Category: Frontiers in Neuroinformatics

Mental disorders are a significant concern in contemporary society, with a pressing need to identify biological markers. Long-range temporal correlations (LRTC) of brain rhythms have been widespread in clinical cohort studies, especially in major depressive disorder (MDD). However, research on LRTC in obsessive-compulsive disorder (OCD) is severely limited. Given the high co-occurrence of OCD and MDD, we conducted a comparative LRTC investigation. We assumed that the LRTC patterns will allow us to compare measures of brain cortical balance of excitation and inhibition in OCD and MDD, which will be useful in the area of differential diagnosis.

In this study, we used the 64-channel resting state EEG of 29 MDD participants, 26 OCD participants, and a control group of 37 volunteers. Detrended fluctuation analyzes was used to assess LRTC.

Our results indicate that all scaling exponents of the three subject groups exhibited persistent LRTC of EEG oscillations. There was a tendency for LRTC to be higher in disorders than in controls, but statistically significant differences were found between the OCD and control groups in the entire frontal and left parietal occipital areas, and between the MDD and OCD groups in the middle and right frontal areas.

We believe that these results indicate abnormalities in the inhibitory and excitatory neurotransmitter systems, predominantly affecting areas related to executive functions.

Pain assessment is extremely important in patients unable to communicate and it is often done by clinical judgement. However, assessing pain using observable indicators can be challenging for clinicians due to the subjective perceptions, individual differences in pain expression, and potential confounding factors. Therefore, the need for an objective pain assessment method that can assist medical practitioners. Functional near-infrared spectroscopy (fNIRS) has shown promising results to assess the neural function in response of nociception and pain. Previous studies have explored the use of machine learning with hand-crafted features in the assessment of pain.

In this study, we aim to expand previous studies by exploring the use of deep learning models Convolutional Neural Network (CNN), Long Short-Term Memory (LSTM), and (CNN-LSTM) to automatically extract features from fNIRS data and by comparing these with classical machine learning models using hand-crafted features.

The results showed that the deep learning models exhibited favourable results in the identification of different types of pain in our experiment using only fNIRS input data. The combination of CNN and LSTM in a hybrid model (CNN-LSTM) exhibited the highest performance (accuracy = 91.2%) in our problem setting. Statistical analysis using one-way ANOVA with Tukey's (post-hoc) test performed on accuracies showed that the deep learning models significantly improved accuracy performance as compared to the baseline models.

Overall, deep learning models showed their potential to learn features automatically without relying on manually-extracted features and the CNN-LSTM model could be used as a possible method of assessment of pain in non-verbal patients. Future research is needed to evaluate the generalisation of this method of pain assessment on independent populations and in real-life scenarios.

The endocannabinoid (eCB) system is named after the discovery that endogenous cannabinoids bind to the same receptors as the phytochemical compounds found in Cannabis. While endogenous cannabinoids include anandamide (AEA) and 2-arachidonoylglycerol (2-AG), exogenous phytocannabinoids include Δ-9 tetrahydrocannabinol (THC) and cannabidiol (CBD). These compounds finely tune neurotransmission following synapse activation, via retrograde signaling that activates cannabinoid receptor 1 (CB1R) and/or transient receptor potential cation channel subfamily V member 1 (TRPV1). Recently, the eCB system has been linked to several neurological diseases, such as neuro-ocular abnormalities, pain insensitivity, migraine, epilepsy, addiction and neurodevelopmental disorders. In the current study, we aim to: (i) highlight a potential link between the eCB system and neurological disorders, (ii) assess if THC exposure alters the expression of eCB-related genes, and (iii) identify evolutionary-conserved residues in CB1R or TRPV1 in light of their function.

To address this, we used several bioinformatic approaches, such as transcriptomic (Gene Expression Omnibus), protein–protein (STRING), phylogenic (BLASTP, MEGA) and structural (Phyre2, AutoDock, Vina, PyMol) analyzes.

