Deborah Apthorp is a cognitive neuroscientist specialising in the human visual system and its interactions with other sensory systems, and also how changes in these systems might serve as non-invasive ways to track disease progression in diseases like Parkinson’s disease and MS. She is also currently collaborating with machine learning experts to develop new and exciting ways to analyse multimodal longitudinal data in these studies.
Our lab collects EEG data to answer questions about how the brain processes faces, about how sleep affects brains, and how the brains of people with Parkinson’s disease and Multiple Sclerosis are affected.
Postural sway can be measuresd using a force plate or balance plate (or even a mobile phone accelerometer or a Nintendo Wii). This is a place for all the projects I am involved in that utilise postural sway.
We are doing online experiments and surveys investigating the impact of sleep on essential human functions such as attention, memory, mood and decision-making.
Our Health in Our Hands aims to transform healthcare by developing new personalised health technologies and solutions in collaboration with patients, clinicians and health care providers.
A biomarker of cognition in Multiple Sclerosis (MS) that is independent from the response of people with MS (PwMS) to test questions would provide a more holistic assessment of cognitive decline. One suggested method involves event-related potentials (ERPs). This systematic review tried to answer five questions about the use of ERPs in distinguishing PwMS from controls: which stimulus modality, which experimental paradigm, which electrodes, and which ERP components are most discriminatory, and whether amplitude or latency is a better measure. Our results show larger pooled effect sizes for visual stimuli than auditory stimuli, and larger pooled effect sizes for latency measurements than amplitude measurements. We observed great heterogeneity in methods and suggest that future research would benefit from more uniformity in methods and that results should be reported for the individual subtypes of PwMS. With more standardised methods, ERPs have the potential to be developed into a clinical tool in MS.
In a novel online study, we explored whether finger tapping differences are evident in people with autism spectrum disorder (ASD) traits in the general population. We hypothesised that those with higher autistic traits would show more impairment in finger tapping, and that age would moderate tapping output. The study included a non-diagnosed population of 159 participants aged 18-78 who completed an online measure of autistic traits (the AQ-10) and a measure of finger tapping (the FTT). Results showed those with higher AQ-10 scores recorded lower tapping scores in both hands. Moderation analysis showed younger participants with more ASD traits recorded lower tapping scores for the dominant hand. This suggests motor differences seen in ASD studies are evident in the general population.
Adoption of artificial intelligence (AI) by the medical community has long been anticipated, endorsed by a stream of machine learning literature showcasing AI systems that yield extraordinary performance. However, many of these systems are likely over-promising and will under-deliver in practice. One key reason is the community’s failure to acknowledge and address the presence of inflationary effects in the data. These simultaneously inflate evaluation performance and prevent a model from learning the underlying task, thus severely misrepresenting how that model would perform in the real world. This paper investigated the impact of these inflationary effects on healthcare tasks, as well as how these effects can be addressed. Specifically, we defined three inflationary effects that occur in medical data sets and allow models to easily reach small training losses and prevent skillful learning. We investigated two data sets of sustained vowel phonation from participants with and without Parkinson’s disease, and revealed that published models which have achieved high classification performances on these were artificially enhanced due to the inflationary effects. Our experiments showed that removing each inflationary effect corresponded with a decrease in classification accuracy, and that removing all inflationary effects reduced the evaluated performance by up to 30 percent. Additionally, the performance on a more realistic test set increased, suggesting that the removal of these inflationary effects enabled the model to better learn the underlying task and generalize. Source code is available at https://github.com/Wenbo-G/pd-phonation-analysis under the MIT license.
In April 2019, Psychological Science published its first issue in which all Research Articles received the Open Data badge. We used that issue to investigate the effectiveness of this badge, focusing on the adherence to its aim at Psychological Science: sharing both data and code to ensure reproducibility of results. Twelve researchers of varying experience levels attempted to reproduce the results of the empirical articles in the target issue (at least three researchers per article). We found that all 14 articles provided at least some data and six provided analysis code, but only one article was rated to be exactly reproducible, and three were rated as essentially reproducible with minor deviations. We suggest that researchers should be encouraged to adhere to the higher standard in force at Psychological Science. Moreover, a check of reproducibility during peer review may be preferable to the disclosure method of awarding badges.
