Geometric and Dosimetric Evaluation of Brain Arteriovenous Malformations Auto-Segmentation using Multimodal Imaging in Stereotactic Radiosurgery

Xing Di¹Wenqian Xu¹Yike Xu²Xiaojia Gong¹Tao Jin¹Minghao Sun¹Lei Zhu³Guanghai Mei^1*Xiaoxia Liu^1*Huaguang Zhu^1*

• ¹Department of Neurology, Huashan Hospital, Fudan University, Shanghai, China
• ²Qingdao Central Hospital of University of Health and Rehabilitation Sciences, Qingdao, China
• ³Shandong First Medical University Cancer Hospital, Jinan, China

Background: Radiation dose optimization for white matter (WM) tracts protection during stereotactic radiosurgery (SRS) of brain arteriovenous malformations (bAVMs) requires integration of diffusion tensor imaging-based WM tractography to delineate WM tracts and establish dose constraints. Conventional manual delineation of perilesional targets demonstrated significant operational inefficiency, primarily attributed to complex structural interdigitation between pathological vasculature and eloquent brain areas. Purpose: To develop a two-stage deep learning method that combines 2D U-Net detection-aided and 3D self-attention segmentation model for bAVMs segmentation. This method aim to improve clinical practice efficiency while protecting WM tracts using multimodal imaging and WM tractography in SRS. Methods: We analyze imaging data from 191 patients who underwent Cyberknife-based SRS at Huashan Hospital, Fudan University with eloquent bAVMs. 153 patients are used to construct the ensemble to segment the bAVMs, the other 38 to validate performance. We introduce spatial and channel attention modules in the U-Net variant, as well as a versatile "Attentional ResBlock", achieving parameter efficiency through cross-dimensional interaction while preserving model fidelity. The accuracy of the auto-segmented contours is evaluated with geometric indices and dosimetric endpoints. Results: Our proposed model demonstrated superior segmentation performance, achieving a dice similarity coefficient of 0.84 ± 0.05, Sensitivity of 0.92 ± 0.09, and F2-score of 0.79 ± 0.08. Furthermore, it attained a low Hausdorff distance (4.55 ± 1.14 mm) and Mean surface distance (0.53 ± 0.08 mm), indicating exceptional boundary delineation precision. The difference in the proportion of WM tracts within the target region between manual and automated contours is minimal (0.08 ± 0.13). Meanwhile, strong concordance between auto-segmented and manual-contoured targets is observed across most dosimetric endpoints with mean difference of 0.46 Gy. The received dose of WM tracts in the two comparison plans also have acceptable representation of dosimetric parameters (R2 = 0.92 for Dmean and 0.88 for V1Gy). Dose exposition of the organ at risk is no statistically significant differences in treatment plans with auto-segmentation targets compared to regular plans. Conclusions: The reliable bAVMs automated-segmentation method has been validated and may support SRS planning for bAVMs and thus avoid neurological sequelae after SRS in considering WM tracts protection.