Seizure Onset Localization in Focal Epilepsy using intracranial-EEG data and the Schrodinger Operator's Spectrum.

Epilepsy is a neurological disorder characterized by recurrent, unprovoked seizures that vary from short attention failure to convulsions. Despite its threats and limitations, existing medications target only specific types of seizures while up to 33% of epileptic conditions are drug-resistant. The best available treatment is surgical resection or neurostimulation and both require accurate localization of the Seizure Onset Zone. Its delineation is performed by analyzing neuronal activity by epileptologists, however, it is time-consuming and error-prone. Therefore, if the said zone could be located faster and more accurately, the seizure freedom of patients would be significantly enhanced. An effort within the field is aiming at developing computer-aided methods to assist medical experts and this starts with characterizing electrical neural activity. In the present paper, a new method for characterizing the epileptic intracranial EEG is proposed. The method is based on a semi-classical signal analysis (SCSA) method. Functional connectivity measures are used to compare patterns observed when feeding these measures with the raw time-series and when feeding them with SCSA features. The obtained results are undeniably promising for further investigation and improvement of the framework.Clinical relevance- The paper contributes to the design methods and algorithms to build reliable software solutions to assist medical experts in identifying Seizure Onset Zone in focal epilepsy.

EEG Epileptic Data Classification Using the Schrodinger Operator's Spectrum.

Epilepsy is a common brain disorder characterized by recurrent, unprovoked seizures which affects over 65 million people. Visual inspection of Electroencephalograms (EEG) is common for diagnosis; however, it requires time and expertise. Therefore, an accurate computer-aided epileptic seizure diagnosis system would be valuable. A new research tendency when tackling epileptic seizure detection tends towards minimizing human manual intervention by designing frameworks with autonomous feature engineering. In this optic, this paper proposes a new approach for EEG epileptic data classification. Features derived from the Semi-Classical Signal Analysis (SCSA) method, a quantum-inspired signal processing method well-suited for the characterization of pulse-shaped physiological signals, are proposed. In addition nonlinear dynamical features that proved efficient in characterizing nonlinear dynamics of neural activity have been extracted. Moreover, hyperparameters' optimization, correlation analysis and feature selection have been performed. The selected features are fed into five different machine learning classifiers. The performance of the proposed approach has been analyzed using Bonn university database. The results show that all classifiers yield a performance accuracy of 93% and above.Clinical relevance- The paper contributes to the design of methods and algorithms to build reliable software solutions to assist medical experts and reduce epilepsy disease's diagnosis time and errors.

Synergy of Small Antiviral Molecules on a Black-Phosphorus Nanocarrier: Machine Learning and Quantum Chemical Simulation Insights.

Favipiravir (FP) and Ebselen (EB) belong to a broad range of antiviral drugs that have shown active potential as medications against many viruses. Employing molecular dynamics simulations and machine learning (ML) combined with van der Waals density functional theory, we have uncovered the binding characteristics of these two antiviral drugs on a phosphorene nanocarrier. Herein, by using four different machine learning models (i.e., Bagged Trees, Gaussian Process Regression (GPR), Support Vector Regression (SVR), and Regression Trees (RT)), the Hamiltonian and the interaction energy of antiviral molecules in a phosphorene monolayer are trained in an appropriate way. However, training efficient and accurate models for approximating the density functional theory (DFT) is the final step in using ML to aid in the design of new drugs. To improve the prediction accuracy, the Bayesian optimization approach has been employed to optimize the GPR, SVR, RT, and BT models. Results revealed that the GPR model obtained superior prediction performance with an R2 of 0.9649, indicating that it can explain 96.49% of the data's variability. Then, by means of DFT calculations, we examine the interaction characteristics and thermodynamic properties in a vacuum and a continuum solvent interface. These results illustrate that the hybrid drug is an enabled, functionalized 2D complex with vigorous thermostability. The change in Gibbs free energy at different surface charges and temperatures implies that the FP and EB molecules are allowed to adsorb from the gas phase onto the 2D monolayer at different pH conditions and high temperatures. The results reveal a valuable antiviral drug therapy loaded by 2D biomaterials that may possibly open a new way of auto-treating different diseases, such as SARS-CoV, in primary terms.