As anthropogenic climate change progresses, there is an increasing need for individuals to make appropriate decisions regarding their approach to extreme weather events. Natural hazards are involuntary risk environments (e.g., flooded roads); interaction with them cannot be avoided (i.e., a decision must be made about how to engage). While the psychological and sociocultural predictors of engagement with voluntary risks (i.e., risk situations that are sought out) are well-documented, less is known about the factors that predict engagement with involuntary risk environments. This exploratory study assessed whether mental health (depression, anxiety, and stress symptoms), personality traits, and cultural worldviews combine to predict engagement with involuntary risk, using the situation of floodwater driving. An Australian sample (N = 235) was assessed via questionnaire and scenario measures. Results were analyzed in a binomial logistic regression assessing which individual factors predicted decision-making in a proxy floodwater driving scenario. Agreeableness and gender were individually significant predictors of floodwater driving intention, and four factors (named affect, progressiveness, insightfulness, and purposefulness) were derived from an exploratory factor analysis using the variables of interest, though only two (progressiveness and insightfulness) predicted floodwater driving intention in an exploratory binomial logistic regression. The findings highlight the need for further research into the differences between voluntary and involuntary risk. The implication of cultural worldviews and personality traits in interaction with mental health indicators on risk situations is discussed.
Parkinson’s Disease (PD) is a progressive chronic disorder with a high misdiagnosis rate. Because finger-tapping tasks correlate with its fine-motor symptoms, they could be used to help diagnose and assess PD. We first designed and developed an Android application to perform finger-tapping tasks without trained supervision, which is not always feasible for patients. Then, we conducted a preliminary user evaluation in Australia with six patients clinically diagnosed with PD and sixteen controls without PD. The application could be used in research and healthcare for regular symptom and progression assessment and feedback.
Computer-assisted quantification and analysis of postural sway may support identifying individuals affected by Parkinson’s disease (PD). Balancing, and its associated postural sway, is a complex process that requires the cooperation of several sensory systems in the brain. Unsurprisingly, a neurodegenerative disease can affect such processes, manifesting itself in the postural sway of affected individuals.
Different aspects of postural sway can be quantified and represented as features, which can be used to distinguish between patients and controls. Our aim, inspired by a recent systematic literature review, was to experimentally determine whether sampling frequency and visual state had a meaningful impact on the effectiveness of features in distinguishing between the two groups, and whether overall discriminability could be improved using machine learning.
We extracted 102 unique features from 78 postural sway recordings and found that the effectiveness (quantified by an effect size and the average area under the receiver operating characteristic curve) with a sampling frequency of 10 Hz was superior to 20, 40, and 100 Hz, though not with high confidence (quantified through Bayesian analysis). We also concluded that effectiveness under the eyes closed condition was higher than the eyes open condition (confirmed through Bayesian analysis), though combining features from both conditions was superior. Finally, we showed that using machine learning to analyse multiple features through feature selection resulted in higher discriminability in almost all cases.
The code for these experiments have been released at https://github.com/Wenbo-G/pd-sway-analysis under the MIT license. When using our code, please cite this paper.
Postural control deficits are well documented in schizophrenia. However, postural stability has not been assessed in first-degree relatives of individuals with schizophrenia to our knowledge. We analyzed postural sway in 27 controls (CTR) and 18 first-degree relatives (REL). The REL group was significantly impaired compared to CTR, with a larger mean sway area and longer mean sway path. These preliminary findings suggest a genetic contribution to postural control deficits observed in schizophrenia spectrum disorders. Future studies should, however, examine the contributions of shared environmental risk factors including stress, toxins, etc. to familial risk to dissociate them from shared genetic risk.
