Cardiac Substructure Segmentation and Dosimetry Using a Novel Hybrid Magnetic Resonance and Computed Tomography Cardiac Atlas

Morris, Eric D.; Ghanem, Ahmed I; Pantelic, Milan; Walker, Eleanor M.; Han, Xiaoxia; Glide-Hurst, Carri K

doi:10.1016/j.ijrobp.2018.11.025

Cited by 37 publications

(47 citation statements)

References 32 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…21 Moreover, our previous MA method required~10 min to generate substructure contours per patient without post-processing. 13 Although the in-plane resolution was 0.7 9 0.7 mm 2 , our study may have been limited by the 8 mm slice thickness of the MRI. Despite our data augmentation techniques, increasing the training sample size may further improve segmentation results.…”

Section: Discussionmentioning

confidence: 96%

“…To generate the contours, a consensus atlas was followed 26 as implemented in our previous work. 13 In brief, 12 cardiac substructures (left/ right ventricles (LV, RV) and atria (LA, RA), superior/inferior venae cavae (SVC, IVC), pulmonary artery/veins (PA, PV), ascending aorta (AA), right coronary artery (RCA), left main coronary artery (LMCA), and LADA) were manually delineated by a radiation oncologist and verified by a radiologist with a cardiac subspecialty. Due to the enhanced soft tissue contrast that MRI provides, preference was given to anatomical information from the MRI.…”

Section: A Imaging and Ground Truth Contour Delineationmentioning

confidence: 99%

“…Several studies have assessed atlas‐based segmentation of cardiac substructures in RT to avoid the time consuming (6–10 h/patient) and tedious task of manual delineation . However, most atlases fail in segmenting the coronary arteries, with Dice similarity coefficients (DSCs) between ground truth and auto‐segmentation of the LADA ranging from 0.09 to 0.27 .…”

Section: Introductionmentioning

confidence: 99%

“…16 However, most atlases fail in segmenting the coronary arteries, with Dice similarity coefficients (DSCs) between ground truth and auto-segmentation of the LADA ranging from 0.09 to 0.27. [13][14][15]17 Incorporating multiple imaging modalities (i.e., contrast enhanced CT and magnetic resonance imaging (MRI)) has improved visualization and yielded successful chamber and great vessel segmentation, yet coronary artery segmentation remains an unmet need. 13,14 Recently, deep neural networks (DNNs), such as U-Net, 18 have shown great promise for generating accurate and rapid delineations for RT.…”

Section: Introductionmentioning

confidence: 99%

“…[13][14][15]17 Incorporating multiple imaging modalities (i.e., contrast enhanced CT and magnetic resonance imaging (MRI)) has improved visualization and yielded successful chamber and great vessel segmentation, yet coronary artery segmentation remains an unmet need. 13,14 Recently, deep neural networks (DNNs), such as U-Net, 18 have shown great promise for generating accurate and rapid delineations for RT. 19 Here, a DNN learns a mapping function between an image and a corresponding feature map (i.e., segmented ground truth).…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Cardiac substructure segmentation with deep learning for improved cardiac sparing

et al. 2019

Self Cite

View full text Add to dashboard Cite

Purpose Radiation dose to cardiac substructures is related to radiation‐induced heart disease. However, substructures are not considered in radiation therapy planning (RTP) due to poor visualization on CT. Therefore, we developed a novel deep learning (DL) pipeline leveraging MRI’s soft tissue contrast coupled with CT for state‐of‐the‐art cardiac substructure segmentation requiring a single, non‐contrast CT input. Materials/methods Thirty‐two left‐sided whole‐breast cancer patients underwent cardiac T2 MRI and CT‐simulation. A rigid cardiac‐confined MR/CT registration enabled ground truth delineations of 12 substructures (chambers, great vessels (GVs), coronary arteries (CAs), etc.). Paired MRI/CT data (25 patients) were placed into separate image channels to train a three‐dimensional (3D) neural network using the entire 3D image. Deep supervision and a Dice‐weighted multi‐class loss function were applied. Results were assessed pre/post augmentation and post‐processing (3D conditional random field (CRF)). Results for 11 test CTs (seven unique patients) were compared to ground truth and a multi‐atlas method (MA) via Dice similarity coefficient (DSC), mean distance to agreement (MDA), and Wilcoxon signed‐ranks tests. Three physicians evaluated clinical acceptance via consensus scoring (5‐point scale). Results The model stabilized in ~19 h (200 epochs, training error <0.001). Augmentation and CRF increased DSC 5.0 ± 7.9% and 1.2 ± 2.5%, across substructures, respectively. DL provided accurate segmentations for chambers (DSC = 0.88 ± 0.03), GVs (DSC = 0.85 ± 0.03), and pulmonary veins (DSC = 0.77 ± 0.04). Combined DSC for CAs was 0.50 ± 0.14. MDA across substructures was <2.0 mm (GV MDA = 1.24 ± 0.31 mm). No substructures had statistical volume differences (P > 0.05) to ground truth. In four cases, DL yielded left main CA contours, whereas MA segmentation failed, and provided improved consensus scores in 44/60 comparisons to MA. DL provided clinically acceptable segmentations for all graded patients for 3/4 chambers. DL contour generation took ~14 s per patient. Conclusions These promising results suggest DL poses major efficiency and accuracy gains for cardiac substructure segmentation offering high potential for rapid implementation into RTP for improved cardiac sparing.

