Seizure Pathways: A Model-Based Investigation: A1111111111 A1111111111 A1111111111 A1111111111 A1111111111

RESEARCH ARTICLE
Seizure pathways: A model-based

investigation
Philippa J. Karoly ID1,2*, Levin Kuhlmann3,4, Daniel Soudry5, David B. Grayden ID1,4,
Mark J. Cook2, Dean R. Freestone ID2,6
1 Department of Biomedical Engineering, The University of Melbourne, Parkville, Australia, 2 Department of
Medicine, St Vincent’s Hospital, The University of Melbourne, Fitzroy, Australia, 3 Brain Dynamics Lab,
Swinburne University of Technology, Hawthorn, Australia, 4 Centre for Neural Engineering, The University of
Melbourne, Parkville, Australia, 5 Department of Electrical Engineering, Technion, Haifa, Israel, 6 Seer
Medical Pty, Melbourne, Australia
a1111111111
a1111111111 ¤ Current address: Department of Medicine, St Vincent’s Hospital, The University of Melbourne, Melbourne,
a1111111111 Australia
a1111111111 * [email protected]
a1111111111
Abstract
We present the results of a model inversion algorithm for electrocorticography (ECoG) data
OPEN ACCESS
recorded during epileptic seizures. The states and parameters of neural mass models were
Citation: Karoly PJ, Kuhlmann L, Soudry D,
tracked during a total of over 3000 seizures from twelve patients with focal epilepsy. These
Grayden DB, Cook MJ, Freestone DR (2018)
Seizure pathways: A model-based investigation. models provide an estimate of the effective connectivity within intracortical circuits over the
PLoS Comput Biol 14(10): e1006403. https://doi. time course of seizures. Observing the dynamics of effective connectivity provides insight
org/10.1371/journal.pcbi.1006403 into mechanisms of seizures. Estimation of patients seizure dynamics revealed: 1) a highly
Editor: Daniele Marinazzo, Ghent University, stereotyped pattern of evolution for each patient, 2) distinct sub-groups of onset mecha-
BELGIUM nisms amongst patients, and 3) different offset mechanisms for long and short seizures.
Received: February 14, 2018 Stereotypical dynamics suggest that, once initiated, seizures follow a deterministic path
Accepted: July 26, 2018 through the parameter space of a neural model. Furthermore, distinct sub-populations of
patients were identified based on characteristic motifs in the dynamics at seizure onset.
Published: October 11, 2018
There were also distinct patterns between long and short duration seizures that were related
Copyright: © 2018 Karoly et al. This is an open
to seizure offset. Understanding how these different patterns of seizure evolution arise may
access article distributed under the terms of the
Creative Commons Attribution License, which provide new insights into brain function and guide treatment for epilepsy, since specific ther-
permits unrestricted use, distribution, and apies may have preferential effects on the various parameters that could potentially be indi-
reproduction in any medium, provided the original vidualized. Methods that unite computational models with data provide a powerful means to
author and source are credited.
generate testable hypotheses for further experimental research. This work provides a dem-
Data Availability Statement: Data used for this onstration that the hidden connectivity parameters of a neural mass model can be dynami-
study contain confidential participant information.
The data are available under a Creative Commons
cally inferred from data. Our results underscore the power of theoretical models to inform
licence with restrictions on ATTRIBUTION, NON- epilepsy management. It is our hope that this work guides further efforts to apply computa-
COMMERICIAL use, (DOI:10.26188/ tional models to clinical data.
5b6a999fa2316<https://doi.org/10.26188/
5b6a999fa2316>). Data access should be
requested from https://www.epilepsyecosystem.
org/howitworks/#data ("Melbourne Seizure
Prediction Trial Seizure Data"). Users will be
required to create an account and sign a terms of
use agreement that requires no commercial use,
and restricts all works derived from the data to be
PLOS Computational Biology | https://doi.org/10.1371/journal.pcbi.1006403 October 11, 2018 1 / 24

Seizure pathways
made publicly available under the same licence.

More information regarding the data download can Author summary
also be found with the code repository https://
github.com/pkaroly/Data-Driven-Estimation. A fundamental question in clinical neuroscience is how and why the brain generates epi-
leptic seizures. To address this problem it is important to unify theoretical models of sei-
Funding: This study was supported by the National
Health and Medical Research Council (Project
zure mechanisms with clinical data. This study investigated a large database of human
Grant 1065638). This research was supported by epileptic seizure recordings. Model inversion was used to track seizure dynamics through
Melbourne Bioinformatics at the University of the lens of a mathematical model for cortical regions. These models can reveal the relative
Melbourne, grant number [VR0003]. PJK was activity and coupling between excitatory, inhibitory and pyramidal neural populations
supported by an Australian Government Research that cannot be directly measured. Measuring cortical dynamics during seizures can pro-
Training Program Scholarship DS was supported
vide insight into epilepsy, and facilitate new treatment strategies. Our analysis of connec-
by the Taub Foundation. The funders had no role in
study design, data collection and analysis, decision tion strengths revealed important aspects of seizure onset and seizure termination. Our
to publish, or preparation of the manuscript. findings have implications for understanding seizure mechanisms and treating epilepsy.
Competing interests: I have read the journal’s
policy and the authors of this manuscript have the
following competing interests: PJK, DRF and MJC
Introduction
are shareholders/employees of Seer Medical Pty
Ltd. Data for this study was obtained from Understanding how and why the brain generates spontaneous seizures is an unsolved problem
NeuroVista Pty Ltd. in neuroscience. The medical implications of seizures are profound, with over 50 million peo-
ple affected by epilepsy, and at least 30% not adequately controlled by available therapies [1].
Surgical treatment does not provide complete seizure freedom for all patients [2], and novel
drugs have not greatly improved on the level of seizure freedom that can be achieved [3]. On
the other hand, data-driven, computational techniques have shown early promise in obtaining
a more individualized picture of a patient’s seizures, which may shed new light on mechanisms
of seizures and lead to targeted treatment strategies [4, 5].
Patient-specific, computational models can provide unique insight into seizure mecha-
nisms, and are well accepted in the study of epilepsy [6]. In particular, lumped parameter neu-
ral mass models [7, 8] have been extensively used to investigate cortical activity during
epileptic seizures [9–11]. These models describe seizures as state transitions in the brain [12]
that arise from endogenous noise perturbations or ‘pathways through the parameter space’ of
a neural model [13]. Clinically, it is recognized that electrographic (EEG) recordings of sei-
zures show stereotypical changes in the signal morphology that are regarded as state changes
(i.e. between interictal, peri-ictal, and ictal states) [14]. Despite the ubiquity of neural mass
models to study seizure transitions, the translation of these theoretical insights into clinical
practice has not been widely realized.
The validation of neural network models to aid clinical decision making has made some
advances in diagnosis [15], and surgical planning [16–19]. Ideally model-based techniques can
also improve outcomes at earlier stage interventions, such as drug selection. Another area
models can aid treatment may be in seizure forecasting [20, 21], or the design of electrical
counter-stimulation (using model predictive control) [22–24]. A fundamental hurdle to over-
come is validating theoretical models of brain dynamics in a clinical setting. This hurdle largely
exists due to the difficulty of obtaining in vivo neural recordings from humans. Whilst simula-
tion has proven valuable to generate new hypotheses regarding the mechanisms of seizures, a
complete validation must unite empirical data with theoretical models and demonstrate that
models have predictive value, as well as being descriptive of the data [4, 5].
Model inversion is a powerful approach to combine patient-specific recordings with
accepted principles of brain structure and function encapsulated by the parameters of a neural
model [25]. Previous work has outlined a generalizable framework to estimate the most likely
states and parameters of a neural model given observed data [25]. For many years this problem
was intractable for non-linear neural models [26]. Previously, model inversion has relied upon

Seizure pathways
simplifying assumptions, such as linearization, or sampling techniques [27]. Another approach

is to re-frame the problem, so that the objective is to find the most likely model to generate the
observed data. In this way, the estimation is conditioned on the model space, which is gener-
ally explored via some heuristic model selection criteria [28, 29]. Alternatively, the inversion
can be conditioned on the data, where the most likely model is identified using an assumed
density (Kalman) filter [30]. This approach has been validated for investigation of seizure
dynamics [31–33]. Recent advances have also incorporated a fast, semi-analytic solution to
handle the propagation of estimates through the non-linear neural mass equations [25, 34, 35].
Model inversion techniques that enable time-varying estimates of key parameters provide a
powerful means of inferring cortical mechanisms from functional neuroimaging data. This is
particularly true for EEG/ECoG data, which has high temporal resolution. The ability to
update estimates with each new data point can lead to insights into ictal dynamics that evolve
over fast time scales.
Statistical observations from data are also important to validate models of seizure transi-
tions. Some studies have investigated the distributions of times spent in different seizure states
[36]. Models that are predictive of higher-order statistics derived from many seizures are more
convincing than models which are only descriptive of (or fitted to) individual seizures [4]. An
important observation is that the distribution of times spent in the ictal state has a patient-spe-
cific peak [37], rather than a uniform distribution. These peaks indicate that patients have a
characteristic seizure duration, or trajectory length. Intriguingly, a subset of patients showed a
distinctly bimodal distribution of seizure durations, indicating two populations of seizures
(long and short) [37]. We hypothesize that these distributions reveal a crucial aspect of seizure
dynamics, which should not be neglected in computational modeling.
Materials and methods

