prediction

1. problems

1.1. data set

1.2. algorithm

1.3. false warning

1.4. fail to forcast

1.5. seizure: patients specific

1.6. MRI

1.6.1. low resolution

2. aim: help take precautions

2.1. trigger a alarm for carers

2.1.1. deliver pharmacological or no-pharmacological control

3. object: EEG, ECG data

3.1. advanced signal processing

3.1.1. identify predictive characteristics

3.2. machine learning

4. development fields

4.1. more continuous database

4.1.1. scalp, implantable, and subdermal selection

4.1.2. electrode placement

4.2. mechanism

4.2.1. underpinning in pre-ictal period

4.2.2. networking theory

4.3. model

4.3.1. mathematics of dynamical system/ biostability

4.3.1.1. physiological parameters

4.3.1.2. endogenous and exogenous parameters

4.4. device, and clinical trials on device

4.5. robust predictor

4.6. algorithm with good performance

4.7. other science field

4.7.1. multiscale electrophysiology (microelectrodes)

4.7.1.1. challenges: number of neurons

4.7.1.2. spatial coverage

4.7.1.3. processing

4.7.1.4. relation between EEG and neuron

4.7.2. multimodal biosensing

4.7.2.1. neurochemistry/Glutamine

4.7.2.2. cardiac, heart rate

4.7.2.3. cerebral haemodynamics

4.7.2.4. vascular coupling

4.7.2.5. blood-oxygen-level