Author name cluster

David F. Abbott

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

16 papers

1 author row

AIIM Journal 2025 Journal Article

Toward responsible artificial intelligence in medicine: Reflections from the Australian epilepsy project

Mangor Pedersen
Heath R. Pardoe
Anton de Weger
Donna Hutchison
David F. Abbott
Karin Verspoor
Graeme D. Jackson

Artificial intelligence (AI) is a multidisciplinary scientific field that uses machines to solve real-world problems and predict outcomes. Despite the current enthusiasm about AI's potential as a clinical support tool, there is also a growing awareness and concern about the potentially harmful effects of AI. Because AI will likely impact expert-based decision-making in medicine, it is critical to consider the issues that AI raises in medical research. This paper outlines the AI guidelines of the Australian Epilepsy Project. This large-scale platform aims to democratise specialist care in epilepsy and use AI for clinical decision support based on prospective multimodal datasets (MRI, genetic, clinical, and cognitive data) from thousands of people with epilepsy. As AI develops rapidly, we focus on key areas of medical AI identified in the literature, including Trust, Responsibility and Safety. We believe AI is changing medicine, and we believe it is imperative to advance and update our AI guidelines adaptably while preparing for an era of augmented-intelligence-based medicine.

YNICL Journal 2017 Journal Article

The diminishing dominance of the dominant hemisphere: Language fMRI in focal epilepsy

Chris Tailby
David F. Abbott
Graeme D. Jackson

"Which is the dominant hemisphere?" is a question that arises frequently in patients considered for neurosurgery. The concept of the dominant hemisphere implies uniformity of language lateralisation throughout the brain. It is increasingly recognised that this is not the case in the healthy control brain, and it is especially not so in neurological diseases such as epilepsy. In the present work we adapt our published objective lateralisation method (based on the construction of laterality curves) for use with sub-lobar cortical, subcortical and cerebellar regions of interest (ROIs). We apply this method to investigate regional lateralisation of language activation in 12 healthy controls and 18 focal epilepsy patients, using three different block design language fMRI paradigms, each tapping different aspects of language processing. We compared lateralisation within each ROI across tasks, and investigated how the quantity of data collected affected the ability to robustly estimate laterality across ROIs. In controls, lateralisation was stronger, and the variance across individuals smaller, in cortical ROIs, particularly in the Inferior Frontal (Broca) region. Lateralisation within temporal ROIs was dependent on the nature of the language task employed. One of the healthy controls was left lateralised anteriorly and right lateralised posteriorly. Consistent with previous work, departures from normality occurred in ~ 15-50% of focal epilepsy patients across the different ROIs, with atypicality most common in the Lateral Temporal (Wernicke) region. Across tasks and ROIs the absolute magnitude of the laterality estimate increased and its across participant variance decreased as more cycles of task and rest were included, stabilising at ~ 4 cycles (~ 4 min of data collection). Our data highlight the importance of considering language as a complex task where lateralisation varies at the subhemispheric scale. This is especially important for presurgical planning for focal resections where the concept of 'hemispheric dominance' may be misleading. This is a precision medicine approach that enables objective evaluation of language dominance within specific brain regions and can reveal surprising and unexpected anomalies that may be clinically important for individual cases.

YNIMG Journal 2015 Journal Article

Resting state functional connectivity changes induced by prior brain state are not network specific

Chris Tailby
Richard A.J. Masterton
Jenny Y. Huang
Graeme D. Jackson
David F. Abbott