This work presents a large-scale, model-based investigation to address the question of how
multiple (long and short) seizure trajectories arise in the brain. Model inversion was per-
formed for the largest database of human seizures recorded in individual patients [38]. Using
this database, we have previously demonstrated that there are two populations of seizure dura-
tion [37]. The current study investigated different seizure pathways and mechanisms through
the lens of a neural mass model (using the formulation of Jansen and Rit, 1995 [8]). The fol-
lowing sections outline the data, model and estimation techniques. Further detail is provided
in S1 Appendix, and code is available online (https://github.com/pkaroly/Data-Driven-
Estimation).
Seizure data
Seizure mechanisms were investigated for continuously recorded ECoG from 12 patients with
focal epilepsy monitored during a previous clinical trial [38]. All subjects were implanted with
intracranial electrode arrays with a total of 16 platinum iridium contacts around the seizure
onset zone. The ECoG was sampled at 400 Hz and wirelessly relayed to an external, portable
personal advisory device. Seizure detection was automated and reviewed by expert clinicians.
This study used data from 3010 clinical seizures (average 250 per patient). Seizures were either
associated with confirmed clinical symptoms or were electrographically similar to clinical sei-
zures. Other epileptiform discharges without clinical symptoms were excluded. All seizures
had onset and offset labelled by expert epileptologists. For further details on the data collection
procedures the reader is referred to Cook et al. (2013) [38].
A similar procedure to that outlined by Cook et al. (2016) [37] was used to identify patients
with bimodal seizure durations. Both k-means clustering and Gaussian mixture model fitting

Seizure pathways
were used to test for bimodality. Clusters were assigned for one, two and three seizure popula-
tions (based on the logarithm of seizure duration in seconds). The optimal number of clusters
was determined using gap criteria [39].
The current study used patients who had at least 20 seizures that had a lead time of one
hour. Recordings were used from five minutes before seizure onset, until one minute after sei-
zure offset. Seizures with telemetry dropouts were excluded from analysis. Data were bandpass
filtered (second-order, zero-phase Butterworth filter) from 1 Hz to 180 Hz with a notch filter
at 50 Hz (second-order, zero-phase Butterworth filter). The energy of the signal was computed
PN
for a 1s sliding window (50% overlap) as energy ¼ n¼1 x2 .
Neural model
The states (mean membrane potentials) and parameters (synaptic connectivity strength) of
neural mass models were fitted to data recorded during epileptic seizures. The formulation of
the neural mass model in the following section is derived from the model introduced by Jansen
and Rit (1995) [8]), and has also been outlined in our previous work [33, 34]. The neural mass
model is suitable to model ECoG measured at this scale (electrodes approximately 5mm in
diameter with spacing on the order of centimeters), in line with similar neural models used to
describe EEG/MEG activity [10, 11, 40]. A single, independent neural model was fitted to each
ECoG channel (16 models in total). Neural models were not coupled between channels; hence,
estimates primarily captured local connection strengths within a single cortical region. The
input parameter, u described non-local inputs to the pyramidal population.
The Jansen and Rit model consists of three neural populations (excitatory, inhibitory and
pyramidal). Neural populations were described by their time varying mean membrane poten-
tial, vn, which is the sum of contributing mean post-synaptic potentials, vmn (post-synaptic and
pre-synaptic neural populations are indexed by n and m, respectively). For the current model,
the index n (post-synaptic) represents either pyramidal (p), excitatory (e) or inhibitory (i) pop-
ulations, as shown in Fig 1.
The post-synaptic potential, vmn arises from the convolution of the input firing rate, f(vn),
with the post-synaptic response kernel,
Z t
vmn ðtÞ ¼ amn hmn ðt t0 Þ�ðvn ðt 0 ÞÞ dt0 ; ð1Þ
1
where αmn, which are the estimation parameters, represent lumped connectivities that incor-
porate average synaptic gain, number of connections, and average maximum firing rate of the
presynaptic populations. ϕ(vn) is the sigmoid function
� � � �
1 v v0
� ðv Þ ¼ erf þ1 ð2Þ
2 B
where v0 = 6mV, and B = 0.0030 (as defined by Freestone et. al. (2014) [25]).
The convolution in Eq 1 can be written as two coupled, first-order, ordinary differential
equations,
dvmn
¼ zmn
dt
ð3Þ
dzmn amn 2 1
¼ � z v :
dt tmn mn tmn mn t2mn mn
where τmn is a lumped time constant. The values of τep, τpe, and τpi were fixed to 10ms and the
value of τip to 20ms, as defined by Jansen & Rit (1995) [8].

Seizure pathways
Fig 1. Data driven estimation using a neural mass model. A. Histograms of seizure durations (log seconds) for the twelve patients
considered in this study, with examples of ECoG during a long and short seizure from a single patient, S8. B. Schematic of the neural
model. Each synapse is defined by a convolution kernel (converting pre-synaptic firing rates to post-synaptic membrane potentials).
Each population is defined by a sigmoid function (converting mean membrane potential to average firing rate). C. Example estimation
time series of the five connectivity parameters during a long and short seizure for one patient. D. Deterministic forward simulation using
the estimated parameters in C (red lines mark actual seizure onset and offset).
https://doi.org/10.1371/journal.pcbi.1006403.g001
External (non-local) inputs to the pyramidal population are modeled as an additive term
affecting the pyramidal membrane potential,
vp ðtÞ ¼ vpe ðtÞ vpi ðtÞ þ uðtÞ: ð4Þ
The recorded ECoG for each channel, i, is derived from the average pyramidal membrane
potential of each independent neural mass model (resulting 16 disconnected models in the

Seizure pathways
estimation),
yi ðtÞ ¼ vip ðtÞ ð5Þ
The neural model can be expressed in matrix notation
xðtÞ
_ ¼ AxðtÞ þ B~
�ðCxðtÞÞ; ð6Þ
where x 2 RNx is a state vector representing the postsynaptic membrane potentials generated
by each population synapse and their time derivatives. There are two states per synapse and Nx
= 2Ns is the total number of states, where, for Ns synaptic connections in the models, the state
vector is of the form,
� �>
x ¼ v1 z 1 . . . vN s z N s :
The definitions of A, B, and C are provided in S1 Appendix.

The observation equation is of the form
yðtÞ ¼ HxðtÞ þ vðtÞ; ð7Þ
where H 2 RNx �Ny is the observation matrix, vðtÞ � N ð0; RÞ 2 RNy is the observation noise,
and Ny is the number of observations (here Ny = 1 as each neural mass model describes a single
ECoG channel). As our measurement function is linear, H is simply an index vector of zeros
and ones that defines the average pyramidal membrane potential given by Eq 4.
State and parameter estimation

A joint state (membrane potentials) and parameter (external input and connectivity strengths)
estimation algorithm was implemented for every sample of the recorded ECoG. To obtain esti-
mates it was necessary to augment the state-space representation of the neural model. To
define the augmented model, we first define a vector of parameters as
� �>
θ ¼ u ape api aip aep :
The dynamics for the parameter are modeled as a random walk
θ_ ¼ 0: ð8Þ
The state vector x and the parameter vector θ are concatenated to form the augmented state
vector,
� �>
ξ ¼ xT θ T : ð9Þ
Our augmented state-space model is

ξt ¼ A y ξ t 1 þ By �ðCy ξt 1 Þ þ wt 1 ; ð10Þ
where wt � N ð0; QÞ. The state vector ξ 2 RNx �1 and matrices Aθ, Bθ, and Cθ are 2 RNx �Nx and
have the form
" # " # " #
A 0 B 0 C 0
Ay ¼ ; By ¼ ; Cy ¼ : ð11Þ
0 I 0 0 0 0
For simplicity we will drop the subscript θ on the system matrices, as the remainder of the
equations refer to the augmented model.

Seizure pathways
The estimation scheme uses an assumed density filter. This filter provides the minimum
mean squared error estimates for the states and parameters, under the assumption that the
underlying probability distribution is Gaussian (the assumed density). Formally stated, the
aim of estimation is to compute the most likely posterior distribution conditioned on previous
measurements,
^ξ þt ¼ E½ξt jy1 ; y2 ; . . . ; yt � ð12Þ
^ þt ¼ E½ðξt
P ^ξ þ Þðξ ^ξ þ Þ> �; ð13Þ
t t t
The estimator proceeds in two stages; prediction and update. In prediction, the prior distri-
bution (obtained from the previous estimate) is propagated though the neural mass equations.
This step provides the so called a priori estimate, which is a Gaussian distribution with mean
and covariance,
^ξ t ¼ E½ξt 1 jy1 ; y2 ; . . . ; yt 1 � ð14Þ
^ t ¼ E½ðξt
P ^ξ þ Þðξ ^ξ þ Þ> �: ð15Þ
1 t 1 t 1 t 1
In the second stage, a Bayesian update is performed to shift the estimated posterior based
on the observed data, giving the a posteriori distribution,
^ξ þt ¼ ξ^t þ Kt ðyt Hξ^t Þ :
|fflfflfflfflfflfflffl {zfflfflfflfflfflfflffl }
ECoG prediction error
þ
^ ¼ ðI
P ^ ;
Kt HÞP ð16Þ
t t
where K is the Kalman gain (readers are referred to [27] for a detailed description of the Kal-
man filter). After each time step, the a posteriori estimate becomes the prior distribution for
the next time step, and the filter proceeds.
In general, the Kalman filter equations do not have a solution for nonlinear model or mea-
surement functions. Previous efforts to use Kalman filtering on the nonlinear neural mass
model have relied on simplifying assumptions (either linearization of the model, or sampling
to estimate the posterior distribution). This work applied an exact, semi-analytic solution for
the mean and covariance of a multivariate Gaussian distribution transformed by the nonlinear
neural mass model. This solution provides the a priori estimate of the mean and covariance
(see S1 Appendix for details).
As the observation function is linear, the updated (a posteriori) mean and covariance are
obtained trivially using Eq 16.
The Kalman filter requires ^ξ þ0 and P
^ þ0 to be initialized to provide the a posteriori state esti-
mate and state estimate covariance for time t = 0. The other parameters that must be initialized
are the model and measurement noise, Q and R, respectively. Further details of filter initializa-
tion are given in the S1 Appendix.
Results
Fig 1A shows an overview of the estimation scheme. The following sections present data from
twelve patients with focal epilepsy. The data consist of over 3000 clinical seizures (average 250
seizures per patient). Of these twelve patients, three showed bimodal distributions of seizure
durations (Patient 3, 8 and 11). Note that all patients showed that either one or two clusters

