Loading...

LabNIng

Neuroengineering Laboratory at National University of San Martin

LabNIng

Neuroengineering Laboratory at National University of San Martin

LabNIng

Neuroengineering Laboratory at National University of San Martin

Welcome

LabNIng is the Neuroengineering Laboratory at National University of San Martin. We are a group of medics, engineerings and biologists with passion for neuroscience. Our goal is to solve relevant neuroscience problems using engineering tools. We apply techniques from theoretical and experimental neuroscience as well as clinical research to find answers to basic questions and design disruptive technologies that improve the quality of life of people with disabilities or disease.

About Us

Daniela Andres

I am a Medical Doctor and Doctor of Science.

My main interests are complex mathematics applied to the study of physiological systems, and how to use this to solve medical problems.

Competences: computational modeling, complex systems, non-linear mathematics, physiology, medicine, movement disorders.

Gianfranco Bianchi

I am a Biomedical Engineer and Electronic Technician.

I’m interested in research and development, and how to apply engineering techniques to solve health problems.

Competences: Software and hardware development for biomedical signals acquisition and processing. Embedded electronic firmware development, electronic signals control and processing. PCB design and development with CAD software.

Oscar Filevich

I am a MSc in Biology and a Doctor of Science in Chemistry.

I’m interested in fundamental properties of systems built with neurons, because they can help us understand complex mechanisms, like perception, information processing, locomotion, pathologies and social phenomena.

I like developing tools to study these mechanisms.

Andrea Cerminati

I am an advanced student of Biomedical Engineering.

I am interested in the development, design and application of new technologies in order to improve medical care.

At LabNIng, I am part of an interdisciplinary team that seeks to develop a software prototype that allows vascular turbulence to be quantified from EcoDoppler studies.

Teresa Politi

I am a cardiologist, specialist in statistics and perfected in vascular and cardiac ultrasound.

Currently, I am a doctoral candidate at the University of Buenos Aires. She collaborated with the LabNIng in studies related to Medicine, especially in the area of cardiovascular images.

Camila Reinaldo

I am an advanced student of Biomedical Engineering.

Currently, I am doing my final integrative project in LabNing, which consists of the design and development of a system for processing signals for accelerometry and clinical information for patients with Parkinson's Disease.

I also work as a Product Specialist in a health technology provider company.

Former members:

Biomedical Eng. Federico Nanni (ITBA)

Biomedical Eng. Alejandro Torres Valencia (Universidad Tecnológica de Valencia, Colombia)

Bch. Sc. Vasco Duarte da Costa (FHNW, Basel, Suiza)

Josefina Bompensieri (UNSAM)

Mariano Paladino (UNSAM)

Gustavo Vinci (UNSAM)

Sebastián Villafañe (UNSAM - Di Tella)

Projects

Participants: Sebastián Villafañe, Oscar Filevich, Daniela Andres

Plug & play software developed at our lab to classify spikes and isolate single cell activity from multi-neuronal microelectrode recordings. We use threshold detection, wavelet transformations and a genetic algorithm for the classification of spikes.

Download software Download dataset

Participants: Daniela Andres, Oscar Filevich, Sebastián Villafañe.

Objective: To understand and characterize information coding by the nervous system in a quantitative and formal manner. To model neuronal activity and information transmission in nervous tissue in the spatio-temporal domain. To use this knowledge in the development of applications and technology that offers new solutions in neurology and neurophysiology.



Relevant publications:

D. S. Andres On the motion of spikes: turbulent-like neuronal activity in the human basal ganglia. Frontiers in Human Neuroscience doi: 10.3389/fnhum.2018.00429, 2018.


Computer simulations. Left column: Time evolution of the velocity of spikes, u(x,t), as the diffusion coefficient δ increases (from top to bottom), with time on the vertical axis and space on the horizontal axis. White areas represent the parts of the integration domain where the module of the velocity of spikes is below an arbitrary limit (108), in opposition to black areas, where it is higher than this limit. As the diffusion coefficient increases, white areas are enlarged, as the total velocity diminishes across the integration domain. Middle column: Sample temporal multifractal spectra ζτ(q) obtained from temporal structure functions of increasing order, at fixed spatial points. Non-linearity indicates temporal multifractality. Right column: Sample spatial multifractal spectra ζx(q) obtained from spatial structure functions of increasing order, at fixed times. Non-linearity indicates spatial multifractality.



