Automatic segmentation of medical images is an important task for many clinical applications. In practice, a wide range of anatomical structures are visualised using different imaging modalities. In this paper, we investigate whether a single convolutional neural network (CNN) can be trained to perform different segmentation tasks. A single CNN is trained to segment six tissues in MR brain images, the pectoral muscle in MR breast images, and the coronary arteries in cardiac CTA. The CNN therefore learns to identify the imaging modality, the visualised anatomical structures, and the tissue classes. For each of the three tasks (brain MRI, breast MRI and cardiac CTA), this combined training procedure resulted in a segmentation performance equivalent to that of a CNN trained specifically for that task, demonstrating the high capacity of CNN architectures. Hence, a single system could be used in clinical practice to automatically perform diverse segmentation tasks without task-specific training.
Results suggest that parenchymal enhancement in the contralateral breast of patients with invasive unilateral breast cancer is significantly associated with long-term outcome, particularly in patients with estrogen receptor-positive, human epidermal growth factor receptor 2-negative breast cancer. Lower value of the mean top 10% enhancement of the parenchyma shows potential as a predictive biomarker for relatively poor outcome in patients who undergo endocrine therapy. These results should, however, be validated in a larger study.
ObjectivesTo assess whether contralateral parenchymal enhancement reproduces as an independent biomarker for patient survival in an independent patient cohort from a different cancer institution.MethodsThis is a HIPAA-compliant IRB approved retrospective study. Patients with ER-positive/HER2-negative operable invasive ductal carcinoma and preoperative dynamic contrast-enhanced MRI were consecutively included between 2005 and 2009. The parenchyma of the breast contralateral to known cancer was segmented automatically on MRI and contralateral parenchymal enhancement (CPE) was calculated. CPE was split into tertiles and tested for association with invasive disease-free survival (IDFS) and overall survival (OS). Propensity score analysis with inverse probability weighting (IPW) was used to adjust CPE for patient and tumour characteristics as well as systemic therapy.ResultsThree hundred and two patients were included. The median age at diagnosis was 48 years (interquartile range, 42-57). Median follow-up was 88 months (interquartile range, 76-102); 15/302 (5%) patients died and 37/302 (13%) had a recurrence or died. In context of multivariable analysis, IPW-adjusted CPE was associated with IDFS [hazard ratio (HR) = 0.27, 95% confidence interval (CI) = 0.05-0.68, p = 0.004] and OS (HR = 0.22, 95% CI = 0.00-0.83, p = 0.032).ConclusionsContralateral parenchymal enhancement on pre-treatment dynamic contrast-enhanced MRI as an independent biomarker of survival in patients with ER-positive/HER2-negative breast cancer has been upheld in this study. These findings are a promising next step towards a practical and inexpensive test for risk stratification of ER-positive/HER2-negative breast cancer.Key points
• High parenchymal-enhancement in the disease-free contralateral breast reproduces as biomarker for survival.
• This is in patients with ER-positive/HER2-negative breast cancer from an independent cancer centre.
• This is independent of patient and pathology parameters and systemic therapy.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.