Normative models for neuroimaging markers: Impact of model selection, sample size and evaluation criteria.

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 sample size. Further, we evaluate linear and non-linear 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 hippocampal volumes in the age range of 45 to 80 years, as an example application. Based on these distributions we repeatedly simulated samples for sizes between 50 and 50,000 data points, and for each simulated sample 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 (1st, 5th, 10th) as these are the most clinically relevant. Our results quantify the expected decreasing trend in variance of the volume estimates with increasing sample size. However, bias in the volume estimates only decreases a modest amount, without much improvement at large sample sizes. The uncertainty of model performance is substantial for what would often be considered large samples in a neuroimaging context and rises dramatically at the ends of the age range, where fewer data points exist. Flexible models perform better across sample sizes, especially for non-linear ground truth. Surprisingly large samples 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, caution is needed when attempting to go near the ends of the age range captured by the source data set and, as is a well known general principle, extrapolation beyond the age range should always be avoided. To help with such evaluations of normative models we have made our code available to guide researchers developing or utilising normative models.

Optimal Magnetic Sensor Vests for Cardiac Source Imaging.

Magnetocardiography (MCG) non-invasively provides functional information about the heart. New room-temperature magnetic field sensors, specifically magnetoresistive and optically pumped magnetometers, have reached sensitivities in the ultra-low range of cardiac fields while allowing for free placement around the human torso. Our aim is to optimize positions and orientations of such magnetic sensors in a vest-like arrangement for robust reconstruction of the electric current distributions in the heart. We optimized a set of 32 sensors on the surface of a torso model with respect to a 13-dipole cardiac source model under noise-free conditions. The reconstruction robustness was estimated by the condition of the lead field matrix. Optimization improved the condition of the lead field matrix by approximately two orders of magnitude compared to a regular array at the front of the torso. Optimized setups exhibited distributions of sensors over the whole torso with denser sampling above the heart at the front and back of the torso. Sensors close to the heart were arranged predominantly tangential to the body surface. The optimized sensor setup could facilitate the definition of a standard for sensor placement in MCG and the development of a wearable MCG vest for clinical diagnostics.

The experience of freedom in decisions - Questioning philosophical beliefs in favor of psychological determinants.

Six experiments tested two competing models of subjective freedom during decision-making. The process model is mainly based on philosophical conceptions of free will and assumes that features of the process of choosing affect subjective feelings of freedom. In contrast, the outcome model predicts that subjective freedom is due to positive outcomes that can be expected or are achieved by a decision. Results heavily favored the outcome model over the process model. For example, participants felt freer when choosing between two equally good than two equally bad options. Process features including number of options, complexity of decision, uncertainty, having the option to defer the decision, conflict among reasons, and investing high effort in choosing generally had no or even negative effects on subjective freedom. In contrast, participants reported high freedom with good outcomes and low freedom with bad outcomes, and ease of deciding increased subjective freedom, consistent with the outcome model.

Perception thresholds and qualitative perceptions for electrocutaneous stimulation.

Our long-term goal is the development of a wearable warning system that uses electrocutaneous stimulation. To find appropriate stimulation parameters and electrode configurations, we investigate perception amplitude thresholds and qualitative perceptions of electrocutaneous stimulation for varying pulse widths, electrode sizes, and electrode positions. The upper right arm was stimulated in 81 healthy volunteers with biphasic rectangular current pulses varying between 20 and [Formula: see text]. We determined perception, attention, and intolerance thresholds and the corresponding qualitative perceptions for 8 electrode pairs distributed around the upper arm. For a pulse width of [Formula: see text], we find median values of 3.5, 6.9, and 13.8 mA for perception, attention, and intolerance thresholds, respectively. All thresholds decrease with increasing pulse width. Lateral electrode positions have higher intolerance thresholds than medial electrode positions, but perception and attention threshold are not significantly different across electrode positions. Electrode size between [Formula: see text] and [Formula: see text] has no significant influence on the thresholds. Knocking is the prevailing perception for perception and attention thresholds while mostly muscle twitching, pinching, and stinging are reported at the intolerance threshold. Biphasic stimulation pulse widths between [Formula: see text] and [Formula: see text] are suitable for electric warning wearables. Within the given practical limits at the upper arm, electrode size, inter-electrode distance, and electrode position are flexible parameters of electric warning wearables. Our investigations provide the basis for electric warning wearables.