Seizure pathways
were optimal for seizure durations, and for all patients, k-means and Gaussian mixture model-
ling aligned on the same optimal number of clusters.
Estimated connectivity within neural models describe seizure transitions

An assumed density filter was used to track the time-varying states and parameters of neural
mass models during every seizure (as seen in Fig 1). This estimation technique finds the most
likely model given the observed ECoG data. Importantly, the model is updated at every time
step, so there is no loss of temporal resolution. Estimated states were mean membrane poten-
tials, and parameters (alpha parameters in Fig 1B and 1C), which were the external input and
average synaptic strengths between pyramidal, non-pyramidal excitatory, and inhibitory neu-
ral populations. In this way, the neural models provided an estimate of the average activity and
effective connectivity within intracortical circuits [25]. We found that slow changes in the syn-
aptic connectivity parameters led to seizure transitions in the neural models. As seen in
Fig 1C, a deterministic forward simulation of neural models using time-varying connectivity
estimates reproduced the beginning and end of seizures.
It is important to also quantify estimation accuracy before proceeding. A full summary of
the model and estimation errors is given in the Supplementary Material. However, it is worth
noting briefly that errors between the estimated signal and real data were small (see S1 Fig).
The mean squared error ranged from 0.2-0.9 mV when averaged across all seizures (note that
the mean amplitude of the measured ECoG signal ranges from approximately 25-100mV).
The uncertainty (covariance) of state variables and parameters was also small (see S2 and S3
Figs), suggesting that key seizure activity was well described by the model, rather than by the
residuals. Across patients, the mean covariance ranged from 2-16% for state variables, and
from 0.1-10% for connectivity parameters (expressed as a percentage of the estimated value).
Numerical instability of the filter was occasionally observed (for all patients, instability
occurred for less than 1% of the data). Estimates that became unstable were removed from fur-
ther analysis.
Long and short seizures show distinct signal evolution

Fig 2 shows the average energy of recorded ECoG during every seizure (averaged across 16
electrode channels). Patients’ seizures showed strikingly consistent patterns of signal energy
between seizures. These patterns were generally time locked to seizure onset, as demonstrated
by the vertical alignment of energy changes (note that Patients 2, and 4 did not show a verti-
cally aligned onset pattern). Long and short seizures began similarly, but evolved differently
(see Patients 3, 8, and 11). Long seizures entered a secondary phase where energy increased.
Short seizures and the early phases of long seizures were characterized by an energy reduction
(note the darker vertical band following seizure onset). Some patients’ seizures only showed
the “long” stereotypical pattern with a high-energy phase (see Patients 2, 4, 6, 9, and 15).
Patient 13 had only low energy, stereotypical “short” seizures. Patient 7 had a large majority of
short seizures, with a small number evolving to have increased energy.
These two patterns of seizure energy suggest that long and short seizures reflect distinct
event types, each with a characteristic electrographic evolution. We hypothesized that these
stereotypical signal patterns represent two alternative seizure trajectories, which could be dif-
ferentiated by their onset and/or offset mechanisms. Note that although the average energy
(averaged across electrode channels) was presented, the observed patterns were consistent
across different channels. Full plots are provided in the Supplementary Material (S8 to
S19 Figs).

Seizure pathways
Fig 2. Seizure energy. Average signal energy during evolution of seizures, sorted by duration, from 10s before seizure onset (marked by arrowhead) to
10s after seizure termination (according to clinicians’ marking). Red lines mark the divisions between long and short populations of seizures within
bimodal subjects. Energy for individual electrode channels is provided in the supplementary material.
Seizure trajectories
Fig 3 presents the dynamic estimation results for the five connectivity parameters of a neural
model during every seizure. Seizures followed a remarkably consistent trajectory through the
parameter space of the neural mass models, showing similar patterns across all events for an
individual. This indicates that seizure transitions follow a stereotypical pathway. Note that
transitions during the seizure are locked to the onset time (as demonstrated by the vertical
banding in parameter changes). A higher contrast (normalized) version of these patterns is
provided in S4 Fig, to more clearly expose connectivity patterns.
For all patients, the strongest ictal changes in connectivity strength occurred for in-going
connections to the pyramidal neurons (the first three columns of Fig 3). Conversely, outgoing
pyramidal connections (to inhibitory and excitatory neurons) were more stable over the dura-
tions of the seizures, demonstrated by values which were closer to zero (reflecting no change
from baseline), and less vertical patterning (reflecting no stereotypical transitions during sei-
zures). Patient 6 was one possible exception, showing some decrease in outgoing pyramidal
connections (6D and 6E).
Note that although neural models were fitted independently to all 16 electrode channels,
Fig 3 shows results for a single example channel per patient. The data associated with every
channel generates 60 full page figures, which are provided in an online repository (https://
github.com/pkaroly/Data-Driven-Estimation). The consistency of stereotypical patterns
between channels is investigated in the following sections.

Seizure pathways
Fig 3. Estimated changes in parameter trajectories during every seizure. Each subpanel represents the connectivity strength for all seizures (sorted by
duration) of a single representative channel for each patient (P1: Ch3, P2: Ch1, P3: Ch2, P4: Ch2, P6: 7, P7: Ch15, P8: Ch11, P9: Ch4, P10: Ch13, P11:
Ch9, P13: Ch5, P15: Ch1). Data for all channels is provided online (https://github.com/pkaroly/Data-Driven-Estimation/tree/master/figures/
connectivity). Parameters are expressed as a percentage change from their pre-ictal background values (where zero reflects no change from the pre-ictal
period). The pre-ictal period was defined from two minutes to one minute before seizure onset. Red lines mark the divisions between long and short
populations of seizures within bimodal subjects.
Fig 4 shows the mean change in connectivity strength during seizures. A consistent motif
was a decrease followed by an increase in ingoing connections to the pyramidal population
(see columns A and C for Patients 1, 3, 6, 8, 9, and 15). Patient 11 showed the same motif, but
with connectivity strength always above baseline. Patients 7, 10, and 13 showed only decreases
in ingoing pyramidal connections. Patients 2 and 4 demonstrated only strengthened
connections.
There were three classes of ictal parameter transitions: decrease, increase, and decrease-
then-increase, where connections into the pyramidal populations were on average weaker,
stronger, or weakened then strengthened (compared to a pre-ictal baseline), respectively.
These classes aligned well with the stereotypical seizure evolution patterns that were identified
based on signal energy (Fig 2). For instance, “decrease-then-increase” patients (1, 3, 6, 8, 9,

Seizure pathways
Fig 4. Average changes in parameter trajectories. Each subpanel represents the mean connectivity strength (averaged across seizures)
for each patient (grey shading represents the 95% confidence bounds of the mean). For each patient a single representative channel was
chosen (P1: Ch3, P2: Ch1, P3: Ch2, P4: Ch2, P6: 7, P7: Ch15, P8: Ch11, P9: Ch4, P10: Ch13, P11: Ch9, P13: Ch5, P15: Ch1). Data for all
channels is provided online (https://github.com/pkaroly/Data-Driven-Estimation/tree/master/figures/connectivity). Parameters are
expressed as a percentage change from their pre-ictal background values (where zero reflects no change from the pre-ictal period). The
pre-ictal period was defined from two minutes to one minute before seizure onset. Significant (p < 0.05) changes are marked in blue
(decrease from baseline) and red (increase from baseline) at the top of each plot. Bimodal patients are shown with red labels. Columns
from A-E show the connectivity parameters: A) external input to pyramidal neurons, B) inhibitory to pyramidal connectivity, C)
excitatory to pyramidal connectivity, D) pyramidal to excitatory connectiviy, E) pyramidal to inhibitory connectiviy.

