrence in patients with LABC.

The imaging features were obtained from the QUS imaging, which included spectral parameters, texture of spectral parameters, and second-order texture analysis of QUS-Tex1 features.

The Czarnota Research Team concluded in their Oncotarget Research Output, "The study presented here has a relatively small number of patients and has been expanded to continue in a larger population. It is possible with a higher number of patients, advanced strategies like deep learning can be used to improve classification performance and the reliability of generated radiomics models. Although we had a relatively longer follow up (median follow up >6 years in patients without recurrence), small groups can exhibit late recurrence with a possible switch of the output groups. Based on the use of QUS-radiomics to predict the response to NAC, a randomized trial is currently underway to study the effect of adaptive chemotherapy (https://clinicaltrials.gov identifier NCT04050228)."

Sign up for free Altmetric alerts about this article

DOI - https://doi.org/10.18632/oncotarget.28139

Correspondence to - Gregory J. Czarnota - gregory.czarnota@sunnybrook.ca

Keywords - radiomics, breast cancer, quantitative ultrasound, recurrence, machine learning

About Oncotarget

Oncotarget is a biweekly, peer-reviewed, open access biomedical journal covering research on all aspects of oncology.

To learn more about Oncotarget, please visit https://www.oncotarget.com or connect with:

SoundCloud - https://soundcloud.com/oncotarget
Facebook - https://www.facebook.com/Oncotarget/
Twitter - https://twitter.com/oncotarget
LinkedIn - https://www.linkedin.com/company/oncotarget
Pinterest - https://www.pinterest.com/oncotarget/
Reddit - https://www.reddit.com/user/Oncotarget/

Oncotarget is published by Impact Journals, LLC please visit https://www.ImpactJournals.com or connect with @ImpactJrnls

Media Contact

Copyright © 2022 Impact Journals, LLC
Impact Journals is a registered trademark of Impact Journals, LLC