Resting state functional connectivity (rFC) is used to identify functionally related brain areas without requiring subjects to perform specific tasks. Previous work suggests that prior brain state, as determined by the activity engaged in immediately prior to collection of resting state data, can influence the networks recovered by rFC analyses. We determined the prevalence and network specificity of rFC changes induced by manipulations of prior state (including an unstructured (unconstrained) state, and language and motor tasks). Three blocks of rest data (one after each of the specified prior states) were acquired on each of 25 subjects. We hypothesised that prior state induced changes in rFC would be greatest within the networks most actively recruited by that prior state. Changes in rFC were greatest following the motor task and, contrary to our hypothesis, were not network specific. This was demonstrated by comparing (1) the timecourses within a set of ROIs selected on the basis of task-related de/activation, and (2) seed-based whole brain voxel-wise connectivity maps, seeded from local maxima in the task-related de/activation maps. Changes in connectivity strength tended to manifest as increases in rFC relative to that in the unstructured rest state, with change maps resembling partially complete maps of the primary sensory cortices and the cognitive control network. The majority of rFC changes occurred in areas moderately (but not weakly) connected to the seeds. Constrained prior states were associated with lower across-participant variance in rFC. This systematic investigation of the effect of prior brain state on rFC indicates that the rFC changes induced by prior brain state occur both in brain networks related to that brain activity and in networks nominally unrelated to that brain activity.

YNIMG Journal 2013 Journal Article

Mapping brain activity using event-related independent components analysis (eICA): Specific advantages for EEG-fMRI

Richard A.J. Masterton
Graeme D. Jackson
David F. Abbott

Event-related analyses of functional MRI (fMRI) typically assume that the onset and offset of neuronal activity match stimuli onset and offset, and that evoked fMRI signal changes follow the canonical haemodynamic response function (HRF). Some event types, however, may be unsuited to this approach: brief stimuli might elicit an extended neuronal response; anticipatory effects might result in activity preceding the event; or altered neurovascular coupling may result in a non-canonical HRF. An example is interictal epileptiform discharges (IEDs), which may show a non-canonical HRF and fMRI signal changes preceding their onset as detected on EEG. In such cases, less constrained analyses – capable of detecting early, non-canonical responses – may be necessary. A consequence of less constrained analyses, however, is that artefactual sources of signal change – motion or physiological noise for example – may also be detected and mixed with the neuronally-generated signals. In this paper, to address this issue, we describe an event-related independent components analysis (eICA) that identifies different sources of event-related signal change that can then be separately assessed to identify likely artefacts and separate primary from propagated activity. We also describe a group analysis that identifies eICA components that are spatially and temporally consistent across subjects and provides an objective approach for selecting group-specific components likely to be of neural origin. We apply eICA to patients with rolandic epilepsy – with stereotypical IEDs arising from a focus in the rolandic fissure – and demonstrate that a single event-related component, concordant with this source location, is detected.

YNIMG Journal 2012 Journal Article

Selecting appropriate voxel-based methods for neuroimaging studies

David F. Abbott
Gaby S. Pell
Heath R. Pardoe
Graeme D. Jackson

We highlight a fundamental difference between voxel-based methods that interrogate signal intensity directly and those that interrogate morphometric features; we discuss how signal intensity changes might erroneously affect morphometric measures, and we provide some guidance for selection of appropriate methods to address particular hyphotheses. Our discussion is motivated by a recent application of voxel-based morphometry methods to T2-weighted images (T2-Voxel Based Morphometry; T2-VBM). In this context we discuss alternative approaches including Voxel-Based T2-Relaxometry (VBR) and Voxel Based Iterative Sensitivity analysis of T2-Weighted Images (VBIS-T2).

YNIMG Journal 2010 Journal Article

fMRI assessment of language lateralization: An objective approach

David F. Abbott
Anthony B. Waites
Leasha M. Lillywhite
Graeme D. Jackson

Language lateralization based on functional magnetic resonance imaging (fMRI) is often used in clinical neurological settings. Currently, interpretation of the distribution, pattern and extent of language activation can be heavily dependent on the chosen statistical threshold. The aim of the present study was to 1) test the robustness of adaptive thresholding of fMRI data to yield a fixed number of active voxels, and to 2) develop a largely threshold-independent method of assessing when individual patients have statistically atypical language lateralization. Simulated data and real fMRI data in 34 healthy controls and 4 selected epilepsy patients performing a verbal fluency language fMRI task were used. Dependence of laterality on the thresholding method is demonstrated for simulated and real data. Simulated data were used to test the hypothesis that thresholding based upon a fixed number of active voxels would yield a laterality index that was more stable across a range of signal strengths (study power) compared to thresholding at a fixed p value. This stability allowed development of a method comparing an individual to a group of controls across a wide range of thresholds, providing a robust indication of atypical lateralization that is more objective than conventional methods. Thirty healthy controls were used as normative data for the threshold-independent method, and the remaining subjects were used as illustrative examples. The method could also be used more generally to assess relative regional distribution of activity in other neuroimaging paradigms (for example, one could apply it to the assessment of lateralization of activation in a memory task, or to the assessment of anterior–posterior distribution rather than laterality).