Mutual Information Neural Estimation for Unsupervised Multi-Modal Registration of Brain Images.

Many applications in image-guided surgery and therapy require fast and reliable non-linear, multi-modal image registration. Recently proposed unsupervised deep learning-based registration methods have demonstrated superior per-formance compared to iterative methods in just a fraction of the time. Most of the learning-based methods have focused on mono-modal image registration. The extension to multi-modal registration depends on the use of an appropriate similarity function, such as the mutual information (MI). We propose guiding the training of a deep learning-based registration method with MI estimation between an image-pair in an end-to-end trainable network. Our results show that a small, 2-layer network produces competitive results in both mono- and multi-modal registration, with sub-second run-times. Comparisons to both iterative and deep learning-based methods show that our MI-based method produces topologically and qualitatively superior results with an extremely low rate of non-diffeomorphic transformations. Real-time clinical application will benefit from a better visual matching of anatomical structures and less registration failures/outliers.

The Reproducibility of Bio-Acoustic Features is Associated With Sample Duration, Speech Task, and Gender.

Bio-acoustic properties of speech show evolving value in analyzing psychiatric illnesses. Obtaining a sufficient speech sample length to quantify these properties is essential, but the impact of sample duration on the stability of bio-acoustic features has not been systematically explored. We aimed to evaluate bio-acoustic features' reproducibility against changes in speech durations and tasks. We extracted source, spectral, formant, and prosodic features in 185 English-speaking adults (98 w, 87 m) for reading-a-story and counting tasks. We compared features at 25% of the total sample duration of the reading task to those obtained from non-overlapping randomly selected sub-samples shortened to 75%, 50%, and 25% of total duration using intraclass correlation coefficients. We also compared the features extracted from entire recordings to those measured at 25% of the duration and features obtained from 50% of the duration. Further, we compared features extracted from reading-a-story to counting tasks. Our results show that the number of reproducible features (out of 125) decreased stepwise with duration reduction. Spectral shape, pitch, and formants reached excellent reproducibility. Mel-frequency cepstral coefficients (MFCCs), loudness, and zero-crossing rate achieved excellent reproducibility only at a longer duration. Reproducibility of source, MFCC derivatives, and voicing probability (VP) was poor. Significant gender differences existed in jitter, MFCC first-derivative, spectral skewness, pitch, VP, and formants. Around 97% of features in both genders were not reproducible across speech tasks, in part due to the short counting task duration. In conclusion, bio-acoustic features are less reproducible in shorter samples and are affected by gender.

FAST-IT: Find A Simple Test - In TIA (transient ischaemic attack): a prospective cohort study to develop a multivariable prediction model for diagnosis of TIA through proteomic discovery and candidate lipid mass spectrometry, neuroimaging and machine learning-study protocol.

INTRODUCTION: 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. METHODS AND ANALYSIS: 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 samples 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, χ2 test or Fisher's exact test. Machine learning methods will also be applied including deep learning using neural networks. ETHICS AND DISSEMINATION: 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).

Coupled CP Decomposition of Simultaneous MEG-EEG Signals for Differentiating Oscillators During Photic Driving.

