Deep transfer learning based on magnetic resonance imaging can improve the diagnosis of lymph node metastasis in patients with rectal cancer

Abstract

Background: Lymph node (LN) metastasis is the main prognostic factor for local recurrence and overall survival of patients with rectal cancer. The accurate evaluation of LN status in rectal cancer patients is associated with improved treatment and prognosis. This study aimed to apply deep transfer learning to classify LN status in patients with rectal cancer to improve N staging accuracy. Methods:The study included 129 patients with 325 rectal cancer screenshots of LN T2-weighted (T2W) images from April 2018 to March 2019. Deep learning was applied through a pre-trained model, Inception-v3, for recognition and detection of LN status. The results were compared to manual identification by experienced radiologists. Two radiologists reviewed images and independently identified their status using various criteria with or without short axial (SA) diameter measurements. The accuracy, positive predictive value (PPV), negative predictive value (NPV), sensitivity, specificity, and the area under the receiver operating characteristic (ROC) curve (AUC) were calculated.Results: When the same radiologist performed the analysis, the AUC was not significantly different in the presence or absence of LN diameter measurements (P>0.05). In the deep transfer learning method, the PPV, NPV, sensitivity, and specificity were 95.2%, 95.3%, 95.3%, and 95.2%, respectively, and the AUC and accuracy were 0.994 and 95.7%, respectively. These results were all higher than that achieved with manual diagnosis by the radiologists. Conclusions:The internal details of LNs should be used as the main criteria for positive diagnosis when using MRI. Deep transfer learning can improve the MRI diagnosis of positive LN metastasis in patients with rectal cancer.

To Investigate the Effect of Magnetic Resonance Imaging (MRI) and Diffusion Tensor Imaging (DTI) in the Diagnosis of Mild Craniocerebral Injury

“…Gattringer et al analyzed the effect of magnetic resonance diffusion imaging in central system diseases [ 9 ]. Li et al found that in acute cerebral infarction, a large number of water molecules enter the cells, resulting in intracellular edema and reduced diffusion of water molecules inside and outside the cells, resulting in reduced free diffusion of total water molecules at the ischemic site, swelling of white matter fiber bundles at the infarct site, limited movement of water molecules, increased diffusion anisotropy, and increased FA value [ 10 ]. Hong et al found that the angiogenesis of malignant brain tumors is more than that of benign tumors, and the density of tumor cells in the parenchyma is also relatively large.…”

RESOLVE-DWI-based deep learning nomogram for prediction of normal-sized lymph node metastasis in cervical cancer: a preliminary study

“…Firstly, we outlined and extracted image information based on primary tumours rather than lymph nodes [ 29 – 31 ]to promote to the clinical simplicity and operability. On the one hand, it is difficult to evaluate every lymph node on cross-sectional images because of the limitations of imaging range and lymph node size [ 12 ]. On the other hand, LNM is fundamentally driven by the primary tumour, and image-based DL analysis of primary tumours may reveal characteristics that relate to the risk of LNM [ 33 ].…”

MRI‐Based Artificial Intelligence in Rectal Cancer

“…Deep learning is also worthy of further research. Li et al 19 identified and detected LN status by using a pretrained deep learning model (Inception‐v3), with an AUC and accuracy of 0.994% and 95.7%, respectively, which showed better performance than radiologists. Lateral lymph node metastasis influences the surgeon's choice of surgical procedure.…”