YNIMG Journal 2010 Journal Article

Focal epileptiform spikes do not show a canonical BOLD response in patients with benign rolandic epilepsy (BECTS)

Richard A.J. Masterton
A. Simon Harvey
John S. Archer
Leasha M. Lillywhite
David F. Abbott
Ingrid E. Scheffer
Graeme D. Jackson

Simultaneous EEG and functional MRI (EEG-fMRI) studies of focal epileptiform spikes commonly use the canonical haemodynamic response function (HRF) to model the blood-oxygenation-level-dependent (BOLD) response to these events. Support for the use of the canonical HRF has come from large studies that contain mixed cohorts of epilepsy syndromes and discharge types, and has demonstrated plausible epileptic localisation results in the majority of patients. Other studies, however, have reported that some patients show a BOLD response that differs markedly from a canonical HRF. Our aim in this study was to see if the BOLD response is well modelled by a canonical HRF in a homogeneous cohort of patients with benign epilepsy with centrotemporal spikes (BECTS), an idiopathic partial epilepsy with stereotypical centrotemporal spikes on the EEG. We studied eight well-characterised and typical BECTS patients and found that the shape of the average BOLD response was different to the canonical HRF. Furthermore, a localisation analysis using the group-average response provided increased sensitivity and specificity compared to the canonical HRF. Our findings suggest that the canonical HRF may not provide the best model for the BOLD response in some epilepsy syndromes or spike-types. In studies of homogeneous patient groups, therefore, localisation results may be improved by using a group-specific BOLD response.

YNIMG Journal 2009 Journal Article

Voxel-Based Iterative Sensitivity (VBIS) analysis: Methods and a validation of intensity scaling for T2-weighted imaging of hippocampal sclerosis

David F. Abbott
Gaby S. Pell
Heath Pardoe
Graeme D. Jackson

Abnormalities in the brain generally manifest on MRI as changes in shape (morphometry) or changes in the nature of the tissue (signal intensity). Voxel Based Morphometry (VBM) is a whole brain quantitative way of assessing morphometric changes. Voxel Based Relaxometry (VBR) directly assesses signal intensity changes in quantitative maps of T2 relaxation time, but this requires specialised multiple-echo acquisition sequences that are not usually available at clinical sites. This paper introduces and assesses an objective voxel-based statistical method for evaluation of signal intensity in groups of routinely acquired qualitative images. We call the method Voxel-Based Iterative Sensitivity (VBIS) analysis. It adaptively optimises the relative global scaling of images to maximise sensitivity to regional effects. We apply and validate the method of analysis for T2-weighted images of the human brain. To validate the method, it was directly compared with VBR by extracting T2-weighted images of a single echo from multi-echo T2 relaxometry acquisitions from a group of 24 patients with left hemisphere hippocampal sclerosis and 97 healthy controls. Expected signal abnormalities in the patients were detectable with VBIS-T2, confirming the feasibility of the technique. This opens the door to the use of a voxel-based analysis approach on the vast amount of T2-weighted image data that has been and is being acquired on MRI scanners. When a quantitative modality is not available, VBIS can be an effective way to quantify differences between groups. We expect the method could also assist quantitative analysis of other qualitative modalities such as T1-weighted MRI, SPECT and CT.