Seizure pathways
and 15) showed long seizures that began with lower energy and evolved into a higher energy
state. The “increase” patients (2 and 4) showed primarily high energy seizures without an obvi-
ous alignment to seizure onset. The “decrease” patients (7, 10, and 13) showed short, low-
energy seizures.
Outgoing connections from pyramidal cells to excitatory/inhibitory populations showed lit-
tle to no change. For some patients (2, 4, 9 and 10), a slight increase in pyramidal to inhibitory
strength was observed.
Fig 5 shows the average seizure trajectory for all channels. Trajectories were qualitatively
similar across channels. Most subjects showed focal patterns in which a subset of channels
demonstrated connectivity changes during seizures while other channels did not have signifi-
cantly increased or decreased connectivity during seizures (compared to a pre-ictal baseline
period). Such patterns were not surprising, given that all subjects had focal seizures, which typ-
ically appear first on a subset of EEG channels before spreading. Apart from focal connectivity
changes, there was some inter-channel variability at the ends of seizures. Subject 6 showed
some channels with increased connection strengths and others with decreased strengths. Sub-
ject 9 showed increased inhibitory connections across most channels but decreased inhibition
on a subset of channels (Fig 5, subpanel 9B), that occurred toward the ends of seizures. Overall,
significant changes in connectivity (above or below baseline) followed the same stereotypical,
patient-specific pattern across all channels. Exceptions to this consistency were observed for a
few subjects in the later stage of seizures (significant changes are shown in Supplementary
Material S6 Fig). In other words, there were no channels with markedly different trajectories;
changes in connectivity were either in the same direction or showed no significant change
from baseline. This consistency supports the finding of characteristic pathways of epileptic sei-
zures, although these pathways were only observed on a subset of (possibly focal) channels,
while other channels did not show altered connectivity patterns during seizures.
Seizure transitions
Overall, there was no difference in the average connection strength trajectories for long com-
pared with short seizures (when connections were averaged across seizures in the long and
short populations, respectively; see S5 Fig for details). Therefore, we hypothesized that short
and long seizures were primarily differentiated by termination (i.e. both types follow a similar
path from onset, with short seizures terminating earlier). This hypothesis was tested by mea-
suring the correlation between connection strength (now averaged across 16 electrode chan-
nels) and seizure duration before onset and offset (correlation results were qualitatively similar
when evaluated for individual electrode channels, and are provided in S7 Fig).
Fig 6 shows that almost no patients showed significant correlation between seizure duration
and onset dynamics. In other words, there was no relationship between average connection
strength and seizure duration at the outset (measured over a 5s window prior to seizure onset).
However, at 5s before seizure offset, there were strong correlations for all connections. In gen-
eral, longer seizures were associated with increased excitatory inputs and decreased inhibition
to the pyramidal cells. Bimodal patients (3, 8, 10, and 11) all showed a similar relationship
between connectivity strength and seizure duration. Four other patients also showed signifi-
cant correlations; therefore, correlations do not arise purely because of the two duration
populations.
Discussion
Analysis of seizure energy (Fig 2) suggests there are two broad categories of focal seizures:
short (low energy) and long (high energy) seizures. Estimation of patients’ seizure trajectories

Seizure pathways
Fig 5. Inter-channel changes in parameter trajectories. Each subpanel represents the mean connectivity strength (averaged across
seizures) for all channels (y-axis) for each patient. Parameters are expressed as a percentage change from their pre-ictal background
values (where zero reflects no change from the pre-ictal period). The pre-ictal period was defined from two minutes to one minute
before seizure onset. Significant changes are shown in the Supplementary Material (S6 Fig). Columns from A-E show the connectivity
parameters: A) external input to pyramidal neurons, B) inhibitory to pyramidal connectivity, C) excitatory to pyramidal connectivity, D)
pyramidal to excitatory connectiviy, E) pyramidal to inhibitory connectiviy.

Seizure pathways
Fig 6. Correlation between connectivity and seizure duration. Correlation was measured between seizure duration
and average connectivity strength (averaged across electrode channels) taken 5s before seizure onset (Panel A), and 5s
before seizure offset (Panel B). Only significant correlation values are shown (p < 0.05). A Bonferroni correction for
multiple comparisons was performed before computing significance, where the 0.05 significance level was divided by
60 (12 patients and 5 parameters).
through the parameter space of a neural model revealed characteristic mechanisms underlying
these different energy states (Fig 3). During the low energy phase of seizures, estimation
showed decreased connectivity strength of ingoing connections to pyramidal cells. Seizures
with high energy showed increased connection strength. This pattern was maintained as sei-
zures evolved through time, with several patients showing a motif of decreased then increased
connection strengths, corresponding to low energy seizures that evolved into a high-energy
state.
Based on their characteristic seizure durations three patient subtypes were defined; those
with exclusively short, exclusively long, or bimodal populations of seizure types. Understand-
ing how these different patterns of seizure evolution arise may provide new insights into brain
function, and guide treatment for epilepsy, as specific therapies may have preferential effects
on the various parameters that could potentially be individualized. This study showed that
long and short seizures reflect different underlying mechanisms in a neural model. Mechanis-
tic differences arose almost exclusively before seizure offset, and were not evident prior to
onset (Fig 6). Therefore, we conclude that seizures follow the same trajectory until termina-
tion. Apart from a bimodal distinction, connectivity patterns were strikingly similar during
the evolution of each patient’s seizures; although highly patient-specific. This suggests that,
once initiated, seizures follow an individualized and deterministic path through the parameter
space of a neural model. It is remarkable to see these parametric pathways maintained across
hundreds of seizures (see Fig 3), and over many years recording duration.
Seizure mechanisms
Patients were classified into three groups of connectivity patterns during seizures (seen in
Fig 4): increased, decreased, and decreased-then-increased strength of ingoing connections to
pyramidal cells. These parameter shifts may relate to distinct mechanisms of seizure onset. For

Seizure pathways
the decrease, and decrease-then-increase patterns we speculate that seizures arise from either
under-regulation or disinhibition of pyramidal neurons. The corresponding rebound of con-
nection strength (in the decrease-then-increase group) may be linked to a regulatory mecha-
nism that was not triggered for patients with shorter seizures (in the decrease group). Previous
work using the neural mass model confirms that inhibitory populations are likely to play a role
in generating epileptiform activity, with the time scale of inhibitory dynamics also highly rele-
vant [10]. There were also two patients who showed only increased connection strength to
pyramidal cells (in Fig 4, Patient 2 showed all connections were increased and Patient 4
showed an increase of excitatory inputs). For these patients, seizures may have been driven by
over-excitation of pyramidal neurons.
There is a lack of consensus as to whether noisy fluctuations (multi-stability) or determin-
istic parameter changes (bifurcations) drive seizure onset/offset [4]. Other mechanisms, such
as intermittency, may also be involved in seizure transitions [41, 42]. This study demonstrated
that the transitions of connectivity parameters were locked to the onset of seizures, and not the
offset (i.e. the patterns in Figs 3 and 2 arise when the seizures are aligned by start time, rather
than end time). This finding suggests that there is a deterministic process conditioned on the
start time of the seizure, whereas the lead up to seizure offset showed more stochasticity. Based
on these results we speculate that seizure onset is more likely to occur through a deterministic
process (as in a bifurcation), where the brain state is driven across some ‘point of no return’.
Offset is more likely to result from noisy fluctuations. Other studies have hypothesized that sei-
zures terminate as the result of a bifurcation [43, 44]. However, the brain’s state during a sei-
zure may merely approach a critical transition, without crossing over [45]. Therefore, it is
possible to observe signs of critical slowing (as in (Kramer et al., 2012)) yet still have seizure
termination driven by noise [4, 46, 47].
The presence of characteristic seizure durations should inform theoretical approaches to
modeling seizure transitions. For instance, in a bistable regime, where noisy fluctuations drive
the transition between a fixed point and oscillatory (‘seizure-like’) state, characteristic dwell
times can emerge for the different states [4, 46, 47]. Dwell times provide one candidate mecha-
nism for characteristic seizure durations. Bimodal populations in some patients suggest that
the brain can support two distinct seizure trajectories (short and long). It has been shown
experimentally that different durations of seizures may arise as the result of distinct onset sti-
muli [48]. Explanations for multiple seizure types can also be derived from computational
models. For instance, different background stability properties in a cortical model can result in
two distinct types of seizures [49]. Multiple seizure trajectories can also arise from different
onset bifurcations [43]. Similarly, multiple offset bifurcations could terminate seizures earlier
or later, giving two populations of duration. The results of this work suggest that long and
short seizures arise from distinct mechanisms of seizure termination. This hypothesis is sup-
ported by a recent study from Payne et al. (2018), which found that long and short seizures
were associated with different durations of post-ictal suppression [50].
Clinical implications
Knowledge of parameter transitions within neural models can increase the information
extracted from EEG, informing new hypotheses of seizure mechanisms and guiding clinical
practice. There is some evidence to suggest that the clinical classification of a seizure is predict-
able soon after its onset [51], in other words, the evolution of a seizure may be somewhat pre-
determined. Our results support the existence of predictable seizure types, and provide
additional metrics (based on the parameters of a neural model) that may extend our under-
standing of traditional seizure types. The consistency of neural model parameters over many