A learning-based image processing approach for pulse wave velocity estimation using spectrogram from peripheral pulse wave signals: An in silico study.

Carotid-to-femoral pulse wave velocity (cf-PWV) is considered a critical index to evaluate arterial stiffness. For this reason, estimating Carotid-to-femoral pulse wave velocity (cf-PWV) is essential for diagnosing and analyzing different cardiovascular diseases. Despite its broader adoption in the clinical routine, the measurement process of carotid-to-femoral pulse wave velocity is considered a demanding task for clinicians and patients making it prone to inaccuracies and errors in the estimation. A smart non-invasive, and peripheral measurement of carotid-to-femoral pulse wave velocity could overcome the challenges of the classical assessment process and improve the quality of patient care. This paper proposes a novel methodology for the carotid-to-femoral pulse wave velocity estimation based on the use of the spectrogram representation from single non-invasive peripheral pulse wave signals [photoplethysmography (PPG) or blood pressure (BP)]. This methodology was tested using three feature extraction methods based on the semi-classical signal analysis (SCSA) method, the Law's mask for texture energy extraction, and the central statistical moments. Finally, each feature method was fed into different machine learning models for the carotid-to-femoral pulse wave velocity estimation. The proposed methodology obtained an $R2\geq0.90$ for all the peripheral signals for the noise-free case using the MLP model, and for the different noise levels added to the original signal, the SCSA-based features with the MLP model presented an $R2\geq0.91$ for all the peripheral signals at the level of noise. These results provide evidence of the capacity of spectrogram representation for efficiently assessing the carotid-to-femoral pulse wave velocity estimation using different feature methods. Future work will be done toward testing the proposed methodology for in-vivo signals.

A novel method for Magnetic Resonance Spectroscopy lipid signal suppression using Semi-classical signal analysis and Bidirectional Long short-term memory.

Magnetic resonance spectroscopy (MRS) is a non-invasive method that enables the analysis and quantification of brain metabolites, which provide useful information about the neuro-biological substrates of brain function. Lactate plays a pivotal role in the diagnosis of various brain diseases. However, accurate lactate quantification is generally difficult to achieve due to the presence of large lipid peaks resonating at a similar spectral position. To overcome this problem several techniques have been proposed. However, most of them suffer from lactate signal loss or poor lipid peak removal. In this paper, a novel method for lipid suppression for MRS signal is proposed. The method combines a semi-classical signal analysis method and a bidirectional long short term memory technique. The method is validated using simulated data that mimics real MRS data.

A Multiple Linear Regression Model for Carotid-to-Femoral Pulse Wave Velocity Estimation Based on Schrodinger Spectrum Characterization.

In this paper, a multiple linear regression model for estimating the Carotid-to-femoral pulse wave velocity (cf-PWV) from a single non-invasive peripheral pulse wave, namely blood pressure or photoplethysmography, is proposed. The training and testing datasets were extracted from in-silico, publicly available, pulse waves and hemodynamics data. The proposed model relies on a preprocessing and features extraction steps, which are performed using a semi-classical signal analysis (SCSA) method. The obtained results provide more evidence for the feasibility of machine learning and the SCSA method as a smart tool for the efficient assessment of the cf-PWV.

Human Hypertension Blood Flow Model Using Fractional Calculus.