Magnetoencephalography (MEG) and electroencephalography (EEG) are contemporary methods to investigate the function and organization of the brain. Simultaneously acquired MEG-EEG data are inherently multi-dimensional and exhibit coupling. This study uses a coupled tensor decomposition to extract the signal sources from MEG-EEG during intermittent photic stimulation (IPS). We employ the Coupled Semi-Algebraic framework for approximate CP decomposition via SImultaneous matrix diagonalization (C-SECSI). After comparing its performance with alternative methods using simulated benchmark data, we apply it to MEG-EEG recordings of 12 participants during IPS with fractions of the individual alpha frequency between 0.4 and 1.3. In the benchmark tests, C-SECSI is more accurate than SECSI and alternative methods, especially in ill-conditioned scenarios, e.g., involving collinear factors or noise sources with different variances. The component field-maps allow us to separate physiologically meaningful oscillations of visually evoked brain activity from background signals. The frequency signatures of the components identify either an entrainment to the respective stimulation frequency or its first harmonic, or an oscillation in the individual alpha band or theta band. In the group analysis of both, MEG and EEG data, we observe a reciprocal relationship between alpha and theta band oscillations. The coupled tensor decomposition using C-SECSI is a robust, powerful method for the extraction of physiologically meaningful sources from multidimensional biomedical data. Unsupervised signal source extraction is an essential solution for rendering advanced multi-modal signal acquisition technology accessible to clinical diagnostics, pre-surgical planning, and brain computer interface applications.

On the Necessity of Consciousness for Sophisticated Human Action.

In this essay, we aim to counter and qualify the epiphenomenalist challenge proposed in this special issue on the grounds of empirical and theoretical arguments. The current body of scientific knowledge strongly indicates that conscious thought is a necessary condition for many human behaviors, and therefore, consciousness qualifies as a cause of those behaviors. We review illustrative experimental evidence for the causal power of conscious thought while also acknowledging its natural limitations. We argue that it is implausible that the metabolic costs inherent to conscious processes would have evolved in humans without any adaptive benefits. Moreover, we discuss the relevance of conscious thought to the issue of freedom. Many accounts hold conscious thought as necessary and conducive to naturalistic conceptions of personal freedom. Apart from these theories, we show that the conscious perception of freedom and the belief in free will provide sources of interesting findings, beneficial behavioral effects, and new avenues for research. We close by proposing our own challenge via outlining the gaps that have yet to be filled to establish hard evidence of an epiphenomenal model of consciousness. To be sure, we appreciate the epiphenomenalist challenge as it promotes critical thinking and inspires rigorous research. However, we see no merit in downplaying the causal significance of consciousness a priori. Instead, we believe it more worthwhile to focus on the complex interplay between conscious and other causal processes.

Functional Freedom: A Psychological Model of Freedom in Decision-Making.

The freedom of a decision is not yet sufficiently described as a psychological variable. We present a model of functional decision freedom that aims to fill that role. The model conceptualizes functional freedom as a capacity of people that varies depending on certain conditions of a decision episode. It denotes an inner capability to consciously shape complex decisions according to one's own values and needs. Functional freedom depends on three compensatory dimensions: it is greatest when the decision-maker is highly rational, when the structure of the decision is highly underdetermined, and when the decision process is strongly based on conscious thought and reflection. We outline possible research questions, argue for psychological benefits of functional decision freedom, and explicate the model's implications on current knowledge and research. In conclusion, we show that functional freedom is a scientific variable, permitting an additional psychological foothold in research on freedom, and that is compatible with a deterministic worldview.

Ordinary people associate addiction with loss of free will.

INTRODUCTION: It is widely believed that addiction entails a loss of free will, even though this point is controversial among scholars. There is arguably a downside to this belief, in that addicts who believe they lack the free will to quit an addiction might therefore fail to quit an addiction. METHODS: A correlational study tested the relationship between belief in free will and addiction. Follow-up studies tested steps of a potential mechanism: 1) people think drugs undermine free will 2) people believe addiction undermines free will more when doing so serves the self 3) disbelief in free will leads people to perceive various temptations as more addictive. RESULTS: People with lower belief in free will were more likely to have a history of addiction to alcohol and other drugs, and also less likely to have successfully quit alcohol. People believe that drugs undermine free will, and they use this belief to self-servingly attribute less free will to their bad actions than to good ones. Low belief in free will also increases perceptions that things are addictive. CONCLUSIONS: Addiction is widely seen as loss of free will. The belief can be used in self-serving ways that may undermine people's efforts to quit.

