Levin Kuhlmann | University of Melbourne (original) (raw)

Papers by Levin Kuhlmann

Computers in Biology and Medicine, 2020

Long noncoding RNAs (lncRNAs) are implicated in various genetic diseases and cancer, attributed t... more Long noncoding RNAs (lncRNAs) are implicated in various genetic diseases and cancer, attributed to their critical role in gene regulation. They are a divergent group of RNAs and are easily differentiated from other types with unique characteristics, functions, and mechanisms of action. In this review, we provide a list of some of the prominent data repositories containing lncRNAs, their interactome, and predicted and validated disease associations. Next, we discuss various wet-lab experiments formulated to obtain the data for these repositories. We also provide a critical review of in silico methods available for the identification purpose and suggest techniques to further improve their performance. The bulk of the methods currently focus on distinguishing lncRNA transcripts from the coding ones. Functional annotation of these transcripts still remains a grey area and more efforts are needed in that space. Finally, we provide details of current progress, discuss impediments, and illustrate a roadmap for developing a generalized computational pipeline for comprehensive annotation of lncRNAs, which is essential to accelerate research in this area.

IEEE Reviews in Biomedical Engineering, 2021

ACM Transactions on Reconfigurable Technology and Systems, 2014

Journal of visualized experiments : JoVE, Jan 13, 2018

Anesthesia arguably provides one of the only systematic ways to study the neural correlates of gl... more Anesthesia arguably provides one of the only systematic ways to study the neural correlates of global consciousness/unconsciousness. However to date most neuroimaging or neurophysiological investigations in humans have been confined to the study of γ-Amino-Butyric-Acid-(GABA)-receptor-agonist-based anesthetics, while the effects of dissociative N-Methyl-D-Aspartate-(NMDA)-receptor-antagonist-based anesthetics ketamine, nitrous oxide (N2O) and xenon (Xe) are largely unknown. This paper describes the methods underlying the simultaneous recording of magnetoencephalography (MEG) and electroencephalography (EEG) from healthy males during inhalation of the gaseous anesthetic agents N2O and Xe. Combining MEG and EEG data enables the assessment of electromagnetic brain activity during anesthesia at high temporal, and moderate spatial, resolution. Here we describe a detailed protocol, refined over multiple recording sessions, that includes subject recruitment, anesthesia equipment setup in t...

Parietal networks are hypothesised to play a central role in the cortical information synthesis t... more Parietal networks are hypothesised to play a central role in the cortical information synthesis that supports conscious experience and behavior. Significant reductions in parietal level functional connectivity have been shown to occur during general anesthesia with propofol and a range of other GABAergic general anesthetic agents. Using two analysis approaches (1) a graph theoretic analysis based on surrogate-corrected zero-lag correlations of scalp EEG, and (2) a global coherence analysis based on the EEG cross-spectrum, we reveal that sedation with the NMDA receptor antagonist nitrous oxide (N2O), an agent that has quite different electroencephalographic effects compared to the inductive general anesthetics, also causes significant alterations in parietal level functional networks, as well as changes in full brain and frontal level networks. A total of 20 subjects underwent N2O inhalation at either 20%, 40% or 60% peak N2O/O2 gas concentration levels. N2O-induced reductions in parietal network level functional connectivity (on the order of 50%) were exclusively detected by utilising a surface Laplacian derivation, suggesting that superficial, smaller spatial scale, cortical networks were most affected. In contrast reductions in frontal network functional connectivity were optimally discriminated using a common-reference derivation (reductions on the order of 10%), indicating that the NMDA antagonist N2O induces spatially coherent and widespread perturbations in frontal activity. Our findings not only give important weight to the idea of agent invariant final network changes underlying drug-induced reductions in consciousness, but also provide significant impetus for the application and development of multiscale functional analyses to systematically characterise the network level cortical effects of NMDA receptor related hypofunction. Future work at the source space level will be needed to verify the consistency between cortical network changes seen at the source level and those presented here at the EEG sensor space level.

BMC Neuroscience, 2010

Visual cortex possesses features such as binocular dis-parity (DP), ocular dominance (OD), orient... more Visual cortex possesses features such as binocular dis-parity (DP), ocular dominance (OD), orientation prefer-ence (OR), direction preference (DS), and spatial frequency. Neurophysiological studies have explored the existence of orthogonal relationship between local maps of DP ...

