Deep learning for automatic head and neck lymph node level delineation provides expert-level accuracy

Background

Deep learning-based head and neck lymph node level (HN_LNL) autodelineation is of high relevance to radiotherapy research and clinical treatment planning but still underinvestigated in academic literature. In particular, there is no publicly available open-source solution for large-scale autosegmentation of HN_LNL in the research setting.

Methods

An expert-delineated cohort of 35 planning CTs was used for training of an nnU-net 3D-fullres/2D-ensemble model for autosegmentation of 20 different HN_LNL. A second cohort acquired at the same institution later in time served as the test set (n = 20). In a completely blinded evaluation, 3 clinical experts rated the quality of deep learning autosegmentations in a head-to-head comparison with expert-created contours. For a subgroup of 10 cases, intraobserver variability was compared to the average deep learning autosegmentation accuracy on the original and recontoured set of expert segmentations. A postprocessing step to adjust craniocaudal boundaries of level autosegmentations to the CT slice plane was introduced and the effect of autocontour consistency with CT slice plane orientation on geometric accuracy and expert rating was investigated.

Results

Blinded expert ratings for deep learning segmentations and expert-created contours were not significantly different. Deep learning segmentations with slice plane adjustment were rated numerically higher (mean, 81.0 vs. 79.6, p = 0.185) and deep learning segmentations without slice plane adjustment were rated numerically lower (77.2 vs. 79.6, p = 0.167) than manually drawn contours. In a head-to-head comparison, deep learning segmentations with CT slice plane adjustment were rated significantly better than deep learning contours without slice plane adjustment (81.0 vs. 77.2, p = 0.004). Geometric accuracy of deep learning segmentations was not different from intraobserver variability (mean Dice per level, 0.76 vs. 0.77, p = 0.307). Clinical significance of contour consistency with CT slice plane orientation was not represented by geometric accuracy metrics (volumetric Dice, 0.78 vs. 0.78, p = 0.703).

Conclusions

We show that a nnU-net 3D-fullres/2D-ensemble model can be used for highly accurate autodelineation of HN_LNL using only a limited training dataset that is ideally suited for large-scale standardized autodelineation of HN_LNL in the research setting. Geometric accuracy metrics are only an imperfect surrogate for blinded expert rating.

Chemoradiation as a definitive treatment for cervical lymph node metastases from unknown primary cancer

AIM

To assess the treatment outcomes of chemoradiation for cervical lymph node metastases from an unknown primary site (CUP), and to identify for prognostic factors.

PATIENTS AND METHODS

Thirty patients diagnosed as having CUP, and receiving chemoradiation as a definitive treatment were included in the analysis. Locoregional control (LRC), disease-free survival (DFS), and overall survival (OS) rates were estimated, and the factors affecting treatment outcomes were analyzed.

RESULTS

After a median follow-up period of 25 months for surviving patients, the two- and five-year LRC, DFS, and OS rates were 56%/45%, 46%/36%, and 69%/52%, respectively. On univariate analysis, lower performance status (PS; p=0.001), and limitation of disease to level 2 or 3 lymph nodes (p=0.009) were significantly associated with better DFS. Low PS (p=0.002) was significantly associated with better LRC. No late toxicity of grade 3 or greater was observed.

CONCLUSION

Definitive chemoradiation for CUP was well-tolerated, with improvement of DFS/LRC for those with good PS and disease limited to level 2 or 3 lymph nodes.