Low HRV entropy is strongly associated with myocardial infarction.

Heart rate variability (HRV) is a marker of autonomous activity in the heart. An important application of HRV measures is the stratification of mortality risk after myocardial infarction. Our hypothesis is that the information entropy of HRV, a non-linear approach, is a suitable measure for this assessment. As a first step, to evaluate the effect of myocardial infarction on the entropy, we compared the entropy to standard HRV parameters. The entropy was estimated by compressing the tachogram with Bzip2. For univariate comparison, statistical tests were used. Multivariate analysis was carried out using automatically generated decision trees. The classification rate and the simplicity of the decision trees were the two evaluation criteria. The findings support our hypothesis. The meanNN-normalized entropy is reduced in patients with myocardial infarction with very high significance. One entropy parameter alone exceeds the discrimination strength of multivariate standards-based trees.

Spatiotemporal correlation analyses: a new procedure for standardisation of DC magnetocardiograms.

There is a lack of standard methods for the analysis of magnetocardiograms (MCGs). MCG signals have a shape similar to the ECG (P wave, QRS complex, T wave). High-quality multichannel recordings can indicate even slight disturbances of de- and repolarisation. The purpose of our study was to apply a new approach in the analysis of signal-averaged DC-MCGs. DC-MCGs (31-channel) were recorded in 182 subjects: 110 patients after myocardial infarction and 72 controls. Spatiotemporal correlation analysis of the QRS complex and T wave patterns throughout the entire heart cycle was used to analyse homogeneity of de- and repolarisation. These plots were compared to standard ECG analyses (electrical axis, Q wave, ST deviation, T polarity and shape). Spatiotemporal correlation analyses seem to be applicable in assessing the course of electrical repolarisation with respect to homogeneity. MCG provided all diagnostic information contained in common ECG recordings at high significance levels. The ECG patterns were included in 5/8 of our parameters for electrical axis, 6/8 for Q-wave, 7/8 for ST deviation and 5/8 for T-polarity based on two time series of correlation coefficients. We conclude that our spatiotemporal correlation approach provides a new tool for standardised analysis of cardiac mapping data such as MCG.

Skull Defects in Finite Element Head Models for Source Reconstruction from Magnetoencephalography Signals.

Magnetoencephalography (MEG) signals are influenced by skull defects. However, there is a lack of evidence of this influence during source reconstruction. Our objectives are to characterize errors in source reconstruction from MEG signals due to ignoring skull defects and to assess the ability of an exact finite element head model to eliminate such errors. A detailed finite element model of the head of a rabbit used in a physical experiment was constructed from magnetic resonance and co-registered computer tomography imaging that differentiated nine tissue types. Sources of the MEG measurements above intact skull and above skull defects respectively were reconstructed using a finite element model with the intact skull and one incorporating the skull defects. The forward simulation of the MEG signals reproduced the experimentally observed characteristic magnitude and topography changes due to skull defects. Sources reconstructed from measured MEG signals above intact skull matched the known physical locations and orientations. Ignoring skull defects in the head model during reconstruction displaced sources under a skull defect away from that defect. Sources next to a defect were reoriented. When skull defects, with their physical conductivity, were incorporated in the head model, the location and orientation errors were mostly eliminated. The conductivity of the skull defect material non-uniformly modulated the influence on MEG signals. We propose concrete guidelines for taking into account conducting skull defects during MEG coil placement and modeling. Exact finite element head models can improve localization of brain function, specifically after surgery.