EEG alpha and theta time-frequency structure during a written mathematical task.

Since the first electroencephalogram (EEG) was obtained, there have been many possibilities to use it as a tool to access brain cognitive dynamics. Mathematical (Math) problem solving is one of the most important cortical processes, but it is still far from being well understood. EEG is an inexpensive and simple indirect measure of brain operation, but only recently has low-cost equipment (mobile EEG) allowed sophisticated analyses in non-clinical settings. The main purpose of this work is to study EEG activation during a Math task in a realistic environment, using mobile EEG. A matching pursuit (MP)-based signal analysis technique was employed, since MP properties render it a priori suitable to study induced EEG activity over long time sequences, when it is not tightly locked to a given stimulus. The study sample comprised sixty healthy volunteers. Unlike the majority of previous studies, subjects were studied in a sitting position with their eyes open. They completed a written Math task outside the EEG lab, wearing a mobile EEG device (EPOC+). Theta [4 Hz-7.5 Hz], alpha (7.5 Hz-13 Hz] and 0.5 Hz micro-bands in the [0.5 Hz-20 Hz] range were studied with a low-density stochastic MP dictionary. Over 1-min windows, ongoing EEG alpha and theta activity was decomposed into numerous MP atoms with median duration around 3 s, similar to the duration of induced, time-locked activity obtained with event-related (des)synchronization (ERS/ERD) studies. Relative to Rest, there was lower right-side and posterior MP alpha atom/min during Math, whereas MP theta atom/min was significantly higher on anteriorly located electrodes, especially on the left side. MP alpha findings were particularly significant on a narrow range around 10 Hz-10.5 Hz, consistent with FFT alpha peak findings from ERS/ERD studies. With a streamlined protocol, these results replicate previous findings of EEG alpha and theta activation obtained during Math tasks with different signal analysis techniques and in different time frames. The efficient application to real-world, noisy EEG data with a low-resolution stochastic MP dictionary shows that this technique is very encouraging. These results provide support for studies of mathematical cognition with mobile EEG and matching pursuit.

NREM sleep alpha and sigma activity in Parkinson's disease: evidence for conflicting electrophysiological activity?

OBJECTIVES: Sleep EEG spectral patterns were investigated in eight newly diagnosed, non-depressed, non-demented, drug-naïve Parkinson's disease patients compared to nine controls. METHODS: Mean relative spectral power density calculated for 0.25 Hz frequency bins and for classical EEG frequency bands. RESULTS: Differences between patients and controls were most prominent in non-REM sleep, specially around 8.6 Hz (slow alpha), 12.5 Hz (fast alpha/slow sigma) and 15 Hz (fast sigma). Slow alpha showed lower p-values over frontal and occipital electrodes, whereas fast sigma activity was more important on central and parietal sites. Significantly increased NREM sleep alpha activity was found in left and right frontal (Mann-Whitney U=12,000, p=.021; U=14,000, p=.036), left and right central (U=14,000, p=.036), left parietal and left occipital (U=13,000, p=.027; U=15,000, p=.046) areas. Increased sigma activity was found in right frontal (U=14,000, p=.036), left central (U=12,000, p=.021), left and right parietal (U=12,000, p=.021; U=13,000, p=.027) and left occipital (U=15,000, p=.046) areas. CONCLUSIONS: Concomitantly increased scalp EEG alpha and sigma activity was found during NREM sleep in initial Parkinson's disease. SIGNIFICANCE: These non-REM sleep microstructure changes may represent evidence for altered electrophysiological mechanisms leading to sleep-wake instability in early disease stages.

Loss of sleep spindle frequency deceleration in Obstructive Sleep Apnea.

OBJECTIVE: Sleep spindles have been suggested as surrogates of thalamo-cortical activity. Internal frequency modulation within a spindle's time frame has been demonstrated in healthy subjects, showing that spindles tend to decelerate their frequency before termination. We investigated internal frequency modulation of slow and fast spindles according to Obstructive Sleep Apnea (OSA) severity and brain topography. METHODS: Seven non-OSA subjects and 21 patients with OSA contributed with 30min of Non-REM sleep stage 2, subjected to a Matching pursuit procedure with Gabor chirplet functions for automatic detection of sleep spindles and quantification of sleep spindle internal frequency modulation (chirp rate). RESULTS: Moderate OSA patients showed an inferior percentage of slow spindles with deceleration when compared to Mild and Non-OSA groups in frontal and parietal regions. In parietal regions, the percentage of slow spindles with deceleration was negatively correlated with global apnea-hypopnea index (rs=-0.519, p=0.005). DISCUSSION: Loss of physiological sleep spindle deceleration may either represent a disruption of thalamo-cortical loops generating spindle oscillations or some compensatory mechanism, an interesting venue for future research in the context of cognitive dysfunction in OSA. SIGNIFICANCE: Quantification of internal frequency modulation (chirp rate) is proposed as a promising approach to advance description of sleep spindle dynamics in brain pathology.