YNIMG Journal 2008 Journal Article

Composite voxel-based analysis of volume and T2 relaxometry in temporal lobe epilepsy

Gaby S. Pell
Regula S. Briellmann
Heath Pardoe
David F. Abbott
Graeme D. Jackson

Voxel-based analyses of tissue characteristics such as volume and T2 are usually carried out in isolation. However, as the images are analysed in a common voxel-based framework, it is possible to directly assess the spatial relationships of abnormalities detected by each technique. We utilize this approach in well-characterized patients with unilateral temporal lobe epilepsy (TLE) with hippocampal sclerosis (HS). TLE is associated with potentially widespread volume and T2 signal abnormalities in MRI images but the relationship between these two aspects of tissue abnormality is not well understood. Here we use a novel approach of combined univariate and multivariate voxel-wise analysis to investigate the spatial relationship of these abnormalities. We studied 19 TLE patients and compared them to 115 control subjects. Grey matter (GM) and white matter (WM) volume changes were assessed with voxel-based morphometry (VBM), and changes in T2 relaxation times were evaluated with voxel-based relaxometry (VBR). The volume and T2 changes obtained using the combined univariate approach were found in an extensive area, prominently in the ipsilateral hippocampus and amygdala (overlap of GM–VBM and VBR), and in the remaining temporal lobe (overlap of WM–VBR and VBR). Other cortical and subcortical areas showed isolated volume or T2 changes. The multivariate analysis based on the Hotelling T 2 statistic, indicated a similar pattern of distributed changes across the brain but with a greater degree of statistical significance in certain areas. The composite analyses appear to identify a network of affected areas not as easily appreciated by the individual analysis of volume or T2 changes.

YNIMG Journal 2008 Journal Article

Multi-site voxel-based morphometry: Methods and a feasibility demonstration with childhood absence epilepsy

Heath Pardoe
Gaby S. Pell
David F. Abbott
Anne T. Berg
Graeme D. Jackson

Aim: Voxel-based morphometry analysis of neurological disorders would benefit if it could use data acquired from different scanners, but scanner based contrast variation could interfere with the detection of disease-specific structural abnormalities. In this study we examine MRI data from three different sites to investigate structural differences between childhood absence epilepsy (CAE) subjects and controls. Methods: T1-weighted structural MRI scans were acquired from: Site A. 10 CAE, 213 controls; Site B. 15 CAE, 33 controls; and Site C. 19 CAE, 11 controls. The images were processed using the optimised VBM protocol. Three statistical analyses were undertaken: (1) Comparisons of CAE subjects and controls stratified by site. (2) Between-site comparison of controls from each site. (3) Factorial analysis of all data with site and disease status as factors. Results: Consistent regions of structural change, located in the thalamic nuclei, were observed in the within-site analysis of CAE vs controls. Analysis of control scans, however, indicated site-specific differences between controls, which required that we adjust for site in combined analyses. Analysis of all data with adjustment for site confirmed the finding of thalamic atrophy in CAE cases. Conclusion: Combined VBM analysis of structural MRI scans acquired from different sites yield consistent patterns of structural change in CAE when site is included as a factor in the statistical analysis of the processed images. In MRI studies of diseases where only a limited number of subjects can be imaged at each site, our study supports the possibility of effective multi-site studies as long as both disease subjects and healthy controls are acquired from each site.

YNIMG Journal 2008 Journal Article

Reproducible activation in BA2, 1 and 3b associated with texture discrimination in healthy volunteers over time

Leeanne M. Carey
David F. Abbott
Gary F. Egan
Geoffrey A. Donnan

We aimed to quantify specific location and reproducibility of brain activation associated with discrimination of a moving textured surface in adult healthy volunteers over a 6-month interval. A sensory stimulation device was developed to provide a texture stimulus to the fingertips at a controlled speed and pressure. Repeat measurements of regional cerebral blood flow, using positron emission tomography (PET), were obtained in 10 healthy individuals, aged 33 to 80 years (mean=55. 8 years), at scanning sessions separated by 6 months. Stimulation and rest conditions were presented to either the right, dominant (n =5) or left non-dominant (n =5) hand. Activation location was objectively quantified with reference to probabilistic cytoarchitectonic maps. Differences in activation over time and regions of common activation were also quantified. Participants consistently activated Brodmann areas (BA) 2, 3b and 1, somatosensory areas of postcentral gyrus, at initial and 6-month studies: 93. 1% of common activation for the right-hand (RH) and 60. 6% for left-hand (LH) stimulation group were in these areas. Reproducible activation in BA6, 4a and 4p was also observed for the RH group (6. 8% of common activation) and LH group (39. 4%). There were no sites of significant difference over time for either hand. Highly consistent location of activation over time suggests that changes in loci of activation may be confidently monitored in adults using this paradigm. Use of probabilistic cytoarchitectonic maps permitted objective quantification of the anatomical location of the core of reproducible activation.

