Development of a magnetic resonance imaging and image-analysis framework for high-precision gamma knife-based stereotactic radiosurgery of brain tumours

脑转移是癌症的常见并发症，发生在所有癌症患者的20-40%。**对于出现有限数量脑转移的患者的标准治疗是立体定向放射手术（SRS）。SRS向目标区域提供高、精确聚焦的单剂量辐射。Sunnybrook最近安装了一种新的图像引导专用设备，用于脑肿瘤的SRS（伽玛刀完美；Elekta，蒙特利尔，加拿大）。**尽管我们在精确地向目标区域提供高剂量辐射的能力方面取得了进步，但我们仍然迫切需要改进我们目前的指导辐射的方法，以确定哪些患者将从SRS中受益，并在治疗后早期评估肿瘤对辐射的反应。为了解决这些缺点，本项目旨在开发和优化一种新的磁共振（MR）成像和图像处理框架，用于使用伽玛刀技术对转移性脑肿瘤进行高精度个性化SRS。**在SRS治疗中，MRI用于描绘肿瘤靶点。然而，这些图像通常是扭曲的，这可能导致“几何缺失”和无效治疗，特别是在小转移目标的情况下。几何上精确的MRI解决方案将用于放疗计划。这些将辅以新的肿瘤代谢MRI扫描，以更准确地描绘治疗目标。结合这些成像方法，新的图像处理技术（称为“放射组学”）将用于开发和优化模型，以预测治疗前或治疗后早期的反应。这样的预测可以帮助临床医生提供有效的辅助治疗，或者在可能为时已晚之前改变无效的治疗方法。**与工业合作伙伴Elekta合作，该项目开发的MRI和放射组学平台可纳入商业SRS治疗计划系统。这些新技术将有助于临床决策并提高产品的适销性

Brain metastasis is a common complication of cancer, occurring in 20-40% of all cancer patients. **The standard of care for patients presenting with a limited number of brain metastases is stereotactic radiosurgery (SRS). SRS delivers a high, precisely-focused, single dose of radiation to a target region. Sunnybrook has recently installed a new image-guided dedicated device for SRS of brain tumours (Gamma Knife Perfexion; Elekta, Montreal, Canada).**Despite advancements in our ability to precisely deliver high radiation doses to target regions, there is still an urgent need to improve upon our current methods to guide radiation, to determine which patients would benefit from SRS, and to assess response of tumours to radiation early after treatment. To address these shortcomings, this project aims to develop and optimize a novel magnetic resonance (MR) imaging and image-processing framework for high-precision personalized SRS of metastatic brain tumours with Gamma Knife technology.**In SRS treatment, MRI is used to delineate the tumour target. However, these images are commonly distorted, which can result in a "geometric miss" and ineffective treatment, especially in the case of small metastatic targets. Geometrically accurate MRI solutions will be developed for radiotherapy planning. These will be complemented by new MRI scans of tumour metabolism, to more accurately delineate treatment targets. In combination with these imaging methods, new image-processing techniques (referred to as "radiomics") will be used to develop and optimize models to predict response prior to or early after a treatment. Such prediction can help clinicians deliver effective adjuvant therapies, or change ineffective treatments before it is potentially too late.**In collaboration with industrial partner Elekta, the MRI and radiomics platform developed in this project can be incorporated into the commercial SRS treatment planning system. The novel technologies will assist clinical decision making and improve marketability of the products.**