Ensembling crowdsourced seizure prediction algorithms using long‐term human intracranial EEG

Reuben, Chip; Karoly, Philippa J.; Freestone, Dean R.; Temko, Andriy; Barachant, Alexandre; Li, Feng; Titericz, Gilberto; Lang, Brian W.; Lavery, Daniel M.; Roman, Kelly; Broadhead, Derek; Jones, Gareth J. F.; Tang, Qingnan; Ivanenko, Irina; Panichev, Oleg; Proix, Timothée; Náhlík, Michal; Grunberg, Daniel B.; Grayden, David B.; Cook, Mark; Kuhlmann, Levin

doi:10.1111/epi.16418

“…Incorporating such multimodal data into a single classifier is likely to result in an improved predictive AI modelling of epilepsy than a classifier relying on only a single data type, as these data sources contain complementary information pertinent to the disease. Additional data sources, such as EEG ( Hosseini et al , 2020 ; Reuben et al , 2020 ) and clinical documentation of patient characteristics ( Cohen et al , 2016 ), may further enrich the modelling. These data are high-dimensional ( Motsinger and Ritchie, 2006 ), so there is a lot of information that can be hard to interpret and compute with conventional statistical methods ( Friston et al , 1994 ; Benjamini and Hochberg, 1995 ).…”

Section: Increase Ai Model Prediction With Multimodal Datamentioning

confidence: 99%

Artificial intelligence for clinical decision support in neurology

Pedersen

¹

,

Verspoor

²

,

Jenkinson

³

et al. 2020

Brain Communications

View full text Add to dashboard Cite

Abstract Artificial intelligence is one of the most exciting methodological shifts in our era. It holds the potential to transform healthcare as we know it, to a system where humans and machines work together to provide better treatment for our patients. It is now clear that cutting edge artificial intelligence models in conjunction with high-quality clinical data will lead to improved prognostic and diagnostic models in neurological disease, facilitating expert-level clinical decision tools across healthcare settings. Despite the clinical promise of artificial intelligence, machine and deep learning algorithms are not a one-size-fits-all solution for all types of clinical data and questions. In this paper, we provide an overview of the core concepts of artificial intelligence, particularly contemporary deep learning methods, to give clinician and neuroscience researchers an appreciation of how artificial intelligence can be harnessed to support clinical decisions. We clarify and emphasise the data quality and the human expertise needed to build robust clinical artificial intelligence models in neurology. Given that artificial intelligence is a rapidly evolving field, we take the opportunity to iterate important ethical principles to guide the field of medicine is it moves into an artificial intelligence enhanced future.

show abstract

“…The optimistic projections that in the next few years part of the burden of epilepsy in some patients may be eased by seizure forecasting methods are motivated by current evidence of the feasibility of anticipating seizures and recent advances in the field [12]. New methods may leverage insights gained not only from active probing approaches discussed here, but also from increasingly available long-term iEEG databases [95], as well as increased adoption of new implantable devices that enable patient-tailored algorithms. Achieving this would be a crucial step that may enable to ultimately close the loop for targeted seizure-controlling interventions.…”

Section: Resultsmentioning

confidence: 99%

Active probing to highlight approaching transitions to ictal states in coupled neural mass models