A robust circle criterion observer is designed and applied to neural mass models. At present, no ... more A robust circle criterion observer is designed and applied to neural mass models. At present, no existing circle criterion observers apply to the considered models, i.e. the required linear matrix inequality is infeasible. Therefore, we generalise available results to derive a suitable estimation algorithm. Additionally, the design also takes into account input uncertainty and measurement noise. We show how to apply the observer to estimate the mean membrane potential of neuronal populations of a popular single cortical column model from the literature.

Australian Control …, 2011

Bayesian Spiking Neurons (BSNs) provide a probabilistic interpretation of how neurons can perform... more Bayesian Spiking Neurons (BSNs) provide a probabilistic interpretation of how neurons can perform inference and learning. Learning in a single BSN can be formulated as an online maximum-likelihood expectation-maximisation (ML-EM) algorithm. This form of learning is quite slow. Here, an alternative to this learning algorithm, called Fast Learning (FL), is presented. The FL algorithm is shown to have acceptable convergence performance when compared to the ML-EM algorithm. Moreover, for our implementations the FL algorithm is approximately 25 times faster than the ML-EM algorithm. Although only approximate, the FL algorithm therefore makes learning in hierarchical BSN networks much more tractable.

Automatica, 2014

ABSTRACT Inspired by the neuro-scientific problem of predicting brain dynamics from electroenceph... more ABSTRACT Inspired by the neuro-scientific problem of predicting brain dynamics from electroencephalography (EEG) measurements of the brain’s electrical activity, this paper presents limitations on the observability of networked oscillators sensed with quantised measurements. The problem of predicting highly complex brain dynamics sensed with relatively limited amounts of measurement is abstracted to a study of observability in a network of oscillators. It is argued that a low-dimensional quantised measurement is in fact, by itself, an exceptionally poor observer for a large-scale oscillator network, even for the case of a completely connected graph. The main rational is based on (i) an information-theoretic argument based on ideas of entropy in measure preserving maps, (ii) a linear deterministic observability argument, and (iii) a linear stochastic approach using Kalman filtering. For prediction of brain network activity, the findings indicate that the classic EEG signal is just not precise enough to be able to provide reliable prediction and tracking in a clinical setting in view of the complexity of underlying neural dynamics.

Australian Control Conference, 2011

2021 International Conference on Microelectronics (ICM)

2018 IEEE International Conference on Bioinformatics and Biomedicine (BIBM), Dec 1, 2018

51st IEEE Conference …, 2012

The cause of peak-splitting in the output phase distribution of the leaky integrate-and-fire sing... more The cause of peak-splitting in the output phase distribution of the leaky integrate-and-fire single neuron model in response to low-frequency periodic nerve fibre inputs is analyzed. It is found that peaksplitting largely arises from an increase of the spiking-rate of individual nerve fibre inputs, or from an increase in amplitude of individual input excitatory postsynaptic potentials, or both. These findings add another dimension to the understanding of how peak-splitting arises in the phase histograms of the responses of neurons in the auditory pathway, given that peak-splitting is typically thought to arise as a result of the non-linear dynamics of the basilar membrane and the hair cells. This research has implications for the understanding of the temporal code in the auditory pathway. Index Terms auditory pathway, leaky integrate-and-fire neuron model, peak-splitting, temporal coding

World Academy of Science, Engineering and Technology, International Journal of Medical, Health, Biomedical, Bioengineering and Pharmaceutical Engineering, 2011

Cardiac pulse-related artifacts in the EEG recorded simultaneously with fMRI are complex and high... more Cardiac pulse-related artifacts in the EEG recorded simultaneously with fMRI are complex and highly variable. Their effective removal is an unsolved problem. Our aim is to develop an adaptive removal algorithm based on the matching pursuit (MP) technique and to compare it to established methods using a visual evoked potential (VEP). We recorded the VEP inside the static magnetic field of an MR scanner (with artifacts) as well as in an electrically shielded room (artifact free). The MP-based artifact removal outperformed average artifact subtraction (AAS) and optimal basis set removal (OBS) in terms of restoring the EEG field map topography of the VEP. Subsequently, a dipole model was fitted to the VEP under each condition using a realistic boundary element head model. The source location of the VEP recorded inside the MR scanner was closest to that of the artifact free VEP after cleaning with the MP-based algorithm as well as with AAS. While none of the tested algorithms offered com...

This is an abstract of a paper from Proceedings of the Australian Neuroscience Society 2001 publi... more This is an abstract of a paper from Proceedings of the Australian Neuroscience Society 2001 published by Australian Neuroscience Society. This version is reproduced with the permission of publisher.