Related Experiment Video
Updated: May 21, 2025
Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model
Published on: August 16, 2020
Accuracy of MRI-based Node-RADS in predicting metastatic lymph node involvement in Head and Neck Squamous Cell
Giovanni Lorusso1, Ilaria Villanova1, Sara Greco1
1Interdisciplinary Department of Medicine, Section of Radiology and Radiation Oncology, University of Bari "Aldo Moro", 70124 Bari, Italy.
Objectives:
This study aimed to assess the accuracy of the Node-RADS scoring system in assessing metastatic lymph node (LN) involvement in patients with head and neck squamous cell carcinoma (HNSCC), a critical factor for treatment planning and prognosis.
Materials And Methods:
A retrospective analysis was conducted on 42 HNSCC patients who underwent preoperative MRI and lymph node dissection. Two radiologists independently evaluated the MRI scans using the Node-RADS system, and the results were compared with postoperative histopathological findings. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and overall accuracy of Node-RADS were calculated, considering LNs with Node-RADS scores of 4 and 5 as positive. LN features such as size, texture, necrosis, border irregularity, and shape were analyzed for their correlation with metastatic involvement. Irregular/ill-defined borders were also evaluated for the detection of Extranodal extension (ENE+).
Results:
Out of 118 LNs assessed, 26 (22.03 %) were metastatic (N+), and 8 (7.78 %) were ENE+. Node-RADS demonstrated a sensitivity of 61.54 %, specificity of 89.69 %, PPV of 76.19 %, NPV of 89.69 %, and an overall accuracy of 87.29 %. Significant correlations were observed between LN characteristics like necrosis, border irregularity, shape, and histopathological results (p < 0.0001).
Conclusion:
The Node-RADS system exhibited high specificity and accuracy in identifying metastatic LNs in HNSCC patients, making it a promising tool for standardizing preoperative imaging assessments. However, further studies are needed to validate its application and improve its integration with advanced imaging techniques for enhanced diagnostic precision.