YNIMG Journal 2008 Journal Article

Selection of the control group for VBM analysis: Influence of covariates, matching and sample size

Gaby S. Pell
Regula S. Briellmann
Chow Huat (Patrick) Chan
Heath Pardoe
David F. Abbott
Graeme D. Jackson

Variability in the control group plays a crucial role in voxel-based morphometry (VBM) detection of structural abnormalities. Two common methods of minimising this variance are inclusion of covariates and matching of control and patient groups. We address two major questions: What are the optimal covariates in the VBM design? When a large pool of controls are available, is it better to choose a subset of matched control subjects at the expense of numbers, or include all available controls? We used regression analysis in a group of 176 controls to determine the contribution of gender, age, and total intracranial volume (TIV) to volume variation. We then used different matching and covariate strategies to determine the optimal design for VBM detection of abnormality in epilepsy patients with hippocampal sclerosis. In the regression analysis, focal gender effects disappeared with inclusion of TIV as an additional regressor. Age had a small but unique contribution to focal volume changes. In the VBM analysis of HS patients, detection of abnormalities was strongly influenced by choice of covariates. The optimal combination was different for grey and white matter (for grey matter: TIV; for temporal lobe white matter: TIV, age and gender). A control group size of 70–90 subjects allowed optimal detection of volume loss in the hippocampus and thalamus. At these group sizes, matched control groups did not consistently prove superior to deliberately “unmatched” groups of the same size. The optimal detection of volume loss was obtained with all available control subjects.

YNIMG Journal 2007 Journal Article

Measurement and reduction of motion and ballistocardiogram artefacts from simultaneous EEG and fMRI recordings

Richard A.J. Masterton
David F. Abbott
Steven W. Fleming
Graeme D. Jackson

Recording the electroencephalogram (EEG) during functional magnetic resonance imaging (fMRI) permits the identification of haemodynamic changes associated with EEG events. However, subject motion within the MR scanner can cause unpredictable and frustrating artefacts on the EEG that may appear focally, bilaterally or unilaterally and can sometimes be confused for epileptiform activity. Motion may arise from a number of sources: small involuntary cardiac-related body movements (ballistocardiogram); acoustic vibrations due to the scanner machinery; and voluntary subject movements. Here we describe a new real-time technique for removing ballistocardiogram (BCG) and movement artefact from EEG recordings in the MR scanner using a novel method for recording subject motion. We record the current induced in a number of wire loops, attached to a cap worn by the subject, due to motion in the static magnetic field of the scanner (Faraday's Law). This is the same process that leads to the motion artefacts on the EEG, and hence these signals are ideally suited to filtering these artefacts from the EEG. Our filter uses a linear adaptive technique based upon the Recursive Least Squares (RLS) algorithm. We demonstrate in both simulations and real EEG recordings from epilepsy patients that our filter significantly reduces the artefact power whilst preserving the underlying EEG signal.