The blood flow dynamics in human arteries with hypertension disease is modeled using fractional calculus. The mathematical model is constructed using five-element lumped parameter arterial Windkessel representation. Fractional-order capacitors are used to represent the elastic properties of both proximal large arteries and distal small arteries measured from the heart aortic root. The proposed fractional model offers high flexibility in characterizing the arterial complex tree network. The results illustrate the validity of the new model and the physiological interpretability of the fractional differentiation order through a set of validation using human hypertensive patients. In addition, the results show that the fractional-order modeling approach yield a great potential to improve the understanding of the structural and functional changes in the large and small arteries due to hypertension disease.

Parameter Sensitivity and Experimental Validation for Fractional-Order Dynamical Modeling of Neurovascular Coupling.

Goal: Modeling neurovascular coupling is very important to understand brain functions, yet challenging due to the complexity of the involved phenomena. An alternative approach was recently proposed where the framework of fractional-order modeling is employed to characterize the complex phenomena underlying the neurovascular. Due to its nonlocal property, a fractional derivative is suitable for modeling delayed and power-law phenomena. Methods: In this study, we analyze and validate a fractional-order model, which characterizes the neurovascular coupling mechanism. To show the added value of the fractional-order parameters of the proposed model, we perform a parameter sensitivity analysis of the fractional model compared to its integer counterpart. Moreover, the model was validated using neural activity-CBF data related to both event and block design experiments that were acquired using electrophysiology and laser Doppler flowmetry recordings, respectively. Results: The validation results show the aptitude and flexibility of the fractional-order paradigm in fitting a more comprehensive range of well-shaped CBF response behaviors while maintaining a low model complexity. Comparison with the standard integer-order models shows the added value of the fractional-order parameters in capturing various key determinants of the cerebral hemody-namic response, e.g., post-stimulus undershoot. This investigation authenticates the ability and adaptability of the fractional-order framework to characterize a wider range of well-shaped cerebral blood flow responses while preserving low model complexity through a series of unconstrained and constrained optimizations. Conclusions: The analysis of the proposed fractional-order model demonstrates that the proposed framework yields a powerful tool for a flexible characterization of the neurovascular coupling mechanism.

Automatic Detection of Epileptiform EEG Discharges based on the Semi-Classical Signal Analysis (SCSA) method.

In this paper we utilize a signal processing tool, which can help physicians and clinical researchers to automate the process of EEG epileptiform spike detection. The semi-classical signal analysis method (SCSA) is a data-driven signal decomposition method developed for pulse-shaped signal characterization. We present an algorithm framework to process and extract features from the patient's EEG recording by deriving the mathematical motivation behind SCSA and quantifying existing spike diagnosis criterion with it. The proposed method can help reduce the amount of data to manually analyse. We have tested our proposed algorithm framework with real data, which guarantees the method's statistical reliability and robustness.

Combining Machine Learning and Blind Estimation for Central Aortic Blood Pressure Reconstruction.

Central blood pressure is a vital signal that provides relevant physiological information about cardiovascular diseases risk factors. The standard clinical protocols for measuring these signals are challenging due to their invasive nature. This makes the estimation-based methods more convenient, however, they are usually not accurate as they fail to capture some important features of the central pressure waveforms. In this paper, we propose a novel data-driven approach that combines machine learning tools and cross-relation-based blind estimation methods to reconstruct the aortic blood pressure waves from the distorted peripheral pressure signals. Due to the lack of large real datasets, in this study, we utilize virtual pulse waves in-silico databases to train the machine learning models. The performance of the proposed approach is compared with the pure machine learning-based model and the cross-relation-based blind estimation approach. In both cases, the hybrid approach shows promising results as the root-mean-squared error has been reduced by 25% with regards to the pure machine learning method and by 40% compared to the cross-relation approach.

Towards Characterization of the Complex and Frequency-dependent Arterial Compliance based on Fractional-order Capacitor.

