Ultrasound Platform for A Priori Therapy Response Prediction

用于先验治疗反应预测的超声平台

• 批准号: 538814-2019
• 负责人: Kolios, Michael
• 金额: $ 22.03万
• 依托单位: Ryerson University
• 依托单位国家: 加拿大
• 项目类别: Collaborative Health Research Projects
• 财政年份: 2020
• 资助国家: 加拿大
• 起止时间: 2020-01-01 至 2021-12-31
• 项目状态: 已结题
• 来源: https://www.nserc-crsng.gc.ca/ase-oro/Details-Detailles_eng.asp?id=697247
• 关键词: Ultrasound; Platform; Priori; Therapy; Response

In this project will be developing methods to predict chemotherapy and radiotherapy responses inadvance of treatment based on ultrasound and photoacoustic characteristics of tumours and machine learning approaches. We have recently discovered that quantitative ultrasound and optical spectroscopy parameters obtained before any chemotherapy is administered can be used to predictchemotherapy outcomes. First results have demonstrated sensitivities and specificities of over 85-90% each for predicting ultimate clinical response based on quantitative ultrasound data before starting 4-6 months of chemotherapy or many weeks of radiotherapy (1,2). The research will develop methods to predict tumour responses to treatment. We will develop a platform for use in patients; optimal ultrasound and photoacoustic predictive characteristics for tumours will be determined based on patient data and clinical outcomes using traditional classification methods and machine learning approaches. The overarching goal of this research is to transform ultrasound and photoacoustics, which are non-invasive, inexpensive and portable, such that they can be used to predict cancer therapy outcomes through functional imaging. By predicting tumour response in advance and/or early on during a treatment rather than using traditional anatomical assessments that take place weeks to months after the completion of 4-6 months of therapy, ineffective therapies could be avoided in favour of more efficacious treatments. This has the potential of increasing the response rate which is associated with improved long term outcome improving survival in addition to sparing patients unnecessary side effects of treatments that may span weeks to months.

在该项目中，将根据肿瘤的超声和光声特性以及机器学习方法，开发预测化疗和放疗反应的方法。 我们最近发现，定量超声和光学光谱参数获得任何化疗前管理可以用来预测化疗的结果。初步结果表明，基于定量超声预测最终临床反应的灵敏度和特异性均超过85- 90 开始4-6个月化疗或多周放疗前的数据（1，2）。该研究将开发预测肿瘤对治疗反应的方法。我们将开发一个用于患者的平台;肿瘤的最佳超声和光声预测特征将基于患者数据和临床结果，使用传统的分类方法和机器学习方法来确定。这项研究的首要目标是改变超声和光声，它们是非侵入性的，廉价的和便携的，这样它们就可以用来预测癌症。 通过功能成像的治疗效果。通过在治疗期间提前和/或早期预测肿瘤反应，而不是使用在完成4-6个月治疗后数周至数月进行的传统解剖学评估，可以避免无效治疗，从而有利于更有效的治疗。这有可能 增加应答率，这与改善长期结果、提高存活率有关，此外还使患者免于可能持续数周至数月的治疗的不必要的副作用。