D.S. Andres, O. Darbin. Complex dynamics in the basal ganglia: health and disease beyond the motor system. The Journal of Neuropsychiatry and Clinical Neurosciences doi: 10.1176/appi.neuropsych.17020039, 2018.

F. Nanni, D.S. Andres. Structure function revisited: a simple tool for complex analysis of neuronal activity. Frontiers in Human Neuroscience doi: 10.3389/fnhum.2017.00409, 2017


Transformation from a raw neuronal recording into a temporal structure function. (Upper) Sample raw extracellular microelectrode recording of neuronal activity. This recording was obtained from the entopeduncular nucleus of a healthy rat (technical details can be found in Andres et al., 2014a). The vertical axis indicates electric potential (mV) and the horizontal axis indicates time (s). The inlet at the lower right shows the whole recording, from which a zoom is shown in the bigger window. Individual spikes are marked with a red arrow. Once spikes are classified as belonging to a single neuron's activity, interspike intervals (ISI) are calculated as shown (ISI = time elapsed between the occurrence of a spike and the next). (Middle) Sample time series of interspike intervals, obtained from a neuronal recording like the one shown in the upper panel. The vertical axis indicates ISI duration (ms) and the horizontal axis indicates ISI number (position in the time series). Notice the high variability of the ISI, typical of complex systems. (Lower) Temporal structure function obtained from a time series of ISI like the one shown in the middle panel. The vertical axis is the value of the function S(τ) and the horizontal axis is the scale τ. In pallidal neurons it is common to observe a positive slope of the function at lower scales, followed by a breakpoint and a plateau at higher scales, also typical of complex systems. The double logarithmic scale helps visualization of smaller τ.


D.S. Andres, D.F. Cerquetti, M. Merello. Neural code alterations and abnormal time patterns in Parkinson's disease. Journal of Neural Engineering 12:026004 (9pp), 2015.



DS. Andres, F. Gomez, F.S. Ferrari, D.F. Cerquetti, M. Merello, R. Viana, R. Stoop. Multiple-time-scale framework for understanding the progression of Parkinson's disease. Physical Review E 90:062709, 2014

D.S. Andres, D.F. Cerquetti, M. Merello, R. Stoop. Neuronal entropy depends on the level of alertness in the parkinsonian Globus Pallidus in vivo. Frontiers in Neurology, 5, 96:1-9, 2014

D.S. Andres, D.F. Cerquetti, M. Merello. Finite dimensional structure of the GPi discharge in patients with Parkinson’s disease. International Journal of Neural Systems 21(3): 175-186, 2011

D.S. Andres, D.F. Cerquetti, M. Merello. Turbulence in Globus pallidum neurons in patients with Parkinson's disease: Exponential decay of the power spectrum. Journal of Neuroscience Methods 197(1): 14-20, 2011

Participants: Oscar Filevich, Daniela Andres.

Objective: employing neuronal cultures with different degrees of complexity, like pure neurons in planar cultures, neurons + glia feeder, neurospheres and brain organoids, we try to get information about emergent properties, which depend on the number of cells in culture, considering patterns of expression of channels and receptors.

We use surface microelectrodes and genetically-encoded Ca++ imaging to record and characterize neuronal activity. We also use photoliberation of Caged-Compounds developed at Etchenique’s lab, FCEN, UBA (www.neuro.qi.fcen.uba.ar) to either stimulate or inhibit activity with high temporal and spatial precision.

Which properties of an intact brain can be recovered in these archaic constructions?

Relevant publications:

A caged nicotine with nanosecond range kinetics and visible light sensitivity

A Visible-Light-Sensitive Caged Serotonin

Ruthenium polypyridyl phototriggers - from beginnings to perspectives

Fast optical pH manipulation and imaging

Participants: Gianfranco Bianchi, Camila Reinaldo, Daniela Andrés

Explanation: Parkinson’s disease is a neurodegenerative disorder with complex symptoms, which makes diagnosis extremely difficult.

Objective: To create technology that brings medical solutions to people who need them.

Advances: This project consists of the creation of an integral system for quantitative diagnosis and follow-up of patients with movement disorders, like Parkinson’s disease, ataxia and Huntington’s chorea. We measure movement with wearable devices and process information with mobile applications, to gain objective measures of movement. This helps non-specialized medical centers to evaluate the motor state of patients, transmitting the results to be evaluated by experts in the field.