Weber’s law predicts that stimulus sensitivity will increase proportionally with increases in stimulus intensity. Does this hold for the stimulus of time – specifically, duration in the milliseconds to seconds range? There is conflicting evidence on the relationship between temporal sensitivity and duration. Weber’s law predicts a linear relationship between sensitivity and duration on interval timing tasks, while two alternative models predict a reverse J-shaped and a U-shaped relationship. Based on previous research, we hypothesised that temporal sensitivity in humans would follow a U-shaped function, increasing and then decreasing with increases in duration, and that this model would provide a better statistical fit to the data than the reverse-J or the simple Weber’s Law model. In a two-alternative forced-choice interval comparison task, 24 participants made duration judgements about six groups of auditory intervals between 100 and 3,200 ms. Weber fractions were generated for each group of intervals and plotted against time to generate a function describing sensitivity to the stimulus of duration. Although the sensitivity function was slightly concave, and the model describing a U-shaped function gave the best fit to the data, the increase in the model fit was not sufficient to warrant the extra free parameter in the chosen model. Further analysis demonstrated that Weber’s law itself provided a better description of sensitivity to changes in duration than either of the two models tested.
We measured postural sway in individuals diagnosed with Parkinson’s disease and age-matched controls. Individuals with Parkinson’s swayed more, as expected, especially when their eyes were closed. In the people with Parkinson’s, sway correlated strongly with cognitive measures, as well as with measures of quality of life and clinical status.
Background: Postural sway may be useful as an objective measure of Parkinson’s disease (PD). Existing studies have analyzed many different features of sway using different experimental paradigms. We aimed to determine what features have been used to measure sway and then to assess which feature(s) best differentiate PD patients from controls. We also aimed to determine whether any refinements might improve discriminative power and so assist in standardizing experimental conditions and analysis of data.
Methods: In this systematic review of the literature, effect size (ES) was calculated for every feature reported by each article and then collapsed across articles where appropriate. The influence of clinical medication status, visual state, and sampling rate on ES was also assessed.
Results: Four hundred and forty‐three papers were retrieved. 25 contained enough information for further analysis. The most commonly used features were not the most effective (e.g., PathLength, used 14 times, had ES of 0.47, while TotalEnergy, used only once, had ES of 1.78). Increased sampling rate was associated with increased ES (PathLength ES increased to 1.12 at 100 Hz from 0.40 at 10 Hz). Measurement during “OFF” clinical status was associated with increased ES (PathLength ES was 0.83 OFF compared to 0.21 ON).
Conclusions: This review identified promising features for analysis of postural sway in PD, recommending a sampling rate of 100 Hz and studying patients when OFF to maximize ES. ES complements statistical significance as it is clinically relevant and is easily compared across experiments. We suggest that machine learning is a promising tool for the future analysis of postural sway in PD.
Background: Parkinson disease (PD) is a common neurodegenerative disorder that affects between 7 and 10 million people worldwide. No objective test for PD currently exists, and studies suggest misdiagnosis rates of up to 34%. Machine learning (ML) presents an opportunity to improve diagnosis; however, the size and nature of data sets make it difficult to generalize the performance of ML models to real-world applications.
Objective: This study aims to consolidate prior work and introduce new techniques in feature engineering and ML for diagnosis based on vowel phonation. Additional features and ML techniques were introduced, showing major performance improvements on the large mPower vocal phonation data set.
Methods: We used 1600 randomly selected /aa/ phonation samples from the entire data set to derive rules for filtering out faulty samples from the data set. The application of these rules, along with a joint age-gender balancing filter, results in a data set of 511 PD patients and 511 controls. We calculated features on a 1.5-second window of audio, beginning at the 1-second mark, for a support vector machine. This was evaluated with 10-fold cross-validation (CV), with stratification for balancing the number of patients and controls for each CV fold.
Results: We showed that the features used in prior literature do not perform well when extrapolated to the much larger mPower data set. Owing to the natural variation in speech, the separation of patients and controls is not as simple as previously believed. We presented significant performance improvements using additional novel features (with 88.6% certainty, derived from a Bayesian correlated t test) in separating patients and controls, with accuracy exceeding 58%.
Conclusions: The results are promising, showing the potential for ML in detecting symptoms imperceptible to a neurologist.
