ASSET: Auto-Segmentation of the Seventeen SEgments for Ventricular Tachycardia Ablation in Radiation Therapy

SIMPLE SUMMARY: The segmentation of the ventricular tachycardia target is complex, error-prone, and time-consuming when performed manually. Currently, target generation for ventricular tachycardia is performed manually through viewing electro-anatomic mapping which is incredibly error-prone due to the current lack of a method to directly register electro-anatomic data to radiation therapy planning images. This work presents a novel model to automatically provide the 17 segments of the left ventricle as defined by the American Heart Association on any imaging modality for radiation therapy planning. This is completed through the use of principal component analysis. These segments can then be used as an aid for radiation therapy planning for physicians and physicists in numerous ways. Namely, by offering significant improvements to target generation consistency and time-saving measures which in turn offers strong potential for widespread application to institutions conducting radio-ablation of the left ventricular myocardium. ABSTRACT: There has been a recent effort to treat high-risk ventricular tachycardia (VT) patients through radio-ablation. However, manual segmentation of the VT target is complex and time-consuming. This work introduces ASSET, or Auto-segmentation of the Seventeen SEgments for Tachycardia ablation, to aid in radiation therapy (RT) planning. ASSET was retrospectively applied to CTs for 26 thoracic RT patients (13 undergoing VT ablation). The physician-defined parasternal long-axis of the left ventricle (LV) and the axes generated from principal component analysis (PCA) were compared using mean distance to agreement (MDA) and angle of separation. The manually selected right ventricle insertion point and LVs were used to apply the ASSET model to automatically generate the 17 segments of the LV myocardium (LVM). Physician-defined parasternal long-axis differed from PCA by 1.2 ± 0.3 mm MDA and 6.9 ± 0.7 degrees. Segments differed by 0.69 ± 0.29 mm MDA and 0.89 ± 0.03 Dice similarity coefficient. Running ASSET takes <5 min where manual segmentation took >2 h/patient. Agreement between ASSET and expert contours was comparable to inter-observer variability. Qualitative scoring conducted by three experts revealed automatically generated segmentations were clinically useable as-is. ASSET offers efficient and reliable automatic segmentations for the 17 segments of the LVM for target generation in RT planning.