Deep Learning for Automated Elective Lymph Node Level Segmentation for Head and Neck Cancer Radiotherapy

Abstract

Depending on the clinical situation, different combinations of lymph node (LN) levels define the elective LN target volume in head-and-neck cancer (HNC) radiotherapy. The accurate auto-contouring of individual LN levels could reduce the burden and variability of manual segmentation and be used regardless of the primary tumor location. We evaluated three deep learning approaches for the segmenting individual LN levels I–V, which were manually contoured on CT scans from 70 HNC patients. The networks were trained and evaluated using five-fold cross-validation and ensemble learning for 60 patients with (1) 3D patch-based UNets, (2) multi-view (MV) voxel classification networks and (3) sequential UNet+MV. The performances were evaluated using Dice similarity coefficients (DSC) for automated and manual segmentations for individual levels, and the planning target volumes were extrapolated from the combined levels I–V and II–IV, both for the cross-validation and for an independent test set of 10 patients. The median DSC were 0.80, 0.66 and 0.82 for UNet, MV and UNet+MV, respectively. Overall, UNet+MV significantly (p < 0.0001) outperformed other arrangements and yielded DSC = 0.87, 0.85, 0.86, 0.82, 0.77, 0.77 for the combined and individual level I–V structures, respectively. Both PTVs were also significantly (p < 0.0001) more accurate with UNet+MV, with DSC = 0.91 and 0.90, respectively. The accurate segmentation of individual LN levels I–V can be achieved using an ensemble of UNets. UNet+MV can further refine this result.

Clinical acceptance and dosimetric impact of automatically delineated elective target and organs at risk for head and neck MR-Linac patients

“…Strijbis et al. ( 40 ) segmented individual levels of the lymph nodes, achieving a combined structure DSC of 0.86, exceeding our model in geometrical evaluation. While the results showcase an impressive performance, it is noteworthy that the sizes of the available datasets for CT scans significantly surpass those for MR images.…”

Deep Learning and Registration-Based Mapping for Analyzing the Distribution of Nodal Metastases in Head and Neck Cancer Cohorts: Informing Optimal Radiotherapy Target Volume Design

“…The planning CT datasets available for the present analysis had a slice thickness of 3 mm, which is in the range radiotherapy head and neck image analysis methods are currently developed and evaluated on [30,42]. A lower slice thickness will increase the resolution in the z-dimension but also the computational time for the mapping analysis.…”

Clinical acceptability of automatically generated lymph node levels and structures of deglutition and mastication for head and neck radiation therapy

“…Sources of such changes can be driven by changes in contouring guidelines, improvements in image-guidance technology, or evolving evidence that alters our understanding of the balance between toxicity and tumor control [28] , [29] . Strijbis et also noted in their work on automated segmentation of levels I-V that contours produced by Cardenas et al, are generous, resembling their institution’s PTVs [27] . Based on physician feedback and changes in clinical practice, we sought to develop a new auto-segmentation model that more accurately reflects the narrower treatment volumes utilized in our clinic’s practice today.…”

“…Automatic CT-based segmentation of these lymph node levels (e.g. I-V) is achievable and has been demonstrated by numerous works [10] , [21] , [22] , [23] , [24] , [25] , [26] , [27] . Our clinic previously integrated a deep-learning based approach to contour elective lymph node levels in CT scans [10] .…”