Arterial compliance is a vital determinant of the ventriculo-arterial coupling dynamic. Its variation is detrimental to cardiovascular functions and associated with heart diseases. Accordingly, assessment and measurement of arterial compliance are essential in the diagnosis and treatment of chronic arterial insufficiency. Recently, experimental and theoretical studies have recognized the power of fractional calculus to perceive viscoelastic blood vessel structure and biomechanical properties. This paper presents five fractional-order model representations to describe the dynamic relationship between the aortic blood pressure input and blood volume. Each configuration incorporates a fractional-order capacitor element (FOC) to lump the apparent arterial compliance's complex and frequency dependence properties. FOC combines both resistive and capacitive attributes within a unified component, which can be controlled through the fractional differentiation order factor, α. Besides, the equivalent capacitance of FOC is by its very nature frequency-dependent, compassing the complex properties using only a few numbers of parameters. The proposed representations have been compared with generalized integer-order models of arterial compliance. Both models have been applied and validated using different aortic pressure and flow rate data acquired from various species such as humans, pigs, and dogs. The results have shown that the fractional-order framework is able to accurately reconstruct the dynamic of the complex and frequency-dependent apparent compliance dynamic and reduce the complexity. It seems that this new paradigm confers a prominent potential to be adopted in clinical practice and basic cardiovascular mechanics research.

Fractional-order model representations of apparent vascular compliance as an alternative in the analysis of arterial stiffness: anin-silicostudy.

Objective. Recent studies have demonstrated the advantages of fractional-order calculus tools for probing the viscoelastic properties of collagenous tissue, characterizing the arterial blood flow and red cell membrane mechanics, and modeling the aortic valve cusp. In this article, we present novel lumped-parameter equivalent circuit models for apparent arterial compliance using a fractional-order capacitor (FOC). FOCs, which generalize capacitors and resistors, display a fractional-order behavior that can capture both elastic and viscous properties through a power-law formulation.Approach. The proposed framework describes the dynamic relationship between the blood-pressure input and the blood volume, using linear fractional-order differential equations.Main results. The results show that the proposed models present a reasonable fit with thein-silicodata of more than 4000 subjects. Additionally, strong correlations have been identified between the fractional-order parameter estimates and the central hemodynamic determinants as well as the pulse-wave velocity indexes.Significance. Therefore, the fractional-order-based paradigm for arterial compliance shows notable potential as an alternative tool in the analysis of arterial stiffness.

Schrödinger Spectrum Based PPG Features for the Estimation of the Arterial Blood Pressure.

In this paper, photoplethysmogram (PPG) features are combined with supervised machine learning algorithms to estimate arterial blood pressure (ABP). Three algorithms for the estimation of cuffless ABP using PPG signals are compared. Since PPG signals are measured non-invasively, this method guarantees an individuals comfort while not omitting important ABP information. The proposed framework predicts the ABP values by processing PPG signals with semi-classical signal analysis (SCSA) method, extracting several categories of features, which reflect the PPG signal morphology variations. Then, regression algorithms are selected for the ABP estimation. The proposed method is evaluated based on a virtual dataset with more than four thousand subjects and MIMIC II database with over eight thousand subjects for model training and testing. Mean average error (MAE) and standard deviation (STD) are evaluated for different machine learning algorithms during the prediction and estimation process. Multiple linear regression (MLR) meets the AAMI standard in terms of estimation accuracy, which proves that the ABP can be accurately estimated in a nonintrusive fashion. Given the easy implementation of the ABP estimation method, we regard that the proposed features and machine learning algorithms for the cuffless estimation of the ABP can potentially provide the means for mobile healthcare equipment to monitor the ABP continuously.

Blind Estimation of Central Blood Pressure Using Least-Squares with Mean Matching and Box Constraints.

