Developing Novel Technology for Data Driven Clinical Decision Making

胶质母细胞瘤(GBM)脑瘤是致命的，无法治愈。迫切需要有效的治疗方案来治疗基底膜。手术切除基底膜肿瘤的范围对患者的生存有很大的积极影响，但对正常脑组织的损害可能严重影响生活质量，因此必须将其降至最低。因此，我们必须开发手术切除工具，该工具可以：A)精确去除癌细胞，同时减少对正常脑细胞的热影响；B)在切除时向外科医生提供反馈，使他们知道哪些脑组织是恶性的，需要切除；以及C)为医生提供癌症性质的适应证，以便可以对术后遗留的任何癌细胞实施最有效的治疗。这项技术的发展将有助于彻底改变外科医生治疗癌症的方式。为了做到这一点，我们将开发高精度的激光，并使用非常高清晰度的相机了解它们对组织的影响，然后我们将收集这些激光在切割时产生的“烟雾”，并将其送入质谱仪，该质谱仪能够测量代表组织的所有单个代谢物化学物质。当癌症出现时，细胞的新陈代谢是不同的，因此可以使用所描述的技术来测量。观察新陈代谢的变化可以帮助我们了解复杂的疾病，如癌症和可能有效的治疗方法。因此，在这个项目中，我们将从我们的新技术中获取数据，并使用机器学习来了解组织化学成分的变化以及这些变化与癌症不同亚型的相关性，并使用这些信息来预测任何复发疾病的性质。如果我们能从原始组织中预测疾病的复发性质，我们可能就能更有效地使用治疗方法来完全阻止复发。

Glioblastoma (GBM) brain tumours are deadly and incurable. Effective treatment options for GBM are urgently needed. The extent of surgical GBM tumour resection substantially positively impacts patient survival, but damage to normal brain tissue can severely affect quality of life and hence must be minimised. Therefore, it is imperative that we develop tools for surgical removal which can:A) precisely remove cancer cells, with reduced thermal impact on normal brain cells B) give feedback at the time of resection to surgeons so that they know which areas of the brain tissue are malignant and need to be removed and C) give doctors indications of the nature of the cancer so that the most effective therapy can be employed post-operatively for any cancer cells left behind. Development of this technology would be instrumental to revolutionising the way surgeons treat cancer. To do this we will develop high-precision lasers and understand their effects on tissue using very high-definition cameras, we will then collect the 'smoke' produced by these lasers when cutting and feed it into a mass spectrometer which is capable of measuring all of the individual metabolite chemicals representative of the tissue it has come from. When cancer is present, cells metabolise differently and hence this can be measured using the technology described. Observing the changes in metabolism can help us to understand complex diseases such as cancer and the therapies which may be effective against them. Therefore, in this project we will take the data from our novel technology and use machine learning to understand changes in the chemical compositions of the tissues and how that correlates to the different subtypes of cancer and use this information to predict the nature of any recurrent disease. If we can predict the recurrent nature of disease from the primary tissue, we may be able to use therapies more effectively to stop recurrence altogether.