Pre-Treatment Prediction of Breast Cancer Response to Neoadjuvant Chemotherapy Using Intratumoral and Peritumoral Radiomics from T2-Weighted and Contrast-Enhanced T1-Weighted MRI DOI Open Access
Deok Hyun Jang, Christopher Kolios, Laurentius O. Osapoetra

et al.

Cancers, Journal Year: 2025, Volume and Issue: 17(9), P. 1520 - 1520

Published: April 30, 2025

(1) Background: Neoadjuvant chemotherapy (NAC) is an integral part of breast cancer management, and response to NAC important prognostic factor associated with improved survival outcomes. However, the current standard for assessment relies on post-surgical histopathological analysis, which limits early therapeutic decision-making treatment personalization. This study aimed develop evaluate a machine learning model that integrates pre-treatment MRI radiomics clinical features predict in patients. (2) Methods: In this study, was developed using magnetic resonance imaging (MRI) data. Radiomic were extracted from contrast-enhanced T1-weighted (CE-T1) T2-weighted (T2) sequences both intratumoral peritumoral segmentations. Furthermore, uniquely examined two criteria: pathologic complete (pCR) versus non-pCR, non-response. A total 254 patients biopsy-confirmed who completed included. (n = 400) 7) analyzed build predictive employing XGBoost classifier. Performance measured terms accuracy, precision, sensitivity, specificity, F1-score, AUC. (3) Results: The integration radiomic data significantly enhanced performance. For pCR non-pCR prediction, combined achieved accuracy 80% AUC 0.85, outperforming (Accuracy 68%, 0.81) 66%, 0.60). Similarly, non-response Accuracy 74% 0.75, 63%, 0.68) 0.57). (4) Conclusions: These findings highlight synergistic effect integrating MRI-based improve has potential enable more precise personalized strategies.

Language: Английский

Radiomic machine learning for predicting prognostic biomarkers and molecular subtypes of breast cancer using tumor heterogeneity and angiogenesis properties on MRI DOI
Ji Young Lee, Kwang‐Sig Lee, Bo Kyoung Seo

et al.

European Radiology, Journal Year: 2021, Volume and Issue: 32(1), P. 650 - 660

Published: July 5, 2021

Language: Английский

Citations

91
Radiomics in breast MRI: current progress toward clinical application in the era of artificial intelligence DOI
Hiroko Satake, Satoko Ishigaki, Rintaro Ito

et al.

La radiologia medica, Journal Year: 2021, Volume and Issue: 127(1), P. 39 - 56

Published: Oct. 26, 2021

Language: Английский

Citations

67
A radiogenomic multimodal and whole-transcriptome sequencing for preoperative prediction of axillary lymph node metastasis and drug therapeutic response in breast cancer: a retrospective, machine learning and international multicohort study DOI Creative Commons
Jianguo Lai, Zijun Chen, Jie Liu

et al.

International Journal of Surgery, Journal Year: 2024, Volume and Issue: 110(4), P. 2162 - 2177

Published: Jan. 11, 2024

Background: Axillary lymph nodes (ALN) status serves as a crucial prognostic indicator in breast cancer (BC). The aim of this study was to construct radiogenomic multimodal model, based on machine learning and whole-transcriptome sequencing (WTS), accurately evaluate the risk ALN metastasis (ALNM), drug therapeutic response avoid unnecessary axillary surgery BC patients. Methods: In study, conducted retrospective analysis 1078 patients from Cancer Genome Atlas (TCGA), Imaging Archive (TCIA), Foshan cohort. These were divided into TCIA cohort ( N =103), validation =51), Duke =138), =106), TCGA =680). Radiological features extracted radiological images differentially expressed gene expression calibrated using technology. A support vector model employed screen genetic features, established clinical pathological predict ALNM. accuracy predictions assessed area under curve (AUC) benefit measured decision analysis. Risk stratification performed by set enrichment analysis, differential comparison immune checkpoint expression, sensitivity testing. Results: For prediction ALNM, rad-score able significantly differentiate between ALN- ALN+ both cohorts P <0.05). Similarly, gene-score nomogram demonstrated satisfactory performance (AUC 0.82, 95% CI: 0.74–0.91) 0.77, 0.63–0.91). sub-stratification there significant differences pathway high low-risk groups Additionally, different may exhibit varying treatment responses Conclusion: Overall, employs data, including images, genetic, clinicopathological typing. can precisely ALNM

Language: Английский

Citations

9
Recent Radiomics Advancements in Breast Cancer: Lessons and Pitfalls for the Next Future DOI Creative Commons
Filippo Pesapane, Anna Rotili, Giorgio Maria Agazzi

et al.

Current Oncology, Journal Year: 2021, Volume and Issue: 28(4), P. 2351 - 2372

Published: June 25, 2021

Radiomics is an emerging translational field of medicine based on the extraction high-dimensional data from radiological images, with purpose to reach reliable models be applied into clinical practice for purposes diagnosis, prognosis and evaluation disease response treatment. We aim provide basic information radiomics radiologists clinicians who are focused breast cancer care, encouraging cooperation scientists mine a better application in practice. investigate workflow as well outlook challenges recent studies. Currently, has potential ability distinguish between benign malignant lesions, predict cancer's molecular subtypes, neoadjuvant chemotherapy lymph node metastases. Even though been used tumor diagnosis prognosis, it still research phase some need faced obtain translation. In this review, we discuss current limitations promises improvement further research.