Triplet entropy analysis of hemagglutinin and neuraminidase sequences measures influenza virus phylodynamics.

The influenza virus has been a challenge to science due to its ability to withstand new environmental conditions. Taking into account the development of virus sequence databases, computational approaches can be helpful to understand virus behavior over time. Furthermore, they can suggest new directions to deal with influenza. This work presents triplet entropy analysis as a potential phylodynamic tool to quantify nucleotide organization of viral sequences. The application of this measure to segments of hemagglutinin (HA) and neuraminidase (NA) of H1N1 and H3N2 virus subtypes has shown some variability effects along timeline, inferring about virus evolution. Sequences were divided by year and compared for virus subtype (H1N1 and H3N2). The nonparametric Mann-Whitney test was used for comparison between groups. Results show that differentiation in entropy precedes differentiation in GC content for both groups. Considering the HA fragment, both triplet entropy as well as GC concentration show intersection in 2009, year of the recent pandemic. Some conclusions about possible flu evolutionary lines were drawn.

Topography-specific spindle frequency changes in obstructive sleep apnea.

BACKGROUND: Sleep spindles, as detected on scalp electroencephalography (EEG), are considered to be markers of thalamo-cortical network integrity. Since obstructive sleep apnea (OSA) is a known cause of brain dysfunction, the aim of this study was to investigate sleep spindle frequency distribution in OSA. Seven non-OSA subjects and 21 patients with OSA (11 mild and 10 moderate) were studied. A matching pursuit procedure was used for automatic detection of fast (≥13 Hz) and slow (<13 Hz) spindles obtained from 30 min samples of NREM sleep stage 2 taken from initial, middle and final night thirds (sections I, II and III) of frontal, central and parietal scalp regions. RESULTS: Compared to non-OSA subjects, Moderate OSA patients had higher central and parietal slow spindle percentage (SSP) in all night sections studied, and higher frontal SSP in sections II and III. As the night progressed, there was a reduction in central and parietal SSP, while frontal SSP remained high. Frontal slow spindle percentage in night section III predicted OSA with good accuracy, with OSA likelihood increased by 12.1%for every SSP unit increase (OR 1.121, 95% CI 1.013-1.239, p=0.027). CONCLUSIONS: These results are consistent with diffuse, predominantly frontal thalamo-cortical dysfunction during sleep in OSA, as more posterior brain regions appear to maintain some physiological spindle frequency modulation across the night. Displaying changes in an opposite direction to what is expected from the aging process itself, spindle frequency appears to be informative in OSA even with small sample sizes, and to represent a sensitive electrophysiological marker of brain dysfunction in OSA.

BacPP: bacterial promoter prediction--a tool for accurate sigma-factor specific assignment in enterobacteria.

Promoter sequences are well known to play a central role in gene expression. Their recognition and assignment in silico has not consolidated into a general bioinformatics method yet. Most previously available algorithms employ and are limited to σ70-dependent promoter sequences. This paper presents a new tool named BacPP, designed to recognize and predict Escherichia coli promoter sequences from background with specific accuracy for each σ factor (respectively, σ24, 86.9%; σ28, 92.8%; σ32, 91.5%; σ38, 89.3%, σ54, 97.0%; and σ70, 83.6%). BacPP is hence outstanding in recognition and assignment of sequences according to σ factor and provide circumstantial information about upstream gene sequences. This bioinformatic tool was developed by weighing rules extracted from neural networks trained with promoter sequences known to respond to a specific σ factor. Furthermore, when challenged with promoter sequences belonging to other enterobacteria BacPP maintained 76% accuracy overall.

Rules extraction from neural networks applied to the prediction and recognition of prokaryotic promoters.