Seizure pathways
seizures suggests that, for some patients, seizure trajectories are established via repetition. The
notion of ‘learned epilepsy’ [52], is an interesting interpretation of epileptogenesis whereby the
abnormal process is learning and spontaneously repeating a pathological sequence, rather than
the sequence itself (all brains can support seizures). For some patients, successful treatment
strategies may involve disrupting or even reversing memorization of the seizure, rather than
addressing an underlying cause [52]. On the other hand, the current results (Fig 3) also showed
that seizure pathways were highly patient-specific and not all subjects’ trajectories were con-
served over time.
Neural mass models have the potential to highlight the relative contributions of excitatory
versus inhibitory connections during seizures. This information can guide whether GABAer-
gic or glutamatergic drugs are required. Previous studies using neural mass models have dem-
onstrated alterations in the balance of excitation and inhibition estimated from data recorded
during seizures [40, 53, 54]. The current estimation technique enables previous efforts to be
extended to investigate a large number of seizures. Some patients showed decreased inhibition
at seizure onset, whereas others demonstrated increased excitation (Fig 4), potentially warrant-
ing different therapies. Furthermore, patients with two duration populations may require dif-
ferent strategies to terminate their seizures. Knowing in advance when two adjunct therapies
are needed is an important clinical insight that can provide crucial benefits to patients with
drug refractory epilepsy. This study found that long seizures were correlated with lower inhibi-
tion and higher excitation (in one patient, the reverse was the case), which can guide electrical
stimulation designed to precipitate early termination of seizures.
The presented model inversion technique and results have wide-ranging applications. The
parameter estimates were consistent across many seizures. Until now, it has not been possible
to show consistency of models of seizure transition in ECoG due to the limited availability of
long-term seizure recordings. Results also generalized across patients. Although the cohort of
12 patients was not large, prior studies have restricted model inversion of seizures to only one
or two patients [13, 55–58]. Another important aspect is that the techniques can be generalized
across models. The estimation filter is not specially formulated for the Jansen and Rit model
used in this study but can be generalized to any model that uses the basic matrix representation
provided in the derivation (see Supplementary Materiall S1 Appendix). That is, the approach
can be applied to any combination of coupled neural populations or indeed any network
model that can be represented by a linear component and a non-linear sigmoidal (error func-
tion) coupling term.
Limitations and future work

Data driven modeling may provide the opportunity to identify which drug could be helpful for
different classes of seizure, as different mechanisms of anti-epileptic drug action may preferen-
tially effect the various connectivity parameters, though further validation of model predic-
tions is needed to translate estimation results to clinical practice. Levels of AEDs have been
related to features of the EEG signal [59, 60]. Therefore, it may be possible to extend this rela-
tionship to predicting the mode of action of an AED from an individual’s EEG. The time scale
of connectivity changes may also be highly relevant to suppressing epileptic activity [10].
Future work should extend estimation to include time constants and investigate the utility of
the outlined neural parameters to detect and predict drug action.
This study provided estimates of independent neural circuits for each channel of ECoG.
Previous studies using coupled neural mass models have highlighted the importance of inter-
channel interactions, particularly for seizure propagation [61]. However, this work considered
local coupling as potentially more relevant to capture the onset of focal seizures. Non-local

Seizure pathways
effects were described by the lumped input parameter, u, rather than explicitly by inter-model
connections. It is possible that long and short seizures could be differentiated earlier based on
inter-channel connectivity patterns. Future work will focus on extending the estimation algo-
rithm for non-locally connected neural regions. An inverse solution to the time-varying,
multi-scale network problem is not trivial and is likely to require additional constraints. For
example, structural MRI data may inform prior probabilities of connection strength [58]. Indi-
vidual neural models can also be coupled within a larger scale network [15]. The approach
taken by Schmidt et. al. (2016) can be adapted to set prior probabilities, or otherwise constrain
the propagation of an assumed density (Kalman) filter.
The challenge of large-scale model inversion is relatively well understood [4, 26]. A more
recent problem in EEG analysis is the challenge of dealing with very high dimensional data.
This study involved separate dimensions for model parameters, seizures, patients, electrode
channels, and time. Distilling insights from such a large dataset is computationally intensive.
To provide some insight into this problem, the estimation results presented were 1.5TB in size.
Generation of each figure can take up to a week to complete for all patients. The use of “big
data” techniques for EEG are becoming more relevant to the study of epilepsy [62]. It is impor-
tant that tools for large scale analysis of EEG are made clinically available. The model inversion
technique presented in this work is generalizable and freely available (https://github.com/
pkaroly/Data-Driven-Estimation).
It is important to note that the presented results are only valid insofar as the connectivity
parameters of a neural model capture the relevant dynamics underlying seizure transitions.
The use of neural mass model to investigate seizures has gained wide acceptance among epi-
lepsy researchers [11, 63–65]. Tracking excitatory and inhibitory strengths within a network is
considered highly relevant to understanding and treating seizures [66]. The ability to infer
directional connections (differentiate between in-going and outgoing pyramidal connections)
is also an important feature of model inversion compared with alternative graph inference
measures. The estimation method was previously validated on simulated data [25]. Neverthe-
less, it is highly challenging to quantify the accuracy of the model reconstructions from real
data, where there is no ground truth. The results showed that the difference between recon-
structed and actual ECoG was small (Supplementary S1, S2 and S3 Figs). The consistency of
results across many seizures provides evidence that the estimation can give overarching insight
into mechanisms of patients’ seizures. It is our hope that this study provides a stepping stone
towards a fully validated model inversion framework to guide the clinical management of epi-
lepsy. Future experimental work should investigate whether modulating connectivity strengths
in a stereotypical fashion does lead to different energy and/or duration of seizures, as predicted
by the current analysis.
Conclusion
This work provided a demonstration that the hidden local connectivity parameters of a neural
mass model can be dynamically inferred from ECoG. Our results showed that seizures follow
stereotypical pathways through parameter space. It is apparent that once a seizure has begun, a
predefined sequence of states must be traversed before termination. For a subset of patients,
there were two routes (short and long) to seizure termination. Short and long seizures began
the same way but showed distinct offset mechanisms. Finally, the connectivity patterns at sei-
zure onset showed common motifs across patients. These distinct sub-groups of onset mecha-
nisms may suggest targeted treatment.
Techniques that unify neural mass models with data provide the means to address some of
the unanswered hypotheses pertaining to epileptic dynamics. For example, theoretical studies

Seizure pathways
have hypothesized that seizure trajectories are “innate”, or “repeatable” [13, 52]. The current
results confirm that seizure pathways are indeed patient-specific and highly stereotyped. It has
also been suggested that there are limited classes of onset mechanisms for seizures [43]. The
current results show that there does appear to be a limited number of seizure onset “motifs”
among patients. Finally, our group had previously hypothesized that long and short seizures
reflect distinct cortical mechanisms [37]. The current results demonstrate that long and short
seizures follow the same pathways but have different termination mechanisms. These results
underscore the power of theoretical models to shed light on seizure mechanisms. It is our hope
that these insights guide further modeling studies and may even prove to be directly translat-
able into clinical practice.
Supporting information
S1 Appendix. Neural model estimation.
(PDF)
S1 Fig. Mean squared error between measured and estimated ECoG. Distribution of average
(across 16 channels) mean squared error of the estimated ECoG for all patients. Box plots
show mean (circle), inter-quartile ranges (black square), 5%-95% ranges (black line), and outli-
ers (black dots) over all seizures for that subject. Mean of the error distributions ranged from
0.2 to 0.9 mV (note that the mean amplitude of the measured ECoG signal ranges from
approximately 25—100mV).
(PNG)
S2 Fig. State estimation covariance. Distribution of average (across 16 channels) estimation
error for the post-synaptic potential state variables. Covariance is expressed as a percentage of
the estimate value. Each subplot shows the results for a given subject. Box plots show mean
(circle), inter-quartile ranges (black square), 5%-95% ranges (black line), and outliers (black
dots) over all seizures for that subject. Mean of the covariance distributions ranged from 2% to
16%.
(PNG)
S3 Fig. Parameter estimation covariance. Distribution of average (across 16 channels) esti-
mation error for the synaptic connectivity parameters. Covariance is expressed as a percentage
of the estimate value. Each subplot shows the results for a given subject. Box plots show mean
(circle), inter-quartile ranges (black square), 5%-95% ranges (black line), and outliers (black
dots) over all seizures for that subject. Mean of the covariance distributions ranged from 0.1%
to 10%.
(PNG)
S4 Fig. Estimated changes in parameter trajectories during every seizure. Each subpanel
represents the connectivity strength for all seizures (sorted by duration) from each patient.
Parameters are expressed as a normalised percentage change from their pre-ictal background
values (where 0 reflects no change from the pre-ictal period). Values are normalised, so -1 and
1 represent the minimal and maximal change for each individual parameter (i.e. absolute value
comparisons between parameters cannot be made). The minimal and maximal parameter val-
ues were computed across all seizures for each individual patient. The pre-ictal period was
defined from 2 minutes to 1 minute before seizure onset.
(PNG)
S5 Fig. Average changes in parameter trajectories for long and short seizures. Each sub-
panel represents the mean connectivity strength (averaged across long and short seizures