Carvalho

¹

,

Moraes

²

,

Cash

³

et al. 2020

Preprint

0

View full text Add to dashboard Cite

The extraction of electrophysiological features that reliably forecast the occurrence of seizures is one of the most challenging goals in epilepsy research. Among possible approaches to tackle this problem is the use of active probing paradigms in which responses to stimuli are used to detect underlying system changes leading up to seizures. This work evaluates the theoretical and mechanistic underpinnings of this strategy using a well-studied seizure-generating neural mass model with two coupled populations. Different simulations are done by shifting parameters (excitability, slow inhibition, and coupling gains) towards ictal states while low-frequency probing stimuli are applied to either one or both populations. The correlation between the extracted features and the shifted parameter indicates if the impending transition to the ictal state may be identified in advance. Results show that not only can the response to the probing stimuli forecast seizures but this is true regardless of the altered ictogenic parameter. That is, similar feature changes are highlighted by probing stimuli responses in advance of the seizure including: increased response variance and lag-1 autocorrelation, decreased skewness, and increased mutual information between the outputs of both model subsets. These changes were restricted to the stimulated population, showing a local effect of this perturbational approach. The transition latencies from normal activity to sustained discharges of spikes were not affected, suggesting that stimuli had no pro-ictal effects. However, stimuli were found to elicit interictal-like spikes just before the transition to the ictal state. Furthermore, the observed feature changes highlighted by probing the neuronal populations may reflect the phenomenon of critical slowing down, where increased recovery times from perturbations may signal the loss of a systems’ resilience and are common hallmarks of an impending critical transition. These results provide more evidence that active probing approaches highlight information about underlying system changes involved in ictogenesis and may be able to play a role in assisting seizure forecasting methods which can be incorporated into early-warning systems that ultimately enable closing the loop for targeted seizure-controlling interventions.Author summaryEpilepsy is characterized by spontaneous and recurrent seizures. Developing a method to forecast the occurrence of these events would ease part of the burden to patients by providing warning systems and identifying critical periods for closed-loop seizure suppressing methods. However, this task is far from trivial and doing this in a clinical setting has been a challenge that is not yet solved. A possible way to achieve this is by using active probing paradigms, in which stimuli are used to highlight otherwise hidden changes in brain dynamics that may lead to seizures. In this work, this approach is tested in a computational model with two coupled neuronal populations. Active probing responses highlighted underlying model parameter changes that led neuronal activity closer to seizures, something that was not detected with the passive observation in most evaluated model settings. Observed feature changes revealed higher excitability, longer recovery times from stimuli, and increased synchrony between populations as the threshold to seizure-like activity approached. This indicates that low frequency stimuli may be used to reveal early-warning signs and assist in seizure forecasting methods.

show abstract

“…The optimistic projections that in the next few years part of the burden of epilepsy in some patients may be eased by seizure forecasting methods are motivated by current evidence of the feasibility of anticipating seizures and recent advances in the field [12]. New methods may leverage insights gained not only from active probing approaches discussed here, but also from increasingly available long-term iEEG databases [97], as well as increased adoption of new implantable devices that enable patient-tailored algorithms. Achieving this would be a crucial step that may enable to ultimately close the loop for targeted seizure-controlling interventions.…”

Section: Discussionmentioning

confidence: 99%

Active probing to highlight approaching transitions to ictal states in coupled neural mass models

Carvalho

¹

,

Moraes

²

,

Cash

³

et al. 2021

View full text Add to dashboard Cite

The extraction of electrophysiological features that reliably forecast the occurrence of seizures is one of the most challenging goals in epilepsy research. Among possible approaches to tackle this problem is the use of active probing paradigms in which responses to stimuli are used to detect underlying system changes leading up to seizures. This work evaluates the theoretical and mechanistic underpinnings of this strategy using two coupled populations of the well-studied Wendling neural mass model. Different model settings are evaluated, shifting parameters (excitability, slow inhibition, or inter-population coupling gains) from normal towards ictal states while probing stimuli are applied every 2 seconds to the input of either one or both populations. The correlation between the extracted features and the ictogenic parameter shifting indicates if the impending transition to the ictal state may be identified in advance. Results show that not only can the response to the probing stimuli forecast seizures but this is true regardless of the altered ictogenic parameter. That is, similar feature changes are highlighted by probing stimuli responses in advance of the seizure including: increased response variance and lag-1 autocorrelation, decreased skewness, and increased mutual information between the outputs of both model subsets. These changes were mostly restricted to the stimulated population, showing a local effect of this perturbational approach. The transition latencies from normal activity to sustained discharges of spikes were not affected, suggesting that stimuli had no pro-ictal effects. However, stimuli were found to elicit interictal-like spikes just before the transition to the ictal state. Furthermore, the observed feature changes highlighted by probing the neuronal populations may reflect the phenomenon of critical slowing down, where increased recovery times from perturbations may signal the loss of a systems’ resilience and are common hallmarks of an impending critical transition. These results provide more evidence that active probing approaches highlight information about underlying system changes involved in ictogenesis and may be able to play a role in assisting seizure forecasting methods which can be incorporated into early-warning systems that ultimately enable closing the loop for targeted seizure-controlling interventions.

show abstract

Ensembling crowdsourced seizure prediction algorithms using long‐term human intracranial EEG

Cited by 20 publications

References 17 publications

Artificial intelligence for clinical decision support in neurology

Artificial intelligence for clinical decision support in neurology

Active probing to highlight approaching transitions to ictal states in coupled neural mass models

Active probing to highlight approaching transitions to ictal states in coupled neural mass models

Contact Info

Product

Resources

About