Using RNA sequencing datasets, we did not observe any dysregulation of eCB-related transcripts in major depressive disorders, bipolar disorder or schizophrenia in the anterior cingulate cortex, nucleus accumbens or dorsolateral striatum. Following in vivo THC exposure in adolescent mice, GPR55 was significantly upregulated in neurons from the ventral tegmental area, while other transcripts involved in the eCB system were not affected by THC exposure. Our results also suggest that THC likely induces neuroinflammation following in vitro application on mice microglia. Significant downregulation of TPRV1 occurred in the hippocampi of mice in which a model of temporal lobe epilepsy was induced, confirming previous observations. In addition, several transcriptomic dysregulations were observed in neurons of both epileptic mice and humans, which included transcripts involved in neuronal death. When scanning known interactions for transcripts involved in the eCB system (n = 13), we observed branching between the eCB system and neurophysiology, including proteins involved in the dopaminergic system. Our protein phylogenic analyzes revealed that CB1R forms a clade with CB2R, which is distinct from related paralogues such as sphingosine-1-phosphate, receptors, lysophosphatidic acid receptors and melanocortin receptors. As expected, several conserved residues were identified, which are crucial for CB1R receptor function. The anandamide-binding pocket seems to have appeared later in evolution. Similar results were observed for TRPV1, with conserved residues involved in receptor activation.

The current study found that GPR55 is upregulated in neurons following THC exposure, while TRPV1 is downregulated in temporal lobe epilepsy. Caution is advised when interpreting the present results, as we have employed secondary analyzes. Common ancestors for CB1R and TRPV1 diverged from jawless vertebrates during the late Ordovician, 450 million years ago. Conserved residues are identified, which mediate crucial receptor functions.

The willingness to trust predictions formulated by automatic algorithms is key in a wide range of domains. However, a vast number of deep architectures are only able to formulate predictions without associated uncertainty.

In this study, we propose a method to convert a standard neural network into a Bayesian neural network and estimate the variability of predictions by sampling different networks similar to the original one at each forward pass.

We combine our method with a tunable rejection-based approach that employs only the fraction of the data, i.e., the share that the model can classify with an uncertainty below a user-set threshold. We test our model in a large cohort of brain images from patients with Alzheimer's disease and healthy controls, discriminating the former and latter classes based on morphometric images exclusively.

We demonstrate how combining estimated uncertainty with a rejection-based approach increases classification accuracy from 0.86 to 0.95 while retaining 75% of the test set. In addition, the model can select the cases to be recommended for, e.g., expert human evaluation due to excessive uncertainty. Importantly, our framework circumvents additional workload during the training phase by using our network “turned into Bayesian” to implicitly investigate the loss landscape in the neighborhood of each test sample in order to determine the reliability of the predictions.

We believe that being able to estimate the uncertainty of a prediction, along with tools that can modulate the behavior of the network to a degree of confidence that the user is informed about (and comfortable with), can represent a crucial step in the direction of user compliance and easier integration of deep learning tools into everyday tasks currently performed by human operators.

We aimed to build a tool which facilitates manual labeling of sleep slow oscillations (SOs) and evaluate the performance of traditional sleep SO detection algorithms on such a manually labeled data set. We sought to develop improved methods for SO detection.

SOs in polysomnographic recordings acquired during nap time from ten older adults were manually labeled using a custom built graphical user interface tool. Three automatic SO detection algorithms previously used in the literature were evaluated on this data set. Additional machine learning and deep learning algorithms were trained on the manually labeled data set.

Our custom built tool significantly decreased the time needed for manual labeling, allowing us to manually inspect 96,277 potential SO events. The three automatic SO detection algorithms showed relatively low accuracy (max. 61.08%), but results were qualitatively similar, with SO density and amplitude increasing with sleep depth. The machine learning and deep learning algorithms showed higher accuracy (best: 99.20%) while maintaining a low prediction time.

Accurate detection of SO events is important for investigating their role in memory consolidation. In this context, our tool and proposed methods can provide significant help in identifying these events.