Epilepsy as a dynamic disease

这个项目将建立在一个数学神经元群体活动模型，最近在诺丁汉开发（使用统计物理和非线性动力学的混合），并扩展到包括生理上重要的成分离子在细胞外空间的扩散，以及由此产生的影响皮层组织内的电通信。这对于临床EEG信号的机械解释至关重要，该模型将被验证，并将借鉴伯明翰世界领先的生物医学专业知识。来自机器学习的工具将与机械模型并列，以创建一个白盒，用于使用新的观察数据预测个体患者未来的大脑状态。这将通过预测关键过渡的例行程序得到进一步加强，从而有能力避免或管理下游缉获的灾难性后果。

This project will build on a mathematical neuronal population activity model, recently developed at Nottingham (using a mixture of statistical physics and nonlinear dynamics), and extend this to include physiologically important components for the diffusion of ions in the extracellular space, and the resulting effect on electrical communication within cortical tissue. This is vitally important for the mechanistic interpretation of clinical EEG signals, against which the model will be validated, and will draw down from world leading biomedical expertise at Birmingham. Tools from machine learning will be juxtaposed with the mechanistic model to create a whitebox for predicting future brain states of individual patients using new observational data. This will further be augmented with routines for the prediction of critical transitions, giving the ability to avert or manage disastrous consequences of downstream seizures.