ORCID Profile
0000-0002-5952-0516
Current Organisation
University of Adelaide
Does something not look right? The information on this page has been harvested from data sources that may not be up to date. We continue to work with information providers to improve coverage and quality. To report an issue, use the Feedback Form.
Publisher: BMJ
Date: 04-2022
DOI: 10.1136/BMJOPEN-2020-045908
Abstract: Transient ischaemic attack (TIA) may be a warning sign of stroke and difficult to differentiate from minor stroke and TIA-mimics. Urgent evaluation and diagnosis is important as treating TIA early can prevent subsequent strokes. Recent improvements in mass spectrometer technology allow quantification of hundreds of plasma proteins and lipids, yielding large datasets that would benefit from different approaches including machine learning. Using plasma protein, lipid and radiological biomarkers, our study will develop predictive algorithms to distinguish TIA from minor stroke (positive control) and TIA-mimics (negative control). Analysis including machine learning employs more sophisticated modelling, allowing non-linear interactions, adapting to datasets and enabling development of multiple specialised test-panels for identification and differentiation. Patients attending the Emergency Department, Stroke Ward or TIA Clinic at the Royal Adelaide Hospital with TIA, minor stroke or TIA-like symptoms will be recruited consecutively by staff-alert for this prospective cohort study. Advanced neuroimaging will be performed for each participant, with images assessed independently by up to three expert neurologists. Venous blood s les will be collected within 48 hours of symptom onset. Plasma proteomic and lipid analysis will use advanced mass spectrometry (MS) techniques. Principal component analysis and hierarchical cluster analysis will be performed using MS software. Output files will be analysed for relative biomarker quantitative differences between the three groups. Differences will be assessed by linear regression, one-way analysis of variance, Kruskal-Wallis H-test, χ 2 test or Fisher’s exact test. Machine learning methods will also be applied including deep learning using neural networks. Patients will provide written informed consent to participate in this grant-funded study. The Central Adelaide Local Health Network Human Research Ethics Committee approved this study (HREC/18/CALHN/384 R20180618). Findings will be disseminated through peer-reviewed publication and conferences data will be managed according to our Data Management Plan (DMP2020-00062).
Publisher: IEEE
Date: 08-2014
Publisher: Elsevier BV
Date: 06-2014
Publisher: Walter de Gruyter GmbH
Date: 10-2006
DOI: 10.1515/BMT.2006.033
Publisher: Walter de Gruyter GmbH
Date: 10-2006
DOI: 10.1515/BMT.2006.036
Publisher: Cold Spring Harbor Laboratory
Date: 23-09-2022
DOI: 10.1101/2022.09.22.509002
Abstract: Modelling population reference curves or normative modelling is increasingly used with the advent of large neuroimaging studies. In this paper we assess the performance of fitting methods from the perspective of clinical applications and investigate the influence of the s le size. Further, we evaluate linear and nonlinear models for percentile curve estimation and highlight how the bias-variance trade-off manifests in typical neuroimaging data. We created plausible ground truth distributions of hippoc al volumes in the age range of 45 to 80 years, as an ex le application. Based on these distributions we repeatedly simulated s les for sizes between 50 and 50,000 data points, and for each simulated s le we fitted a range of normative models. We compared the fitted models and their variability across repetitions to the ground truth, with specific focus on the outer percentiles (1 th , 5 th , 10 th ) as these are the most clinically relevant. Our results quantify the expected decreasing trend in variance of the volume estimates with increasing s le size. However, bias in the volume estimates only decreases a modest amount, without much improvement at large s le sizes. The uncertainty of model performance is substantial for what would often be considered large s les in a neuroimaging context and rises dramatically at the ends of the age range, where fewer data points exist. Flexible models perform better across s le sizes, especially for nonlinear ground truth. Surprisingly large s les of several thousand data points are needed to accurately capture outlying percentiles across the age range for applications in research and clinical settings. Performance evaluation methods should assess both, bias and variance. Furthermore, extreme caution is needed when attempting to extrapolate beyond the age range included in the source dataset. To help with such evaluations of normative models we have made our code available to guide researchers developing or utilising normative models.
