2021
DOI: 10.3390/cancers13102447
|View full text |Cite
|
Sign up to set email alerts
|

MRI-Based Radiomics Analysis for the Pretreatment Prediction of Pathologic Complete Tumor Response to Neoadjuvant Systemic Therapy in Breast Cancer Patients: A Multicenter Study

Abstract: This retrospective study investigated the value of pretreatment contrast-enhanced Magnetic Resonance Imaging (MRI)-based radiomics for the prediction of pathologic complete tumor response to neoadjuvant systemic therapy in breast cancer patients. A total of 292 breast cancer patients, with 320 tumors, who were treated with neo-adjuvant systemic therapy and underwent a pretreatment MRI exam were enrolled. As the data were collected in two different hospitals with five different MRI scanners and varying acquisit… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

1
15
0

Year Published

2021
2021
2024
2024

Publication Types

Select...
8
1

Relationship

0
9

Authors

Journals

citations
Cited by 26 publications
(16 citation statements)
references
References 59 publications
1
15
0
Order By: Relevance
“…In this study, we used routine contrast-enhanced CT and clinical data to build a prediction model with high accuracy (AUC 0.87). Compared to previous studies using MRI (AUC from 0.52 to 0.98) or PET/CT (AUC from 0.70 to 0.73) [ 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 ], pCR could be predicted from the available images without any additional imaging studies, saving significant costs (time and money). Use of a relatively large sample size and the standardized radiomics feature definition and extraction process are additional strengths of this study, rendering this process more reproducible and comparable.…”
Section: Discussionmentioning
confidence: 86%
See 1 more Smart Citation
“…In this study, we used routine contrast-enhanced CT and clinical data to build a prediction model with high accuracy (AUC 0.87). Compared to previous studies using MRI (AUC from 0.52 to 0.98) or PET/CT (AUC from 0.70 to 0.73) [ 8 , 9 , 10 , 11 , 12 , 13 , 14 , 15 , 16 ], pCR could be predicted from the available images without any additional imaging studies, saving significant costs (time and money). Use of a relatively large sample size and the standardized radiomics feature definition and extraction process are additional strengths of this study, rendering this process more reproducible and comparable.…”
Section: Discussionmentioning
confidence: 86%
“…Several studies have also investigated the value of breast MRI for the a priori prediction of treatment response to NST, where a combination of radiomics and deep learning methods was used to build prediction models. However, the results varied significantly, with AUC ranging from 0.52 to 0.98 [ 10 , 11 , 12 , 13 , 14 , 15 , 16 ]. Because of the variations in pulse sequence selection, image acquisition, reconstruction parameters, and feature extraction, the performance of MR-based prediction models is limited due to a lack of reproducibility and generalizability.…”
Section: Introductionmentioning
confidence: 99%
“…Dataset 2. Dataset 2 consists of the images from 134 subjects with histologically confirmed invasive breast cancer imaged between 2011 and 2017 in Maastricht University Medical Cen-ter+ and collected retrospectively (Granzier et al (2020(Granzier et al ( , 2021).…”
Section: Datasetsmentioning
confidence: 99%
“…Nevertheless, we do agree that the relationship of MRI-structure-based analysis and BRAF V600E mutation was likely worth studying even further in order to improve the potential pretreatment medication needed detected by MRI-based prediction in patients with cancer or carcinoma such as PTC. 6 , 7…”
Section: Dear Editormentioning
confidence: 99%