Different approaches to volume assessment of lymph nodes in computer tomography scans of head and neck squamous cell carcinoma in comparison with a real gold standard

摘要

Background: Volume assessment in head and neck squamous cell carcinoma (HNSCC) is becoming a more and more clinical important parameter, especially in treatment planning and response control. Various authors showed a significant impact of tumour volume on treatment outcome and local control. Regarding the increasing impact of induction chemotherapy and primary chemoradiation on HNSCC, the need for an adequate measuring tool to judge treatment response becomes obvious. This study was performed to compare the momentary 'gold standard', the diameter-based approach, and tumour volume assessment in HNSCC with approaches based on segmentation algorithms in computer tomography (CT) scans. Methods: CT scans were taken as part of the standardized staging investigations. Using these image data, 30 lymph nodes were defined and segmented. The segmentations were carried out with the newly developed software called 'NeckSegmenter'. After obtaining informed consent from the patient, neck dissection was performed and the excised lymph nodes underwent analysis of their true volume. The datasets were compared with each other and put in correlation with the segmented volumes. Results: Pearson's correlation index showed a higher correlation of the segmented volumes (r = 0.7979) with the true volumes than the results generated via diameter-based equation (r = 0.7974). Furthermore, the diameter-generated volumes show clearly too high volumes at 130% (confidence interval: 107.7-156.7%). The volumes generated with the segmentation are at 89.18% (confidence interval: 73.52-108.16%). Conclusion: The data show a higher reliability for volumes estimated by the segmentation-based approach than the widely used diameter-based approach.