Presentations: FENS 2018, Berlin “Quantitative diagnosis of Parkinson´s disease based on scale invariance of acceleration signals”

Participants: Daniela Andres, Gustavo Vinci, Andrea Cerminati, Josefina Bompensieri y Teresa Politi

Advances:

Automatic detection of ischemia from a single ECG channel
Myocardial ischemia is the pathology due to lack of oxygen in the heart tissue, whose most serious expression is myocardial infarction. Using the public Physionet database, we reviewed and classified electrocardiograms to obtain representative isolated cardiac cycles from healthy and ischemic cases. We apply our own cardiac cycle detection algorithm, and use them to train a neural network capable of detecting ischemia from a single ECG cycle. We obtained an accuracy of 86%. This type of analysis can be applied to the early detection of ischemia, in stages in which myocardial infarction can be prevented or treated without leaving sequelae.

Turbulence quantification in color Doppler of neck vessels
Vascular echo-Doppler studies were obtained with a Philips ClearVue 650 ultrasound equipment and Philips Active Array 12-4 MHz linear transducer L12-4, under the carotid artery preset. Images were taken of 50 patients with a median age of 63 [60-67] years, the majority being male (65%) and presenting with carotid atherosclerotic plaques (72%). From the image delivered by the team, the color reference was extracted and processed, breaking it down into its RGB values. The color was thresholded to extract the region of the image corresponding to the Doppler signal. Turbulent areas were defined based on the presence of aliasing, understood as those regions in which an apparent change in the direction of blood flow is observed in high-velocity areas. These areas appear with values ​​at the upper limits of the forward and backward velocity scales. To delimit them and define the turbulent areas, the color dispersion graph of each frame was analyzed in cylindrical HSV coordinates. The regions of interest were found by optimizing the angle (corresponding to the hue or hue) capable of capturing the appropriate density of aliasing zones. This angle functions as a critical parameter of the algorithm.
Once the turbulent zones were defined, a box-counting method was implemented to measure their fractal dimension frame by frame, that is, to measure the coverage of the plane by turbulent zones. The box-counting method is a traditional and robust method to calculate the fractal dimension from images. The dimension D is calculated as the slope of a linear regression on a double logarithmic scale of the relationship between the number of occupied boxes N, and the scale s, that is, between mass and scale (Haudorff-Besicovitch dimension):

log⁡〖(N)∝-D∙log⁡(s)〗. (1)

The coefficient R2 of the linear regression works as the second critical parameter of the algorithm, allowing to decide which regressions are accepted and which ones are discarded, concluding which frames do or do not have a quantifiable fractal dimension.
Our analysis allowed us to obtain images with turbulent zones with dimension D in about 400 frames per study. The geometry of these turbulent regions turned out to be monofractal, that is, D=constant. Figure 1 shows the results obtained for a representative patient with obstructive atherosclerotic plaque and a representative patient with normal Doppler ultrasound. D fluctuates between 0 and 1, oscillating smoothly with the cardiac cycle, with higher values in patients with obstructive atherosclerotic plaque.



Figure 1. Evaluation of the fractal dimension in the Doppler echocardiography of neck vessels. A. Color scatter plot in HSV cylindrical coordinates for a carotid Doppler ultrasound frame of representative patients with normal carotid Doppler ultrasound (left) and with obstructive atherosclerotic plaque (right). The marked angle (acute angle between black lines) shows the area defined as aliasing or turbulence, at the upper limits of the forward and backward velocity scales (blue and red, respectively). In the patient with obstruction, a higher density of points can be observed in the angle of interest. B. Fractal dimension D as a function of frame number n on carotid Doppler ultrasound of the representative patients shown above (left carotid Doppler normal, right with obstructive atherosclerotic plaque - see text in section 2). Higher values of D are observed in the patient with obstruction.

Participants: Daniela Andres, Gianfranco Bianchi, Sebastián Villafañe, Mariano Paladino

Start-Up originated at LabNIng, dedicated to the development of medical technology.

www.ideme.com.ar

Collaborations

Movement Disorders Group at FLENI Institute, directed by Dr. Marcelo Merello
Prof. Simone Hemm-Ode, FHNW, Basel, Suiza
Roberto Etchenique
Arancha del Campo

Contact

Get in contact with us

Documents

Download information in PDF