Language: Английский

Citations

56
Radiomics MRI for lymph node status prediction in breast cancer patients: the state of art DOI

Alessandro Calabrese,

Domiziana Santucci, Roberta Landi

et al.

Journal of Cancer Research and Clinical Oncology, Journal Year: 2021, Volume and Issue: 147(6), P. 1587 - 1597

Published: March 23, 2021

Language: Английский

Citations

49
3D DCE-MRI Radiomic Analysis for Malignant Lesion Prediction in Breast Cancer Patients DOI Creative Commons
Carmelo Militello, Leonardo Rundo, Mariangela Dimarco

et al.

Academic Radiology, Journal Year: 2021, Volume and Issue: 29(6), P. 830 - 840

Published: Sept. 29, 2021

To develop and validate a radiomic model, with features extracted from breast Dynamic Contrast-Enhanced Magnetic Resonance Imaging (DCE-MRI) 1.5T scanner, for predicting the malignancy of masses enhancement. Images were acquired using an 8-channel coil in axial plane. The rationale behind this study is to show feasibility radiomics-powered model that could be integrated into clinical practice by exploiting only standard-of-care DCE-MRI goal reducing required image pre-processing (ie, normalization quantitative imaging map generation).107 manually annotated dataset 111 patients, which was split discovery test sets. A feature calibration step performed find robust non-redundant features. An in-depth analysis define predictive model: purpose, Support Vector Machine (SVM) trained nested 5-fold cross-validation scheme, several unsupervised selection methods. performance evaluated terms Area Under Receiver Operating Characteristic (AUROC), specificity, sensitivity, PPV NPV. on unseen held-out data.The combining Unsupervised Discriminative Feature Selection (UDFS) SVMs average achieved best blinded set: AUROC = 0.725±0.091, sensitivity 0.709±0.176, specificity 0.741±0.114, 0.72±0.093, NPV 0.75±0.114.In study, we built based DCE-MRI, strongest enhancement phase, promising results accuracy differentiation malignant benign lesions.

Language: Английский

Citations

48
Intra- and peritumoral radiomics on assessment of breast cancer molecular subtypes based on mammography and MRI DOI

Shuxian Niu,

Wenyan Jiang,

Nannan Zhao

et al.

Journal of Cancer Research and Clinical Oncology, Journal Year: 2021, Volume and Issue: 148(1), P. 97 - 106

Published: Oct. 8, 2021

Language: Английский

Citations

46
Pretreatment DCE-MRI-Based Deep Learning Outperforms Radiomics Analysis in Predicting Pathologic Complete Response to Neoadjuvant Chemotherapy in Breast Cancer DOI Creative Commons
Yunsong Peng,

Ziliang Cheng,

Chang Gong

et al.

Frontiers in Oncology, Journal Year: 2022, Volume and Issue: 12

Published: March 10, 2022

To compare the performances of deep learning (DL) to radiomics analysis (RA) in predicting pathological complete response (pCR) neoadjuvant chemotherapy (NAC) based on pretreatment dynamic contrast-enhanced MRI (DCE-MRI) breast cancer.This retrospective study included 356 cancer patients who underwent DCE-MRI before NAC and surgery after NAC. Image features kinetic parameters tumors were derived from DCE-MRI. Molecular information was assessed immunohistochemistry results. The image-based RA DL models constructed by adding or molecular image-only linear discriminant (LDA) convolutional neural network (CNN) models. predictive developed receiver operating characteristic (ROC) curve compared with DeLong method.The overall pCR rate 23.3% (83/356). area under ROC (AUROC) image-kinetic-molecular model 0.781 [95% confidence interval (CI): 0.735, 0.828], which higher than that image-kinetic (0.629, 95% CI: 0.595, 0.663; P < 0.001) comparable image-molecular (0.755, 0.708, 0.802; = 0.133). AUROC 0.83 (95% 0.816, 0.847), (0.707, 0.654, 0.761; 0.79, 0.768, 0.812; (0.778, 0.828; 0.001).The DCE-MRI-based is superior patients. has best prediction performance.

Language: Английский

Citations

31
Radiomics as a New Frontier of Imaging for Cancer Prognosis: A Narrative Review DOI Creative Commons
Alfonso Reginelli, Valerio Nardone,

Giuliana Giacobbe

et al.

Diagnostics, Journal Year: 2021, Volume and Issue: 11(10), P. 1796 - 1796

Published: Sept. 29, 2021

The evaluation of the efficacy different therapies is paramount importance for patients and clinicians in oncology, it usually possible by performing imaging investigations that are interpreted, taking consideration response criteria. In last decade, texture analysis (TA) has been developed order to help radiologist quantify identify parameters related tumor heterogeneity, which cannot be appreciated naked eye, can correlated with endpoints, including cancer prognosis. aim this work analyze impact prediction prognosis stratification into pathologies (lung cancer, breast gastric hepatic rectal cancer). Key references were derived from a PubMed query. Hand searching clinicaltrials.gov also used. This paper contains narrative report critical discussion radiomics approaches fields diseases.

Language: Английский

Citations

41
Radiomics features based on automatic segmented MRI images: Prognostic biomarkers for triple-negative breast cancer treated with neoadjuvant chemotherapy DOI
Mingming Ma, Liangyu Gan, Yinhua Liu

et al.

European Journal of Radiology, Journal Year: 2021, Volume and Issue: 146, P. 110095 - 110095

Published: Dec. 4, 2021

Language: Английский

Citations

36