YNIMG Journal 2005 Journal Article

Differential amygdala responses to happy and fearful facial expressions depend on selective attention

Mark A. Williams
Francis McGlone
David F. Abbott
Jason B. Mattingley

Facial expressions of emotion elicit increased activity in the human amygdala. Such increases are particularly evident for expressions that convey potential threat to the observer, and arise even when the face is masked from awareness. We used functional magnetic resonance imaging (fMRI) to examine whether the amygdala responds differentially to threatening (fearful) versus nonthreatening (happy) facial expressions depending on whether the face is attended or actively ignored. In separate runs, participants were cued to attend to a face or a house within semitransparent, spatially overlaid composite pairs, presented either side of fixation, and were required to perform a demanding same/different judgment. We found significant attentional modulation of activity in category-specific ‘face’ (fusiform gyrus) and ‘place’ (parahippocampal gyrus) regions, with activity in each area increasing selectively when its preferred stimulus was attended versus ignored. In contrast, activity in the amygdala differed according to the valence of the facial expression and the category of the attended stimulus. For happy faces, activity in the amygdala was greater in the attend-face than in the attend-house condition, whereas for fearful faces, activity was greater in the attend-house than in the attend-face condition. We conclude that differential amygdala responses to fearful versus happy facial expressions are tuned by mechanisms of attention and that the amygdala gives preference to potentially threatening stimuli under conditions of inattention.

YNIMG Journal 2005 Journal Article

How reliable are fMRI–EEG studies of epilepsy? A nonparametric approach to analysis validation and optimization

Anthony B. Waites
Marnie E. Shaw
Regula S. Briellmann
Angelo Labate
David F. Abbott
Graeme D. Jackson

Simultaneously acquired functional magnetic resonance imaging (fMRI) and electroencephalography (EEG) data hold great promise for localizing the spatial source of epileptiform events detected in the EEG trace. Despite a number of studies applying this method, there has been no independent and systematic validation of the approach. The present study uses a nonparametric method to show that interictal discharges lead to a blood oxygen level dependent (BOLD) response that is significantly different to that obtained by examining random ‘events’. We also use this approach to examine the optimization of analysis strategy for detecting these BOLD responses. Two patients with frequent epileptiform events and a healthy control were studied. The fMRI data for each patient were analyzed using a model derived from the timings of the epileptiform events detected on EEG during fMRI scanning. Twenty sets of random pseudoevents were used to generate a null distribution representing the level of chance correlation between the EEG events and fMRI data. The same pseudoevents were applied to control data. We demonstrate that it is possible to detect blood oxygen level-dependent (BOLD) changes related to interictal discharges with specific and independent knowledge about the reliability of this activation. Biologically generated events complicate the fMRI–EEG experiment. Our proposed validation examines whether identified events have an associated BOLD response beyond chance and allows optimization of analysis strategies. This is an important step beyond standard analysis. It informs clinical interpretation because it permits assessment of the reliability of the connection between interictal EEG events and the BOLD response to those events.

YNIMG Journal 2000 Journal Article

The Functional Neuroanatomy and Long-Term Reproducibility of Brain Activation Associated with a Simple Finger Tapping Task in Older Healthy Volunteers: A Serial PET Study

Leeanne M. Carey
David F. Abbott
Gary F. Egan
Henri J. Tochon-Danguy
Geoffrey A. Donnan

We examined long-term reproducibility of the functional organization of the brain associated with a simple finger tapping movement using positron emission tomography (PET). Repeat measurements of regional cerebral blood flow were obtained in 10 individuals, ages 35 to 82 years (mean 52 years), at scanning sessions separated by 6 months. Although the functional neuroanatomy of hand movements has previously been investigated with PET by a number of groups, none has reported systematic investigation of the consistency of brain activation over an extended time. As expected, we found significant activation in the left precentral gyrus [Talairach coordinate (−32, −34, 52)], postcentral gyrus (−22, −48, 56), and supplementary motor area (SMA) (−2, −18, 52) at the initial study, consistent with previous studies in younger subjects. For the follow-up study we also found significant activation in the left precentral (−36, −28, 52) and postcentral (−28, −36, 52) gyri and in the SMA (2, −16, 56). Our group results demonstrate consistent anatomical location and extent of motor activation over time. More importantly, analysis of individuals confirmed the presence of consistent sites of activation in primary sensorimotor cortex and SMA over the 6-month interval in most subjects. A high degree of consistency in location of activation in the group, and within individuals, over time suggests that changes in loci of activation may be confidently monitored using the PET method. Evidence of individual differences in extent of activation over time highlights the need for caution when interpreting similar changes in patient studies.