Central aortic blood pressure (CABP) is a very-well recognized source of information to asses the cardiovascular system conditions. However, the clinical measurement protocol of this pulse wave is very intrusive and burdensome as it requires expert staff and complicated invasive settings. On the other hand, the measurement of peripheral blood pressure is much more straightforward and easy-to-get non-invasively. Several mathematical tools have been employed in the past few decades to reconstruct CABP waveforms from distorted peripheral pressure signals. More specifically, the cross-relation approach together with the widely used least-squares method, are shown to be effective as a way to estimate CABP waves. In this paper, we propose an improved cross-relation method that leverages the values of the diastolic and systolic pressures as box constraints. In addition, a mean-matching criterion is introduced to relax the need for the input and output mean values to be strictly equal. Using the proposed method, the root mean squared error is reduced by approximately 20% while the computational complexity is not significantly increased.

A Position Weight Matrix Feature Extraction Algorithm Improves Hand Gesture Recognition.

Recent advances in the biomedical field have generated a massive amount of data and records (signals) that are collected for diagnosis purposes. The correct interpretation and understanding of these signals present a big challenge for digital health vision. In this work, Quantization-based position Weight Matrix (QuPWM) feature extraction method for multiclass classification is proposed to improve the interpretation of biomedical signals. This method is validated on surface Electromyogram (sEMG) signals recognition for eight different hand gestures. The used CapgMyo dataset consists of high-density sEMG signals across 128 channels acquired from 9 intact subjects. Our pilot results show that an accuracy of up to 83% can be achieved for some subjects using a support vector machine classifier, and an average accuracy of 75% has been reached for all studied subjects using the CapgMyo dataset. The proposed method shows a good potential in extracting relevant features from different biomedical signals such as Electroencephalogram (EEG) and Magnetoencephalogram (MEG) signals.

QuPWM: Feature Extraction Method for Epileptic Spike Classification.

Epilepsy is a neurological disorder ranked as the second most serious neurological disease known to humanity, after stroke. Inter-ictal spiking is an abnormal neuronal discharge after an epileptic seizure. This abnormal activity can originate from one or more cranial lobes, often travels from one lobe to another, and interferes with normal activity from the affected lobe. The common practice for Inter-ictal spike detection of brain signals is via visual scanning of the recordings, which is a subjective and a very time-consuming task. Motivated by that, this article focuses on using machine learning for epileptic spikes classification in magnetoencephalography (MEG) signals. First, we used the Position Weight Matrix (PWM) method combined with a uniform quantizer to generate useful features from time domain and frequency domain through a Fast Fourier Transform (FFT) of the framed raw MEG signals. Second, the extracted features are fed to standard classifiers for inter-ictel spikes classification. The proposed technique shows great potential in spike classification and reducing the feature vector size. Specifically, the proposed technique achieved average sensitivity up to 87% and specificity up to 97% using 5-folds cross-validation applied to a balanced dataset. These samples are extracted from nine epileptic subjects using a sliding frame of size 95 samples-points with a step-size of 8 sample-points.

Fractional-Order SEIQRDP Model for Simulating the Dynamics of COVID-19 Epidemic.

Goal: Coronavirus disease (COVID-19) is a contagious disease caused by a newly discovered coronavirus, initially identified in the mainland of China, late December 2019. COVID-19 has been confirmed as a higher infectious disease that can spread quickly in a community population depending on the number of susceptible and infected cases and also depending on their movement in the community. Since January 2020, COVID-19 has reached out to many countries worldwide, and the number of daily cases remains to increase rapidly. Method: Several mathematical and statistical models have been developed to understand, track, and forecast the trend of the virus spread. Susceptible-Exposed-Infected-Quarantined-Recovered-Death-Insusceptible (SEIQRDP) model is one of the most promising epidemiological models that has been suggested for estimating the transmissibility of the COVID-19. In the present study, we propose a fractional-order SEIQRDP model to analyze the COVID-19 pandemic. In the recent decade, it has proven that many aspects in many domains can be described very successfully using fractional order differential equations. Accordingly, the Fractional-order paradigm offers a flexible, appropriate, and reliable framework for pandemic growth characterization. In fact, due to its non-locality properties, a fractional-order operator takes into consideration the variables' memory effect, and hence, it takes into account the sub-diffusion process of confirmed and recovered cases. Results-The validation of the studied fractional-order model using real COVID-19 data for different regions in China, Italy, and France show the potential of the proposed paradigm in predicting and understanding the pandemic dynamic. Conclusions: Fractional-order epidemiological models might play an important role in understanding and predicting the spread of the COVID-19, also providing relevant guidelines for controlling the pandemic.