Seizure pathways
separately) from each patient (shading represents the 95% confidence bounds of the mean).
Parameters are expressed as a percentage change from their pre-ictal background values
(where 0 reflects no change from the pre-ictal period). The pre-ictal period was defined from 2
minutes to 1 minute before seizure onset. Significant (p < 0.05) differences between long and
short trajectories are marked in black above each plot.
(PNG)
S6 Fig. Significant changes in average parameter trajectories for all channels. Each sub-
panel represents the significance of changes in mean connectivity strength (averaged across
seizures, x-axis) for all channels (y-axis) from each patient. Parameters were expressed as a per-
centage change from their pre-ictal background values (shown in Fig 5). Significant (p < 0.05)
increase in connectivity strength is shown in red, and significant decrease is shown in blue.
Columns from A-E show the connectivity parameters: A) external input to pyramidal neurons,
B) inhibitory to pyramidal connectivity, C) excitatory to pyramidal connectivity, D) pyramidal
to excitatory connectiviy, E) pyramidal to inhibitory connectiviy.
(PNG)
S7 Fig. Correlation between connectivity and seizure duration. Correlation was measured
between seizure duration and average connectivity strength taken 5s before seizure onset
(Panel A), and 5s before seizure offset (Panel B). For each patient and parameter one coloured
vertical bar is shown per channel (16 electrodes). Only significant correlation values are shown
(p < 0.05). A Bonferonni correction for multiple comparisons was performed before comput-
ing significance, where the 5% significance level was divided by 60 (12 patients and 5 parame-
ters).
(PNG)
S8 Fig. Seizure energy for Patient 1. Signal energy during evolution of seizures, sorted by
duration, from 10s before seizure onset (marked by arrowhead) to 10s after seizure termina-
tion (according to clinicians’ marking). Energy was computed for a 1s sliding window (50%
overlap).
(TIF)
overlap).
(TIF)
overlap).
(TIF)
overlap).
(TIF)
duration, from 10s before seizure onset (marked by arrowhead) to 10s after seizure

Seizure pathways
termination (according to clinicians’ marking). Energy was computed for a 1s sliding window
(50% overlap).
(TIF)
overlap).
(TIF)
overlap).
(TIF)
overlap).
(TIF)
overlap).
(TIF)
overlap).
(TIF)
overlap).
(TIF)
overlap).
(TIF)
Acknowledgments
The authors thank John Terry and Wessel Woldman for their helpful insights and review of
the manuscript.

Seizure pathways
Author Contributions
Conceptualization: Philippa J. Karoly, David B. Grayden, Mark J. Cook, Dean R. Freestone.
Data curation: Philippa J. Karoly, Mark J. Cook, Dean R. Freestone.
Formal analysis: Philippa J. Karoly, Daniel Soudry, Dean R. Freestone.
Funding acquisition: David B. Grayden, Mark J. Cook, Dean R. Freestone.
Investigation: Philippa J. Karoly, Dean R. Freestone.
Methodology: Philippa J. Karoly, Levin Kuhlmann, Daniel Soudry, David B. Grayden, Dean
R. Freestone.
Project administration: Philippa J. Karoly, David B. Grayden, Mark J. Cook, Dean R.
Freestone.
Resources: David B. Grayden, Mark J. Cook, Dean R. Freestone.
Software: Philippa J. Karoly, Dean R. Freestone.
Supervision: David B. Grayden, Mark J. Cook.
Validation: Philippa J. Karoly, Dean R. Freestone.
Visualization: Philippa J. Karoly, David B. Grayden, Dean R. Freestone.
Writing – original draft: Philippa J. Karoly, Dean R. Freestone.
Writing – review & editing: Philippa J. Karoly, Levin Kuhlmann, David B. Grayden, Mark J.
Cook, Dean R. Freestone.
References
1. Kwan Patrick and Brodie Martin J. Early identification of refractory epilepsy. New England Journal of
Medicine, 342(5):314–319, 2000. https://doi.org/10.1056/NEJM200002033420503 PMID: 10660394
2. Téllez-Zenteno José F, Dhar Raj, and Wiebe Samuel. Long-term seizure outcomes following epilepsy
surgery: a systematic review and meta-analysis. Brain, 128(5):1188–1198, 2005. https://doi.org/10.
1093/brain/awh449 PMID: 15758038
3. Chen Zhibin, Brodie Martin J, Liew Danny, and Kwan Patrick. Treatment outcomes in patients with
newly diagnosed epilepsy treated with established and new antiepileptic drugs: a 30-year longitudinal
cohort study. JAMA neurology, 75(3):279–286, 2018. https://doi.org/10.1001/jamaneurol.2017.3949
PMID: 29279892
4. Breakspear Michael. Dynamic models of large-scale brain activity. Nature neuroscience, 20(3):
340–352, 2017. https://doi.org/10.1038/nn.4497 PMID: 28230845
5. Woldman Wessel and Terry John R. Multilevel computational modelling in epilepsy: classical studies
and recent advances. In Validating Neuro-Computational Models of Neurological and Psychiatric Disor-
ders, pages 161–188. Springer, 2015.
6. Prinz Astrid A. Understanding epilepsy through network modeling. Proceedings of the National Acad-
emy of Sciences, 105(16):5953–5954, 2008. https://doi.org/10.1073/pnas.0802299105
7. Wilson Hugh R and Cowan Jack D. Excitatory and inhibitory interactions in localized populations of
model neurons. Biophysical journal, 12(1):1–24, 1972. https://doi.org/10.1016/S0006-3495(72)86068-
5 PMID: 4332108
8. Jansen Ben H and Rit Vincent G. Electroencephalogram and visual evoked potential generation in a
mathematical model of coupled cortical columns. Biological cybernetics, 73(4):357–366, 1995. https://
doi.org/10.1007/BF00199471 PMID: 7578475
9. Coombes Stephen and Terry John R. The dynamics of neurological disease: integrating computational,
experimental and clinical neuroscience. European Journal of Neuroscience, 36(2):2118–2120, 2012.
https://doi.org/10.1111/j.1460-9568.2012.08185.x PMID: 22805057
10. F Wendling, F Bartolomei, JJ Bellanger, and P Chauvel. Epileptic fast activity can be explained by a
model of impaired gabaergic dendritic inhibition. European Journal of Neuroscience, 15(9):1499–1508,
2002. https://doi.org/10.1046/j.1460-9568.2002.01985.x

Seizure pathways
11. David Olivier and Friston Karl J. A neural mass model for MEG/EEG:: coupling and neuronal dynamics.
NeuroImage, 20(3):1743–1755, 2003. https://doi.org/10.1016/j.neuroimage.2003.07.015 PMID:
14642484
12. Da Silva Fernando Lopes, Blanes Wouter, Kalitzin Stiliyan N., Parra Jaime, Suffczynski Piotr, and Velis
Demetrios N. Epilepsies as dynamical diseases of brain systems: basic models of the transition
between normal and epileptic activity. Epilepsia, 44(s12):72–83, 2003. https://doi.org/10.1111/j.0013-
9580.2003.12005.x
13. Nevado-Holgado Alejo J., Marten Frank, Richardson Mark P., and Terry John R. Characterising the
dynamics of EEG waveforms as the path through parameter space of a neural mass model: application
to epilepsy seizure evolution. Neuroimage, 59(3):2374–2392, 2012. https://doi.org/10.1016/j.
neuroimage.2011.08.111 PMID: 21945471
14. Badawy R. a. B., D. R. Freestone, Lai A., and Cook M. J. Epilepsy: Ever-changing states of cortical
excitability. Neuroscience, 222:89–99, October 2012. https://doi.org/10.1016/j.neuroscience.2012.07.
015 PMID: 22813999
15. Schmidt Helmut, Woldman Wessel, Goodfellow Marc, Chowdhury Fahmida A, Koutroumanidis Micha-
lis, Jewell Sharon, Richardson Mark P, and Terry John R. A computational biomarker of idiopathic gen-
eralized epilepsy from resting state EEG. Epilepsia, 57(10), 2016.
16. Khambhati Ankit N, Davis Kathryn A, Lucas Timothy H, Litt Brian, and Bassett Danielle S. Virtual cortical
resection reveals push-pull network control preceding seizure evolution. Neuron, 91(5):1170–1182,
2016. https://doi.org/10.1016/j.neuron.2016.07.039 PMID: 27568515
17. Goodfellow M, Rummel Christian, Abela Eugenio, Richardson MP, Schindler K, and Terry JR. Estima-
tion of brain network ictogenicity predicts outcome from epilepsy surgery. Scientific reports, 6:29215,
2016. https://doi.org/10.1038/srep29215 PMID: 27384316
18. Proix Timothée, Bartolomei Fabrice, Guye Maxime, and Jirsa Viktor K. Individual brain structure and
modelling predict seizure propagation. Brain, 140(3):641–654, 2017. https://doi.org/10.1093/brain/
awx004 PMID: 28364550
19. Sinha Nishant, Dauwels Justin, Kaiser Marcus, Cash Sydney S, Brandon Westover M, Wang Yujiang,
and Taylor Peter N. Predicting neurosurgical outcomes in focal epilepsy patients using computational
modelling. Brain, 140(2):319–332, 2016. https://doi.org/10.1093/brain/aww299 PMID: 28011454
20. Freestone Dean R, Karoly Philippa J, Peterson Andre DH, Kuhlmann Levin, Lai Alan, Goodarzy Farhad,
and Cook Mark J. Seizure prediction: science fiction or soon to become reality? Current neurology and
neuroscience reports, 15(11):73, 2015. https://doi.org/10.1007/s11910-015-0596-3 PMID: 26404726
21. Kuhlmann Levin, Grayden David B, Wendling Fabrice, and Schiff Steven J. The role of multiple-scale
modelling of epilepsy in seizure forecasting. Journal of clinical neurophysiology: official publication of
the American Electroencephalographic Society, 32(3):220, 2015.
22. Kramer Mark A, Lopour Beth A, Kirsch Heidi E, and Szeri Andrew J. Bifurcation control of a seizing
human cortex. Physical Review E, 73(4):041928, 2006. https://doi.org/10.1103/PhysRevE.73.041928
23. Taylor Peter N, Thomas Jijju, Sinha Nishant, Dauwels Justin, Kaiser Marcus, Thesen Thomas, and
Ruths Justin. Optimal control based seizure abatement using patient derived connectivity. Frontiers in
neuroscience, 9:202, 2015. https://doi.org/10.3389/fnins.2015.00202 PMID: 26089775
24. Taylor Peter Neal, Wang Yujiang, Goodfellow Marc, Dauwels Justin, Moeller Friederike, Stephani
Ulrich, and Baier Gerold. A computational study of stimulus driven epileptic seizure abatement. PLOS
one, 9(12):e114316, 2014. https://doi.org/10.1371/journal.pone.0114316 PMID: 25531883
25. Freestone Dean R., Karoly Philippa J., Nešić Dragan, Aram Parham, Cook Mark J., and Grayden David
B. Estimation of effective connectivity via data-driven neural modeling. Frontiers in neuroscience, 8,
2014. https://doi.org/10.3389/fnins.2014.00383 PMID: 25506315
26. Schiff Steven J. Neural control engineering: the emerging intersection between control theory and neu-
roscience. MIT Press, 2012.
27. Simon Dan. Optimal state estimation Kalman H infinity and nonlinear approaches. John Wiley & Sons,
2006.
28. Daunizeau Jean, David Olivier, and Stephan Klaas E. Dynamic causal modelling: a critical review of the
biophysical and statistical foundations. Neuroimage, 58(2):312–322, 2011. https://doi.org/10.1016/j.
neuroimage.2009.11.062 PMID: 19961941
29. Friston Karl J., Harrison Lee, and Penny Will. Dynamic causal modelling. Neuroimage, 19(4):
1273–1302, 2003. https://doi.org/10.1016/S1053-8119(03)00202-7 PMID: 12948688
30. Kalman Rudolph Emilet al. A new approach to linear filtering and prediction problems. Journal of basic
Engineering, 82(1):35–45, 1960. https://doi.org/10.1115/1.3662552