Publisher: IEEE
Date: 12-2017
Publisher: Frontiers Media SA
Date: 07-04-2016
Publisher: Alfons W. Gentner Verlag GmbH & Co. KG
Date: 29-09-2020
Abstract: Methods for the determination of parameters for the electrical warning of persons Objectives: The aims of this publication are to present the main findings of research into the development of a wearable system for electrical warning, identify the current challenges and introduce the next research objectives. Methods: A basic study (n = 81) with self-adhesive electrodes on the right upper arm was used to investigate the influence of pulse width, electrode size and electrode position on perceived thresholds as well as qualitative and spatial perception. Varying textile cuff types were developed and tested. The suitability of varying support materials and textile electrodes was investigated with regard to adaptability, comfort, electrical conductivity, DC resistance and traction elastic behaviour. The textile and self-adhesive electrodes were compared with regard to thresholds as well as qualitative and spatial perceptions (n = 30). Results: Practical parameter sets of the thresholds (perception, attention, intolerance) were determined for various pulse widths, electrode sizes and positions. The dominant qualitative perceptions were “Knocking” (perception and attention threshold) and “Muscle twitch” (intolerance threshold). The spatial perception was located at the stimulation area. The resulting textile cuff contains a knitted fabric with electrically conductive surfaces and a layer of an electrically conductive silicone compound. The comparison between textile and self-adhesive electrodes showed no differences regarding thresholds and qualitative and spatial perceptions. The impedance of the textile electrodes was (1.5 to 3 times) higher than that of the self-adhesive electrodes. Conclusions: Future studies will investigate the influences of working conditions, climatic conditions, age, gender and skin properties. The further development of the textile cuffs is focused on the improvement of the contact between electrode and skin to optimise the transition impedance. Keywords: electric stimulation – electrode – TENS – smart textile – wearable
Publisher: Institute of Electrical and Electronics Engineers (IEEE)
Date: 2022
Publisher: Elsevier BV
Date: 06-2017
Publisher: Walter de Gruyter GmbH
Date: 07-01-2013
Publisher: Springer Science and Business Media LLC
Date: 13-07-2009
Publisher: Springer Science and Business Media LLC
Date: 03-1992
DOI: 10.1007/BF02657009
Publisher: MDPI AG
Date: 24-05-2016
DOI: 10.3390/S16060754
Publisher: Walter de Gruyter GmbH
Date: 2017
Publisher: No publisher found
Date: 2008
Publisher: Walter de Gruyter GmbH
Date: 04-01-2012
Publisher: Elsevier BV
Date: 06-2014
Publisher: Frontiers Media SA
Date: 09-04-2020
Publisher: Walter de Gruyter GmbH
Date: 30-01-2012
Publisher: Walter de Gruyter GmbH
Date: 04-01-2012
Publisher: Walter de Gruyter GmbH
Date: 2016
Publisher: Elsevier BV
Date: 11-2008
DOI: 10.1016/J.CLINPH.2008.08.011
Abstract: The source of somatosensory evoked high-frequency activity at about 600 Hz is still not completely clear. Hence, we aimed to study the influence of double stimulation on the human somatosensory system by analyzing both the low-frequency activity and the high-frequency oscillations (HFOs) at about 600 Hz. We used median nerve stimulation at seven interstimuli intervals (ISIs) with a high time resolution between 2.4 and 4.8 ms to investigate the N15, N20 and superimposed HFOs. Simultaneously, the electroencephalogram and the magnetoencephalogram of 12 healthy participants were recorded. Subsequently, the source analysis of precortical and cortical dipoles was performed. The difference computations of precortical dipole activation curves showed in both the low- and high-frequency range a correlation between the ISI and the latency of the second stimulus response. The cortical low-frequency response showed a similar behavior. Contrarily, in the second response of cortical HFOs this latency shift could not be confirmed. We found litude fluctuations that were dependent on the ISI in the low-frequency activity and the HFOs. These nonlinear interactions occurred at ISIs, which differ by one full HFO period (1.6 ms). Low-frequency activity and HFOs originate from different generators. Precortical and cortical HFOs are independently generated. The litude fluctuations dependent on ISI indicate nonlinear interference between successive stimuli. Information processing in human somatosensory system includes nonlinearity.
Publisher: Springer Science and Business Media LLC
Date: 06-2015
Publisher: Elsevier BV
Date: 08-2014
DOI: 10.1016/J.CLINPH.2013.12.099
Abstract: Magnetoencephalography (MEG) signals had previously been hypothesized to have negligible sensitivity to skull defects. The objective is to experimentally investigate the influence of conducting skull defects on MEG and EEG signals. A miniaturized electric dipole was implanted in vivo into rabbit brains. Simultaneous recording using 64-channel EEG and 16-channel MEG was conducted, first above the intact skull and then above a skull defect. Skull defects were filled with agar gels, which had been formulated to have tissue-like homogeneous conductivities. The dipole was moved beneath the skull defects, and measurements were taken at regularly spaced points. The EEG signal litude increased 2-10 times, whereas the MEG signal litude reduced by as much as 20%. The EEG signal litude deviated more when the source was under the edge of the defect, whereas the MEG signal litude deviated more when the source was central under the defect. The change in MEG field-map topography (relative difference measure, RDM(∗)=0.15) was geometrically related to the skull defect edge. MEG and EEG signals can be substantially affected by skull defects. MEG source modeling requires realistic volume conductor head models that incorporate skull defects.
Publisher: Elsevier BV
Date: 03-2023
Location: Germany
No related grants have been discovered for Stephan Lau.