Assessment of Fractional-Order Arterial Windkessel as a Model of Aortic Input Impedance.

Goal: Fractional-order Windkessel model is proposed to describe the aortic input impedance. Compared with the conventional arterial Windkessel, the main advantage of the proposed model is the consideration of the viscoelastic nature of the arterial wall using the fractional-order capacitor (FOC). Methods: The proposed model, along with the standard two-element Windkessel, three-element Windkessel, and the viscoelastic Windkessel models, are assessed and compared using in-silico data. Results: The results show that the fractional-order model fits better the moduli of the aortic input impedance and fairly approximates the phase angle. In addition, by its very nature, the pseudo-capacitance of FOC makes the proposed model's dynamic compliance complex and frequency-dependent. Conclusions: The analysis of the proposed fractional-order model indicates that fractional-order impedance yields a powerful tool for a flexible characterization of the arterial hemodynamics.

Sparse Reconstruction of Glucose Fluxes Using Continuous Glucose Monitors.

A new technique for estimating postprandial glucose flux profiles without the use of glucose tracers is proposed. A sparse vector space representation is first found for the space of plausible glucose flux profiles using sparse encoding. A Lasso formulation is then used to estimate the glucose fluxes that combines (1) known patient model parameters; (2) the vector space of plausible glucose flux profiles; (3) continuous glucose monitor measurements taken during the meal; (4) amount of insulin injected; (5) amount of meal carbohydrates; and (6) an estimate of the initial conditions. Three glucose fluxes are then estimated, namely; glucose rate of appearance from the intestine; endogenous glucose production from the liver; insulin dependent glucose utilization; and other important state variables. The simulation results show that the technique is capable of estimating the glucose fluxes with high accuracy, even for complex meal scenarios. The experimental results indicate that the technique is capable of reproducing the triple tracer measurements for three T1DM undergoing the triple tracer protocol while estimating the missing measurements for a certain model parameter selection.

Hybrid model for efficient prediction of poly(A) signals in human genomic DNA.

Polyadenylation signals (PAS) are found in most protein-coding and some non-coding genes in eukaryotes. Their accurate recognition improves understanding gene regulation mechanisms and recognition of the 3'-end of transcribed gene regions where premature or alternate transcription ends may lead to various diseases. Although different methods and tools for in-silico prediction of genomic signals have been proposed, the correct identification of PAS in genomic DNA remains challenging due to a vast number of non-relevant hexamers identical to PAS hexamers. In this study, we developed a novel method for PAS recognition. The method is implemented in a hybrid PAS recognition model (HybPAS), which is based on deep neural networks (DNNs) and logistic regression models (LRMs). One of such models is developed for each of the 12 most frequent human PAS hexamers. DNN models appeared the best for eight PAS types (including the two most frequent PAS hexamers), while LRM appeared best for the remaining four PAS types. The new models use different combinations of signal processing-based, statistical, and sequence-based features as input. The results obtained on human genomic data show that HybPAS outperforms the well-tuned state-of-the-art Omni-PolyA models, reducing the classification error for different PAS hexamers by up to 57.35% for 10 out of 12 PAS types, with Omni-PolyA models being better for two PAS types. For the most frequent PAS types, 'AATAAA' and 'ATTAAA', HybPAS reduced the error rate by 35.14% and 34.48%, respectively. On average, HybPAS reduces the error by 30.29%. HybPAS is implemented partly in Python and in MATLAB available at https://github.com/EMANG-KAUST/PolyA_Prediction_LRM_DNN.