Seizure pathways
31. Freestone Dean R, Layton Kelvin J, Kuhlmann Levin, and Cook Mark J. Statistical performance analysis
of data-driven neural models. International journal of neural systems, 27(01):1650045, 2017. https://
doi.org/10.1142/S0129065716500453 PMID: 27776437
32. Aram Parham, Freestone Dean R, Cook Mark J, Kadirkamanathan Visakan, and Grayden David B.
Model-based estimation of intra-cortical connectivity using electrophysiological data. NeuroImage,
118:563–575, 2015. https://doi.org/10.1016/j.neuroimage.2015.06.048 PMID: 26116963
33. Freestone Dean R, Aram Parham, Dewar Michael, Scerri Kenneth, Grayden David B, and Kadirkama-
nathan Visakan. A data-driven framework for neural field modeling. NeuroImage, 56(3):1043–1058,
2011. https://doi.org/10.1016/j.neuroimage.2011.02.027 PMID: 21329758
34. Kuhlmann Levin, Freestone Dean R, Manton Jonathan H, Heyse Bjorn, Vereecke Hugo EM, Lipping
Tarmo, Struys Michel MRF, and Liley David TJ. Neural mass model-based tracking of anesthetic brain
states. NeuroImage, 133:438–456, 2016. https://doi.org/10.1016/j.neuroimage.2016.03.039 PMID:
27018048
35. Ahmadizadeh Saeed, Karoly Philippa J, Grayden David B, Nešić Dragan, and Freestone Dean R. Bifur-
cation analysis of two coupled jansen-rit neural mass models. PLoS ONE, 2018. https://doi.org/10.
1371/journal.pone.0192842 PMID: 29584728
36. Suffczynski Piotr, Lopes Da Silva Fernando H., Parra Jaime, Velis Demetrios N., Bouwman Brigitte M.,
Van Rijn Clementina M., Van Hese Peter, Boon Paul, Khosravani Houman, and Derchansky Miron.
Dynamics of epileptic phenomena determined from statistics of ictal transitions. IEEE Transactions on
Biomedical Engineering, 53(3):524–532, 2006. https://doi.org/10.1109/TBME.2005.869800 PMID:
16532779
37. Cook Mark J., Karoly Philippa J., Freestone Dean R., Himes David, Leyde Kent, Berkovic Samuel,
O’brien Terence, Grayden David B., and Boston Ray. Human focal seizures are characterized by popu-
lations of fixed duration and interval. Epilepsia, 57(3):359–368, 2016. https://doi.org/10.1111/epi.13291
PMID: 26717880
38. Cook Mark J., O’Brien Terence J., Berkovic Samuel F., Murphy Michael, Morokoff Andrew, Fabinyi
Gavin, Wendyl D’Souza, Yerra Raju, Archer John, and Litewka Lucas. Prediction of seizure likelihood
with a long-term, implanted seizure advisory system in patients with drug-resistant epilepsy: a first-in-
man study. The Lancet Neurology, 12(6):563–571, 2013. https://doi.org/10.1016/S1474-4422(13)
70075-9 PMID: 23642342
39. Tibshirani Robert, Walther Guenther, and Hastie Trevor. Estimating the number of clusters in a data set
via the gap statistic. Journal of the Royal Statistical Society: Series B (Statistical Methodology), 63(2):
411–423, 2001. https://doi.org/10.1111/1467-9868.00293
40. Aarabi Ardalan and He Bin. Seizure prediction in hippocampal and neocortical epilepsy using a model-
based approach. Clinical Neurophysiology, 125(5):930–940, 2014. https://doi.org/10.1016/j.clinph.
2013.10.051 PMID: 24374087
41. Goodfellow Marc, Schindler Kaspar, and Baier Gerold. Intermittent spike–wave dynamics in a heteroge-
neous, spatially extended neural mass model. Neuroimage, 55(3):920–932, 2011. https://doi.org/10.
1016/j.neuroimage.2010.12.074 PMID: 21195779
42. Goodfellow Marc and Glendinning Paul. Mechanisms of intermittent state transitions in a coupled het-
erogeneous oscillator model of epilepsy. The Journal of Mathematical Neuroscience, 3(1):17, 2013.
https://doi.org/10.1186/2190-8567-3-17 PMID: 23945016
43. Jirsa Viktor K., Stacey William C., Quilichini Pascale P., Ivanov Anton I., and Bernard Christophe. On
the nature of seizure dynamics. Brain, 137(8):2210–2230, 2014. https://doi.org/10.1093/brain/awu133
PMID: 24919973
44. Kramer Mark A., Truccolo Wilson, Eden Uri T., Lepage Kyle Q., Hochberg Leigh R., Eskandar Emad N.,
Madsen Joseph R., Lee Jong W., Maheshwari Atul, and Halgren Eric. Human seizures self-terminate
across spatial scales via a critical transition. Proceedings of the National Academy of Sciences,
109(51):21116–21121, 2012. https://doi.org/10.1073/pnas.1210047110
45. Scheffer Marten, Bascompte Jordi, Brock William A, Brovkin Victor, Carpenter Stephen R, Dakos Vasi-
lis, Held Hermann, Van Nes Egbert H, Rietkerk Max, and Sugihara George. Early-warning signals for
critical transitions. Nature, 461(7260):53–59, 2009. https://doi.org/10.1038/nature08227 PMID:
19727193
46. Benjamin Oscar, Fitzgerald Thomas HB, Ashwin Peter, Krasimira Tsaneva-Atanasova, Chowdhury
Fahmida, Richardson Mark P., and Terry John R. A phenomenological model of seizure initiation sug-
gests network structure may explain seizure frequency in idiopathic generalised epilepsy. The Journal
of Mathematical Neuroscience, 2(1):1, 2012. https://doi.org/10.1186/2190-8567-2-1 PMID: 22657571
47. P. Suffczynski, S. Kalitzin, and FH Lopes Da Silva. Dynamics of non-convulsive epileptic phenomena
modeled by a bistable neuronal network. Neuroscience, 126(2):467–484, 2004. https://doi.org/10.
1016/j.neuroscience.2004.03.014