show abstract

Section: Discussionmentioning

confidence: 96%

Section: A Imaging and Ground Truth Contour Delineationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Cardiac substructure segmentation with deep learning for improved cardiac sparing

et al. 2019

Self Cite

View full text Add to dashboard Cite

show abstract

Incorporating sensitive cardiac substructure sparing into radiation therapy planning

Morris

Aldridge²,

Ghanem

et al. 2020

J Applied Clin Med Phys

Self Cite

View full text Add to dashboard Cite

Purpose: Rising evidence suggests that cardiac substructures are highly radiosensitive. However, they are not routinely considered in treatment planning as they are not readily visualized on treatment planning CTs (TPCTs). This work integrated the soft tissue contrast provided by low-field MRIs acquired on an MR-linac via image registration to further enable cardiac substructure sparing on TPCTs. Methods: Sixteen upper thoracic patients treated at various breathing states (7 end-exhalation, 7 end-inhalation, 2 free-breathing) on a 0.35T MR-linac were retrospectively evaluated. A hybrid MR/CT atlas and a deep learning three-dimensional (3D) U-Net propagated 13 substructures to TPCTs. Radiation oncologists revised contours using registered MRIs. Clinical treatment plans were re-optimized and evaluated for beam arrangement modifications to reduce substructure doses. Dosimetric assessment included mean and maximum (0.03cc) dose, left ventricular volume receiving 5Gy (LV-V5), and other clinical endpoints. As metrics of plan complexity, total MU and treatment time were evaluated between approaches. Results: Cardiac sparing plans reduced the mean heart dose (mean reduction 0.7 ± 0.6, range 0.1 to 2.5 Gy). Re-optimized plans reduced left anterior descending artery (LADA) mean and LADA 0.03cc (0.0-63.9% and 0.0-17.3 Gy, respectively). LV 0.03cc was reduced by >1.5 Gy for 10 patients while 6 cases had large reductions (>7%) in LV-V5. Left atrial mean dose was equivalent/reduced in all sparing plans (mean reduction 0.9 ± 1.2 Gy). The left main coronary artery was better spared in all cases for mean dose and D 0.03cc. One patient exhibited >10 Gy reduction in D 0.03cc to four substructures. There was no statistical difference in treatment time and MU, or clinical endpoints to the planning target volume, lung, esophagus, or spinal cord after re-optimization. Four patients benefited from new beam arrangements, leading to further dose reductions. Conclusions: By introducing 0.35T MRIs acquired on an MR-linac to verify cardiac substructure segmentations for CT-based treatment planning, an opportunity was presented for more effective sparing with limited increase in plan complexity. Validation in a larger cohort with appropriate margins offers potential to reduce radiation-related cardiotoxicities.

show abstract

Automatic delineation of cardiac substructures using a region‐based fully convolutional network

et al. 2021

View full text Add to dashboard Cite

Purpose Radiation dose to specific cardiac substructures, such as the atria and ventricles, has been linked to post‐treatment toxicity and has shown to be more predictive of these toxicities than dose to the whole heart. A deep learning‐based algorithm for automatic generation of these contours is proposed to aid in either retrospective or prospective dosimetric studies to better understand the relationship between radiation dose and toxicities. Methods The proposed method uses a mask‐scoring regional convolutional neural network (RCNN) which consists of five major subnetworks: backbone, regional proposal network (RPN), RCNN head, mask head, and mask‐scoring head. Multiscale feature maps are learned from computed tomography (CT) via the backbone network. The RPN utilizes these feature maps to detect the location and region‐of‐interest (ROI) of all substructures, and the final three subnetworks work in series to extract structural information from these ROIs. The network is trained using 55 patient CT datasets, with 22 patients having contrast scans. Threefold cross validation (CV) is used for evaluation on 45 datasets, and a separate cohort of 10 patients are used for holdout evaluation. The proposed method is compared to a 3D UNet. Results The proposed method produces contours that are qualitatively similar to the ground truth contours. Quantitatively, the proposed method achieved average Dice score coefficients (DSCs) for the whole heart, chambers, great vessels, coronary arteries, the valves of the heart of 0.96, 0.94, 0.93, 0.66, and 0.77 respectively, outperforming the 3D UNet, which achieved DSCs of 0.92, 0.87, 0.88, 0.48, and 0.59 for the corresponding substructure groups. Mean surface distances (MSDs) between substructures segmented by the proposed method and the ground truth were <2 mm except for the left anterior descending coronary artery and the mitral and tricuspid valves, and <5 mm for all substructures. When dividing results into noncontrast and contrast datasets, the model performed statistically significantly better in terms of DSC, MSD, centroid mean distance (CMD), and volume difference for the chambers and whole heart with contrast. Notably, the presence of contrast did not statistically significantly affect coronary artery segmentation DSC or MSD. After network training, all substructures and the whole heart can be segmented on new datasets in less than 5 s. Conclusions A deep learning network was trained for automatic delineation of cardiac substructures based on CT alone. The proposed method can be used as a tool to investigate the relationship between cardiac substructure dose and treatment toxicities.

show abstract

Cardiac Substructure Segmentation and Dosimetry Using a Novel Hybrid Magnetic Resonance and Computed Tomography Cardiac Atlas

Cited by 37 publications

References 32 publications

Cardiac substructure segmentation with deep learning for improved cardiac sparing

Cardiac substructure segmentation with deep learning for improved cardiac sparing

Incorporating sensitive cardiac substructure sparing into radiation therapy planning

Automatic delineation of cardiac substructures using a region‐based fully convolutional network

Contact Info

Product

Resources

About