Prediction of Mandibular ORN Incidence from 3D Radiation Dose Distribution Maps Using Deep Learning

Humbert-Vidan, Laia; Patel, Vinod; Andlauer, Robin; King, Andrew P.; Urbano, T. Guerrero

doi:10.1007/978-3-031-17721-7_6

Cited by 6 publications

(6 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A 3D (DN40) [26] convolutional neural network (CNN) was trained on 3D radiation dose distribution maps of the mandible for the binary classification of ORN vs. no ORN subjects using the GSTT cohort [18]. The 3D DN40 CNN (Figure 1) consists of three dense blocks and two transition blocks, where all the convolutional, pooling, batch normalisation and dropout operations are three-dimensional.…”

Section: Methodsmentioning

confidence: 99%

“…The current study aimed to externally validate an existing DL-based ORN prediction model [18]. The model was developed in a UK population treated at Guy’s and St Thomas’ Hospitals (GSTT) and was externally validated in an independent dataset from a Danish population treated at Odense University Hospital (OUH) [9].…”

Section: Introductionmentioning

confidence: 99%

“…The use of radiation dose distribution maps as the dose information for predicting radiation-induced toxicities has been explored with deep learning (DL) methods [16,17]. More recently, this approach has also been introduced in the prediction of mandibular ORN and compared to DVH-based approaches [18,19,20]. However, these studies featured only internal validation on a holdout subset of the data used for training.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

External validation of a deep-learning mandibular ORN prediction model trained on 3D radiation distribution maps

Humbert-Vidan,

Hansen,

Patel

et al. 2023

Preprint

Self Cite

View full text Add to dashboard Cite

Background and purposeMandibular osteoradionecrosis (ORN) is a severe side effect affecting patients undergoing radiation therapy for head and neck cancer. Variations in the bone’s vascularization and composition across the mandible may influence the susceptibility to ORN. Recently, deep learning-based models have been introduced for predicting mandibular ORN using radiation dose distribution maps to incorporate spatial information. These studies, however, only feature internal validation on a holdout subset of the data used for training.Materials and methodsThis study externally validated a 3D DenseNet-40 (DN40) ORN prediction model on an independent dataset. Model performance was evaluated in terms of discrimination and calibration, with Platt scaling applied for improved external calibration. The DN40 model’s discriminative ability on the external dataset was compared to a Random Forest model on corresponding dose-volume histogram (DVH) data.ResultsThe overall model performance was worse at external validation than at internal validation, with Platt scaling improving balance between recall and specificity but not significantly improving the overall calibration. Although the discrimination ability of the DN40 model was slightly lower at external validation (AUROC 0.63 vs. 0.69), this was statistically comparable to that of a DVH-based RF model for the same dataset (p-value 0.667).ConclusionsOur results suggest that, in addition to potential model overfitting issues, dosimetric data distribution differences between the two datasets could explain the low generalisability of the DN40 ORN prediction model. Future work will involve a larger and more diverse cohort.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

External validation of a deep-learning mandibular ORN prediction model trained on 3D radiation distribution maps

Humbert-Vidan,

Hansen,

Patel

et al. 2023

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…Furthermore, recent studies have shown that the toxicity observed can depend on which subregion of a segmented organ was irradiated ( 2 – 4 , 16 – 37 ). Functionally distinct subregions within a single OAR contour may not be accounted for in current treatment planning.…”

Section: Introductionmentioning

confidence: 99%

Impact of radiation dose distribution on nutritional supplementation needs in head and neck cancer radiotherapy: a voxel-based machine learning approach

Madhavan,

Gamez,

Garces

et al. 2024

Front. Oncol.

View full text Add to dashboard Cite

ObjectivesTo investigate the relationship between nutritional supplementation and radiation dose to the pharyngeal constrictor muscles and larynx for head and neck (HN) cancer patients undergoing radiotherapy.MethodsWe retrospectively analyzed radiotherapy (RT) dose for 231 HN cancer patients, focusing on the pharyngeal constrictors and larynx. We defined nutritional supplementation as feeding tube utilization or >10% weight loss from baseline within 90 days after radiotherapy completion. Using deformable image registration (DIR), we mapped each patient’s anatomical structures to a reference coordinate system, and corresponding deformations were applied to dose matrices. Voxel doses were utilized as features for ridge logistic regression models, optimized through 5-fold cross-validation. Model performance was assessed with area under the curve of a receiver operating curve (AUC) and F1 score. We built and compared models using 1) pharyngeal constrictor voxels, 2) larynx voxels, 3) clinical factors and mean regional dose metrics, and 4) clinical factors and dose-volume histogram metrics. Test set AUCs were compared among the models, and feature importance was evaluated.ResultsDIR of the pharyngeal constrictors and larynx yielded mean Dice coefficients of 0.80 and 0.84, respectively. Pharyngeal constrictors voxels and larynx voxel models had AUC of 0.88 and 0.82, respectively. Voxel-based dose modeling identified the superior to middle regions of the pharyngeal constrictors and the superior region of larynx as most predictive of feeding tube use/weight loss. Univariate analysis found treatment setting, treatment laterality, chemotherapy, baseline dysphagia, weight, and socioeconomic status predictive of outcome. An aggregated model using mean doses of pharyngeal constrictors and larynx subregions had an AUC of 0.87 and the model using conventional DVH metrics had an AUC of 0.85 with p-value of 0.04. Feature importance calculations from the regional dose model indicated that mean doses to the superior-middle pharyngeal constrictor muscles followed by mean dose to the superior larynx were most predictive of nutritional supplementation.ConclusionsMachine learning modeling of voxel-level doses enables identification of subregions within organs that correlate with toxicity. For HN radiotherapy, doses to the superior-middle pharyngeal constrictors are most predictive of feeding tube use/weight loss followed by the doses to superior portion of the larynx.

show abstract

“…We congratulate the authors on their study, which notably includes a large cohort of 173 ORN cases and 1086 controls. In our earlier study, 2 we performed a similar comparison, albeit on a smaller but balanced cohort. We therefore take this opportunity to discuss the findings of these studies ( Table 1 ), which we believe is important to set the agenda for future research in this area.…”

mentioning

confidence: 99%

In Regard to Reber et al

Humbert-Vidan

Patel

King

et al. 2023

Advances in Radiation Oncology

View full text Add to dashboard Cite

Prediction of Mandibular ORN Incidence from 3D Radiation Dose Distribution Maps Using Deep Learning

Cited by 6 publications

References 20 publications

External validation of a deep-learning mandibular ORN prediction model trained on 3D radiation distribution maps

External validation of a deep-learning mandibular ORN prediction model trained on 3D radiation distribution maps

Impact of radiation dose distribution on nutritional supplementation needs in head and neck cancer radiotherapy: a voxel-based machine learning approach

In Regard to Reber et al

Contact Info

Product

Resources

About