Abstract
Background: The prognosis of postoperative tumor deposits (TDs) is worse than positive lymph node metastases alone.
Objective: To detect TDs by using multiplanar high-resolution T2-weighted imaging (HRT2WI).
Material and Methods: This retrospective study enrolled 130 patients with locally advanced rectal cancer (LARC). Using pathology-proven tumor deposits (pTDs) as the gold standard, all patients were divided into the pTDs-negative and pTDs-positive groups, the correlation of clinicopathological factors and image features [such as MRI-detected tumor deposits (mTDs), MRI-detected metastatic lymph node (mLN), MRI-detected extramural vascular invasion (mEMVI), maximal extramural depth (EMD), etc.] with pTDs were analyzed by univariate analysis and multivariate binary logistic regression analysis, and the nomogram was established based on the latter. The diagnostic efficiency was evaluated by the receiver operating characteristic curve (ROC) analysis and area under curve (AUC).
Results: mTDs, mLN, mEMVI, and EMD were significantly different between the pTDs-positive and pTDs-negative groups (P < 0.05), with the AUC of 0.767, 0.746, 0.664 and 0.644, respectively. mTDs and mLN were independent risk factors for pTDs (odds ratio: 5.74 and 3.90, P < 0.05). The AUC, sensitivity, specificity, negative predictive value, and accuracy of the nomogram were 0.814 (95% CI: 0.720 ~ 0.908), 73.9%, 79.4%, 93.4%, and 78.5%, respectively. Seventeen of 23 patients with pTDs were identified as mTDs, with a moderate agreement between pTDs and mTDs (Kappa=0.419).
Conclusion: Multiplanar HRT2WI can be used as a preoperative diagnostic tool to identify TDs in LARC. The combined model constructed by mTDs and mLN shows a good diagnostic performance for TDs.
Keywords: Rectal cancer, Tumor deposits, Lymph node, EMVI, Magnetic resonance imaging, Nomogram.
Call for Papers in Thematic Issues
Deep Learning Approaches for Image Enhancement in Medical Diagnostics
The medical and healthcare sectors are very different from one another. Consumers in these sectors need more affordable, superior offerings. Health or medical specialists are typically the ones who analyse medical data. In particular, medical imaging evaluation heavily relies on machine learning and deep learning. They not only yield outstanding ...read more
Deep Learning for Advanced Image Analysis in Health Informatics
The field of health informatics has endured a sharp increase in the use of image analysis due to the abundance of versatile data. This has also led to a rise in fascination with the development of computational models based on information and deep learning in the field of health informatics. ...read more
Emerging Biomedical Imaging Informatic for Smart Healthcare
Biomedical imaging informatics is a field that deals with the acquisition, processing, and interpretation of images in biomedicine using computational approaches. Biomedical imaging has progressed from early, simple X-rays for fracture diagnosis and foreign body identification to various techniques used to research biological structure and function and answer fundamental biomedical ...read more
Emerging Trends & Techniques in Medical Imaging using Machine Learning and Deep Learning Techniques
This special issue will aim to discuss different approaches, initiatives and applications in Medical Imaging fields (including Machine Intelligence, Mining Engineering, Modeling and Simulation, Computer, Communication, Networking and Information Engineering, Systems Engineering, Innovative Computing Systems, Adaptive Technologies for Sustainable Growth, and Theoretical and Applied Sciences). This collection should inspire various ...read more
23