Sleep restriction affects attention in different ways. Performance on an attentional blink task was unaffected by sleep restriction in two studies, but performance on a vigilance task was affected in both. In the second study, we looked at resting state EEG and found alpha was reduced after sleep restriction, which may have balanced out performance on the attentional blink task.
Abnormalities of cerebellar function have been implicated in the pathophysiology of schizophrenia. Since the cerebellum has afferent and efferent projections to diverse brain regions, abnormalities in cerebellar lobules could affect functional connectivity with multiple functional systems in the brain. Prior studies, however, have not examined the relationship of individual cerebellar lobules with motor and nonmotor resting‐state functional networks. We evaluated these relationships using resting‐state fMRI in 30 patients with a schizophrenia‐spectrum disorder and 37 healthy comparison participants. For connectivity analyses, the cerebellum was parcellated into 18 lobular and vermal regions, and functional connectivity of each lobule to 10 major functional networks in the cerebrum was evaluated. The relationship between functional connectivity measures and behavioral performance on sensorimotor tasks (i.e., finger‐tapping and postural sway) was also examined. We found cerebellar–cortical hyperconnectivity in schizophrenia, which was predominantly associated with Crus I, Crus II, lobule IX, and lobule X. Specifically, abnormal cerebellar connectivity was found to the cerebral ventral attention, motor, and auditory networks. This cerebellar–cortical connectivity in the resting‐state was differentially associated with sensorimotor task‐based behavioral measures in schizophrenia and healthy comparison participants—that is, dissociation with motor network and association with nonmotor network in schizophrenia. These findings suggest that functional association between individual cerebellar lobules and the ventral attentional, motor, and auditory networks is particularly affected in schizophrenia. They are also consistent with dysconnectivity models of schizophrenia suggesting cerebellar contributions to a broad range of sensorimotor and cognitive operations.
We measured postural sway in individuals diagnosed with schizotypal personality disorder, but otherwise free of medication and other comorbidities. They swayed significantly more than matched controls, and as much as people dignosed with schizophrenia, in all conditions.
Symmetry is a ubiquitous feature in the visual environment and can be detected by a variety of species, ranging from insects through to humans. Here we show it can also bias estimates of basic scene properties. Mirror (reflective) symmetry can be detected in as little as 50 ms, in both natural and artificial visual scenes, and even when embedded within cluttered backgrounds. In terms of its biological relevance, symmetry is a key determinant in mate selection; the degree of symmetry in a face is positively associated with perceived healthiness and attractiveness ratings. In short, symmetry processing mechanisms are an important part of the neural machinery of vision. We reveal that the importance of symmetry extends beyond the processing of shape and objects. Mirror symmetry biases our perception of scene content, with symmetrical patterns appearing to have fewer components than their asymmetric counterparts. This demonstrates an interaction between two fundamental dimensions of visual analysis: symmetry and number. We propose that this numerical underestimation results from a processing bias away from the redundant information within mirror symmetrical displays, extending existing theories regarding redundancy in visual analysis.
Temporal integration in the visual system causes fast-moving objects to generate static, oriented traces (‘motion streaks’), which could be used to help judge direction of motion. While human psychophysics and single-unit studies in non-human primates are consistent with this hypothesis, direct neural evidence from the human cortex is still lacking. First, we provide psychophysical evidence that faster and slower motions are processed by distinct neural mechanisms: faster motion raised human perceptual thresholds for static orientations parallel to the direction of motion, whereas slower motion raised thresholds for orthogonal orientations. We then used functional magnetic resonance imaging to measure brain activity while human observers viewed either fast (‘streaky’) or slow random dot stimuli moving in different directions, or corresponding static-oriented stimuli. We found that local spatial patterns of brain activity in early retinotopic visual cortex reliably distinguished between static orientations. Critically, a multivariate pattern classifier trained on brain activity evoked by these static stimuli could then successfully distinguish the direction of fast (‘streaky’) but not slow motion. Thus, signals encoding static-oriented streak information are present in human early visual cortex when viewing fast motion. These experiments show that motion streaks are present in the human visual system for faster motion