Seizure pathways
48. Nelson Timothy S, Suhr Courtney L, Lai Alan, Halliday Amy J, Freestone Dean R, McLean Karen J, Bur-
kitt Anthony N, and Cook Mark J. Seizure severity and duration in the cortical stimulation model of
experimental epilepsy in rats: A longitudinal study. Epilepsy research, 89(2):261–270, 2010. https://doi.
org/10.1016/j.eplepsyres.2010.01.010 PMID: 20153951
49. Wang Yujiang, Trevelyan Andrew J., Valentin Antonio, Alarcon Gonzalo, Taylor Peter N., and Kaiser
Marcus. Mechanisms underlying different onset patterns of focal seizures. PLoS computational biology,
13(5):e1005475, 2017. https://doi.org/10.1371/journal.pcbi.1005475 PMID: 28472032
50. Payne Daniel E, Karoly Philippa J, Freestone Dean R, Boston Ray, D’souza Wendyl, Nurse Ewan, Kuhl-
mann Levin, Cook Mark J, and Grayden David B. Postictal suppression and seizure durations: A
patient-specific, long-term ieeg analysis. Epilepsia, 59(5):1027–1036, 2018. https://doi.org/10.1111/
epi.14065 PMID: 29633239
51. Donos Cristian, Maliia Mihai Dragos, Matthias Dümpelmann, and Andreas Schulze-Bonhage. Seizure
onset predicts its type. Epilepsia, 2018. https://doi.org/10.1111/epi.13997
52. Hsu David, Chen Wei, Hsu Murielle, and Beggs John M. An open hypothesis: is epilepsy learned, and
can it be unlearned? Epilepsy & Behavior, 13(3):511–522, 2008. https://doi.org/10.1016/j.yebeh.2008.
05.007
53. Fan Xiaoya, Gaspard Nicolas, Legros Benjamin, Lucchetti Federico, Ercek Rudy, and Nonclercq
Antoine. Dynamics underlying interictal to ictal transition in temporal lobe epilepsy: insights from a neu-
ral mass model. European Journal of Neuroscience, 2017.
54. Blenkinsop Alex, Valentin Antonio, Richardson Mark P, and Terry John R. The dynamic evolution of
focal-onset epilepsies–combining theoretical and clinical observations. European Journal of Neurosci-
ence, 36(2):2188–2200, 2012. https://doi.org/10.1111/j.1460-9568.2012.08082.x PMID: 22805064
55. Cooray Gerald K, Sengupta Biswa, Douglas Pamela K, and Friston Karl. Dynamic causal modelling of
electrographic seizure activity using bayesian belief updating. NeuroImage, 125:1142–1154, 2016.
https://doi.org/10.1016/j.neuroimage.2015.07.063 PMID: 26220742
56. Murta Teresa, Leal Alberto, Garrido Marta I, and Figueiredo Patrı́cia. Dynamic causal modelling of epi-
leptic seizure propagation pathways: a combined eeg–fmri study. Neuroimage, 62(3):1634–1642,
2012. https://doi.org/10.1016/j.neuroimage.2012.05.053 PMID: 22634857
57. Papadopoulou Margarita, Leite Marco, van Mierlo Pieter, Vonck Kristl, Lemieux Louis, Friston Karl, and
Marinazzo Daniele. Tracking slow modulations in synaptic gain using dynamic causal modelling: valida-
tion in epilepsy. NeuroImage, 107:117–126, 2015. https://doi.org/10.1016/j.neuroimage.2014.12.007
PMID: 25498428
58. Jirsa Viktor K., Proix T., Perdikis Dionysios, Woodman Michael Marmaduke, Wang H., Gonzalez-Marti-
nez J., Bernard C., Bénar Christian, Guye Maxime, and Chauvel Patrick. The virtual epileptic patient:
individualized whole-brain models of epilepsy spread. Neuroimage, 145:377–388, 2017. https://doi.
org/10.1016/j.neuroimage.2016.04.049 PMID: 27477535
59. Meisel Christian, Plenz Dietmar, Schulze-Bonhage Andreas, and Reichmann Heinz. Quantifying antiep-
ileptic drug effects using intrinsic excitability measures. Epilepsia, 57(11), 2016. https://doi.org/10.
1111/epi.13517 PMID: 27562603
60. Meisel Christian, Andreas Schulze-Bonhage, Freestone Dean, Cook Mark James, Achermann Peter,
and Plenz Dietmar. Intrinsic excitability measures track antiepileptic drug action and uncover increas-
ing/decreasing excitability over the wake/sleep cycle. Proceedings of the National Academy of Sci-
ences, 112(47):14694–14699, 2015. https://doi.org/10.1073/pnas.1513716112
61. Kameneva Tatiana, Ying Tianlin, Guo Ben, and Freestone Dean R. Neural mass models as a tool to
investigate neural dynamics during seizures. Journal of computational neuroscience, 42(2):203–215,
2017. https://doi.org/10.1007/s10827-017-0636-x PMID: 28102460
62. Wagenaar Joost B, Worrell Gregory A, Ives Zachary, Matthias Dümpelmann, Litt Brian, and Andreas
Schulze-Bonhage. Collaborating and sharing data in epilepsy research. Journal of clinical neurophysiol-
ogy: official publication of the American Electroencephalographic Society, 32(3):235, 2015. https://doi.
org/10.1097/WNP.0000000000000159
63. Touboul Jonathan, Wendling Fabrice, Chauvel Patrick, and Faugeras Olivier. Neural mass activity,
bifurcations, and epilepsy. Neural computation, 23(12):3232–3286, 2011. https://doi.org/10.1162/
NECO_a_00206 PMID: 21919787
64. Walter J. Freeman. Mass action in the nervous system. Academic Press, New York, 1975.
65. Deco Gustavo, Jirsa Viktor K., Robinson Peter A., Breakspear Michael, and Friston Karl. The dynamic
brain: from spiking neurons to neural masses and cortical fields. PLoS computational biology, 4(8):
e1000092, 2008. https://doi.org/10.1371/journal.pcbi.1000092 PMID: 18769680
66. Kramer Mark A. and Cash Sydney S. Epilepsy as a disorder of cortical network organization. The Neu-
roscientist, 18(4):360–372, 2012. https://doi.org/10.1177/1073858411422754 PMID: 22235060

Seizure Pathways: A Model-Based Investigation: A1111111111 A1111111111 A1111111111 A1111111111 A1111111111

Uploaded by

Document Informationclick to expand document informationjurnal

Document Informationclick to expand document information

Copyright:

Available Formats

Seizure Pathways: A Model-Based Investigation: A1111111111 A1111111111 A1111111111 A1111111111 A1111111111

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Seizure Pathways: A Model-Based Investigation: A1111111111 A1111111111 A1111111111 A1111111111 A1111111111

Uploaded by

Copyright:

Available Formats

RESEARCH ARTICLE

Seizure pathways: A model-based

PLOS Computational Biology | https://doi.org/10.1371/journal.pcbi.1006403 October 11, 2018 1 / 24

made publicly available under the same licence.

PLOS Computational Biology | https://doi.org/10.1371/journal.pcbi.1006403 October 11, 2018 2 / 24

simplifying assumptions, such as linearization, or sampling techniques [27]. Another approach

Materials and methods

PLOS Computational Biology | https://doi.org/10.1371/journal.pcbi.1006403 October 11, 2018 3 / 24

PLOS Computational Biology | https://doi.org/10.1371/journal.pcbi.1006403 October 11, 2018 4 / 24

vp ðtÞ ¼ vpe ðtÞ vpi ðtÞ þ uðtÞ: ð4Þ

PLOS Computational Biology | https://doi.org/10.1371/journal.pcbi.1006403 October 11, 2018 5 / 24

The neural model can be expressed in matrix notation

The definitions of A, B, and C are provided in S1 Appendix.

State and parameter estimation

The dynamics for the parameter are modeled as a random walk

Our augmented state-space model is

PLOS Computational Biology | https://doi.org/10.1371/journal.pcbi.1006403 October 11, 2018 6 / 24

PLOS Computational Biology | https://doi.org/10.1371/journal.pcbi.1006403 October 11, 2018 7 / 24

Estimated connectivity within neural models describe seizure transitions

Long and short seizures show distinct signal evolution

PLOS Computational Biology | https://doi.org/10.1371/journal.pcbi.1006403 October 11, 2018 8 / 24

PLOS Computational Biology | https://doi.org/10.1371/journal.pcbi.1006403 October 11, 2018 9 / 24

PLOS Computational Biology | https://doi.org/10.1371/journal.pcbi.1006403 October 11, 2018 10 / 24

PLOS Computational Biology | https://doi.org/10.1371/journal.pcbi.1006403 October 11, 2018 11 / 24

PLOS Computational Biology | https://doi.org/10.1371/journal.pcbi.1006403 October 11, 2018 12 / 24

PLOS Computational Biology | https://doi.org/10.1371/journal.pcbi.1006403 October 11, 2018 13 / 24

PLOS Computational Biology | https://doi.org/10.1371/journal.pcbi.1006403 October 11, 2018 14 / 24

PLOS Computational Biology | https://doi.org/10.1371/journal.pcbi.1006403 October 11, 2018 15 / 24

Limitations and future work

PLOS Computational Biology | https://doi.org/10.1371/journal.pcbi.1006403 October 11, 2018 16 / 24

PLOS Computational Biology | https://doi.org/10.1371/journal.pcbi.1006403 October 11, 2018 17 / 24

PLOS Computational Biology | https://doi.org/10.1371/journal.pcbi.1006403 October 11, 2018 18 / 24

PLOS Computational Biology | https://doi.org/10.1371/journal.pcbi.1006403 October 11, 2018 19 / 24

PLOS Computational Biology | https://doi.org/10.1371/journal.pcbi.1006403 October 11, 2018 20 / 24

PLOS Computational Biology | https://doi.org/10.1371/journal.pcbi.1006403 October 11, 2018 21 / 24

PLOS Computational Biology | https://doi.org/10.1371/journal.pcbi.1006403 October 11, 2018 22 / 24

PLOS Computational Biology | https://doi.org/10.1371/journal.pcbi.1006403 October 11, 2018 23 / 24

PLOS Computational Biology | https://doi.org/10.1371/journal.pcbi.1006403 October 11, 2018 24 / 24

You might also like