Promoters are DNA sequences located upstream of the gene region and play a central role in gene expression. Computational techniques show good accuracy in gene prediction but are less successful in predicting promoters, primarily because of the high number of false positives that reflect characteristics of the promoter sequences. Many machine learning methods have been used to address this issue. Neural Networks (NN) have been successfully used in this field because of their ability to recognize imprecise and incomplete patterns characteristic of promoter sequences. In this paper, NN was used to predict and recognize promoter sequences in two data sets: (i) one based on nucleotide sequence information and (ii) another based on stability sequence information. The accuracy was approximately 80% for simulation (i) and 68% for simulation (ii). In the rules extracted, biological consensus motifs were important parts of the NN learning process in both simulations.

Quantifying chirp in sleep spindles.

Sleep spindles are considered as a marker of integrity for thalamo-cortical circuits. Recently, attention has been given to internal frequency variation in sleep spindles. In this study, a procedure based on matching pursuit with a Gabor-chirplet dictionary was applied in order to measure chirp rate in atoms representing sleep spindles, also categorized into negative, positive or zero chirp types. The sample comprised 707 EEG segments containing visual sleep spindles, labeled TP, obtained from nine healthy male volunteers (aged 20-34, average 24.6 y). Control datasets were 333 non-REM (NREM) sleep background segments and 287 REM sleep intervals, each with 16s duration. Analyses were carried out on the C3-A2 EEG channel. In TP and NREM groups, the proportion of non-null chirp types was non-random and total chirp distribution was asymmetrical towards negative values, in contrast to REM. Median negative chirp rate in the TP and NREM groups was significantly lower than in REM (-0.4 Hz/s vs -0.3 Hz/s, P < 0.05). Negative chirp atoms outnumbered positives by 50% in TP, while in NREM and REM, they were, respectively, only 22% and 12% more prevalent. TP negative chirp atoms were significantly higher in amplitude compared to positive or zero types. Considering individual subjects, 88.9% had a TP negative/positive chirp ratio above 1 (mean ± sd=1.64 ± 0.65). We propose there is increasing evidence, corroborated by the present study, favoring systematic measurement of sleep spindle chirp rate or internal frequency variation. Preferential occurrence of negatively chirping spindles is consistent with the hypothesis of electrophysiological modulation of neocortical memory consolidation.

Rules extraction from neural networks applied to the prediction and recognition of prokaryotic promoters

Promoters are DNA sequences located upstream of the gene region and play a central role in gene expression. Computational techniques show good accuracy in gene prediction but are less successful in predicting promoters, primarily because of the high number of false positives that reflect characteristics of the promoter sequences. Many machine learning methods have been used to address this issue. Neural Networks (NN) have been successfully used in this field because of their ability to recognize imprecise and incomplete patterns characteristic of promoter sequences. In this paper, NN was used to predict and recognize promoter sequences in two data sets: (i) one based on nucleotide sequence information and (ii) another based on stability sequence information. The accuracy was approximately 80 percent for simulation (i) and 68 percent for simulation (ii). In the rules extracted, biological consensus motifs were important parts of the NN learning process in both simulations.

Benchmarking matching pursuit to find sleep spindles.

The aim of this study is to evaluate performance of Matching Pursuit (MP) algorithm against visual analysis for automatic sleep spindle (SS) detection in a sample of sleep stages 2-4 and REM pertaining to nine healthy young subjects. MP-SS voltage, frequency and duration characteristics were investigated for the amplitude threshold (AT) that maximized yield between test sensitivity and specificity. Parameter distribution curves were also built for correctly detected (true positive) and false-positive events. For sleep stage 2, MP reached 80.6% sensitivity and specificity for an AT value of 58.8. For all stages together, 81.2% sensitivity and specificity were reached for an AT value of 46.6. Specificity curves were adequate for all stages; sensitivity was lower for S3+4. Sigma frequency range activity with atypical characteristics was detected within REM sleep. Prevalence indexes obtained with MP were much higher than visual prevalence indexes for all stages; similar voltage, frequency and duration distribution curves were obtained for true positive and false positive events. For this sample of young male healthy subjects, the free-ware MP algorithm showed satisfactory performance for SS detection in sleep stage 2 as reported earlier, acceptable performance in sleep stages 3+4, although with lowered sensitivity, and sigma frequency range activity within REM sleep that needs better understanding. Within NREM sleep, correspondence between the MP automatic and the visual method was supported.