Predictive value of MRI-based deep learning model for lymphovascular invasion status in node-negative invasive breast cancer.

Sci Rep

Department of Breast Imaging, Tianjin Medical University Cancer Institute and Hospital, National Clinical Research Center for Cancer, West Huan-Hu Road, Ti Yuan Bei, Hexi District, Tianjin, 300060, People's Republic of China.

Published: July 2024

AI Article Synopsis

  • This study evaluated how well a deep learning model could predict lymphovascular invasion (LVI) status using breast MRI images in patients with invasive breast cancer that did not have axillary lymph node metastasis.
  • Data was analyzed from 280 patients, with various machine learning and deep learning algorithms used to identify key features that predict LVI status.
  • The deep learning model that combined radiomic and clinical features had the best performance, achieving an area under the curve (AUC) of 0.896 for predicting LVI, outperforming other models used in the study.

Article Abstract

To retrospectively assess the effectiveness of deep learning (DL) model, based on breast magnetic resonance imaging (MRI), in predicting preoperative lymphovascular invasion (LVI) status in patients diagnosed with invasive breast cancer who have negative axillary lymph nodes (LNs). Data was gathered from 280 patients, including 148 with LVI-positive and 141 with LVI-negative lesions. These patients had undergone preoperative breast MRI and were histopathologically confirmed to have invasive breast cancer without axillary LN metastasis. The cohort was randomly split into training and validation groups in a 7:3 ratio. Radiomics features for each lesion were extracted from the first post-contrast dynamic contrast-enhanced (DCE)-MRI. The Least Absolute Shrinkage and Selection Operator (LASSO) regression method and logistic regression analyses were employed to identify significant radiomic features and clinicoradiological variables. These models were established using four machine learning (ML) algorithms and one DL algorithm. The predictive performance of the models (radiomics, clinicoradiological, and combination) was assessed through discrimination and compared using the DeLong test. Four clinicoradiological parameters and 10 radiomic features were selected by LASSO for model development. The Multilayer Perceptron (MLP) model, constructed using both radiomic and clinicoradiological features, demonstrated excellent performance in predicting LVI, achieving a high area under the curve (AUC) of 0.835 for validation. The DL model (MLP-radiomic) achieved the highest accuracy (AUC = 0.896), followed by DL model (MLP-combination) with an AUC of 0.835. Both DL models were significantly superior to the ML model (RF-clinical) with an AUC of 0.720. The DL model (MLP), which integrates radiomic features with clinicoradiological information, effectively aids in the preoperative determination of LVI status in patients with invasive breast cancer and negative axillary LNs. This is beneficial for making informed clinical decisions.

Download full-text PDF

Source
http://www.ncbi.nlm.nih.gov/pmc/articles/PMC11246470PMC
http://dx.doi.org/10.1038/s41598-024-67217-0DOI Listing

Publication Analysis

Top Keywords

invasive breast
16
breast cancer
16
radiomic features
12
deep learning
8
model
8
learning model
8
lymphovascular invasion
8
lvi status
8
status patients
8
cancer negative
8

Similar Publications

Breast cancer is a significant health challenge worldwide, and disproportionately affects women of African ancestry (AA) who experience higher mortality rates relative to other racial/ethnic groups. Several studies have pointed to biological factors that affect breast cancer outcomes. A recently discovered stromal cell population that expresses P ROCR, Z EB1 and P DGFRα (PZP cells) was found to be enriched in normal healthy breast tissue from AA donors, and only in tumor adjacent tissues from donors of European ancestry (EA).

View Article and Find Full Text PDF

Automatic image generation and stage prediction of breast cancer immunobiological through a proposed IHC-GAN model.

BMC Med Imaging

January 2025

Electronics and Communications, Arab Academy for Science, Heliopolis, Cairo, 2033, Egypt.

Invasive breast cancer diagnosis and treatment planning require an accurate assessment of human epidermal growth factor receptor 2 (HER2) expression levels. While immunohistochemical techniques (IHC) are the gold standard for HER2 evaluation, their implementation can be resource-intensive and costly. To reduce these obstacles and expedite the procedure, we present an efficient deep-learning model that generates high-quality IHC-stained images directly from Hematoxylin and Eosin (H&E) stained images.

View Article and Find Full Text PDF

Time-dependent diffusion MRI and kinetic heterogeneity as potential imaging biomarkers for diagnosing suspicious breast lesions with 3.0-T breast MRI.

Magn Reson Imaging

January 2025

Department of Radiology, Beijing Hospital, National Center of Gerontology, Institute of Geriatric Medicine, Chinese Academy of Medical Sciences, No. 1 Da Hua Road, Dong Dan, Beijing 100730, PR China; Graduate School of Peking Union Medical College, Chinese Academy of Medical Sciences, Beijing, PR China.

Purpose: This study aimed to evaluate the diagnostic efficacy of time-dependent diffusion magnetic resonance imaging (td-dMRI) and dynamic contrast-enhanced MRI (DCE-MRI)-based kinetic heterogeneity in differentiating suspicious breast lesions (categorised as Breast Imaging Reporting and Data System 4 or 5).

Methods: This prospective study included 51 females with suspicious breast lesions who underwent preoperative breast MRI, including DCE-MRI and td-dMRI. Six kinetic parameters, namely peak, persistent, plateau, washout component, predominant curve type, and heterogeneity, were extracted from the DCE series using MATLAB and SPM software.

View Article and Find Full Text PDF

Triple-negative breast cancer (TNBC) is highly prone to early relapse and metastasis following standard treatment. CXCL8 is a key factor in tumor invasion and metastasis, but its role in TNBC prognosis and clinicopathological correlations remains poorly understood. This study investigated CXCL8 expression and its clinical significance in TNBC to develop a prognostic nomogram for guiding intensive treatment and follow-up strategies.

View Article and Find Full Text PDF

[Factors associated with breastfeeding one month after discharge in very low weight premature infants hospitalized at birth].

Andes Pediatr

October 2024

Instituto de Enfermería, Facultad de Medicina, Universidad Austral de Chile, Valdivia, Chile.

Article Synopsis
  • The study shows that breastfeeding (BF) has significant benefits for both full-term and premature newborns, but maintaining it after hospitalization can be challenging, especially for preterm infants.
  • The research analyzed data from 218 infants who were breastfeeding at discharge from a neonatology unit, revealing that 35.8% had stopped breastfeeding one month later.
  • Factors influencing BF cessation included lower gestational age and birth weight, with mothers who were married and had higher educational levels being less likely to stop breastfeeding.
View Article and Find Full Text PDF

Want AI Summaries of new PubMed Abstracts delivered to your In-box?

Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!