Dimitrios Boursinos


pcb3 pcb2


The goal of this project was to build a cost efficient electromyograph (EMG) that can be connected to the computer through USB and be capable of supporting a large number of channels.


EMG is a technique for recording the electrical activity of muscles.  An EMG detects the electrical potential generated by muscle cells when these cells are electrically or neurologically activated. The signals can be analyzed to detect medical abnormalities, muscle activation level, or muscle recruitment order. The signals can also be used to analyze the biomechanics of human or animal movement. Another very common usage of EMG is for the control of prosthetic devices where EMG signals from the remaining muscles in a missing limb can be used for controlling a robotic prosthetic.

The reason I built this device is that we needed a cost efficient EMG acquisition hardware in order to use it with an existing simulator software and eventually improve the existing control algorithms and try it in an actual robotic arm.





Figure 1 – Channel circuit


In order to control a prosthetic device we need to have some EMG signals from the muscle activity of the amputee as system inputs. The activity is related to the voltage difference along a muscle.

A channel consists of 2 electrodes that measure the voltage at 2 points on the surface of the muscle, an instrumentation amplifier that calculates the difference between them, analog filters and an Analog-to-Digital Converter (ADC).

The hardware consists of the analog part which inputs the differential signal for all the channels and the digital part which samples the analog signals and sends the data to a computer at a high frequency of fs=1kHz. The user can stack up to 6 boards on top of each other to get a total of 42 channels.

Figure 1 shows the structure of each channel.

The active electodes ensure:

  1. low ouput impedance which means that the signal in the cable is stable against interference without the use of heavy shielding and guarding
  2. matched input impedance to the instrumentation amplifier.


The instrumentation amplifier calculates the difference between 2 electrodes. The average voltage of the 2 electrodes is injected back to the person by the Driven Right Leg (DRL) circuit. DRL is a feedback loop for the system that is used to reduce common mode interference.

The filtering stage is consisted of a 2nd order High Pass Filter (HPF) and a 2nd order Low Pass Filter (LPF). I chose the HPF cut-off frequency at 19.4 Hz in order to remove the drift of the signal and the movement transients. The signal is sensitive to disturbances caused by movement because of our use of dry electrodes but the aforementioned filter can remove most of the movement artifacts. The LPF cut off frequency, fc=152.7 Hz, was chosen lower than the Nyquist frequency (fs/2 = 500 Hz) as we observed that the power of the signal above fc was negligible for our project. This fc value is also more computational efficient for a real time application. Even though filters of larger order had slightly improved Signal-to-Noise Ratios (SNR), these results were not steady between all of the subjects so we chose a 2nd order filter for design simplicity and cost efficiency.

Finally, the ADC samples the signal and connects with the digital circuit through the Serial Peripheral Interface (SPI).



Figure 2 – High level circuit of the PCB


Figure 2 shows a high level view of the board. There are 7 channels, each of them is like the one described above. Moreover, there are 8 active electrodes (plus 1 passive for the DRL) which can be connected in bipolar or monopolar configuration using jumpers. This result to 4 bipolar or 7 monopolar channels. The average voltage of all of the electrodes is injected back to the user through the DRL circuit. An Arduino Due was used to connect the computer with the board as it inherently supports SPI and USB communications, is fast, is relatively cheap and easily programmed. Arduino Due accesses the ADCs consecutively and when it receives a sample for all the signals it sends them to the virtual port of the computer.



raw signal

                                                                           Figure 3 – EMG signal
                                                                             Figure 4 – 7 channels


The hardware was tested in a simulator software written in Matlab. The 7 channels (1 PCB) were more than enough in controlling 3 Degrees Of Freedom (DOF) when the electrodes were connected around the forearm.

Figure 3 shows the raw EMG signal of a single channel as it is recorded by Matlab when a subject extended his hand without any digital post processing. Figure 4 shows all of the signals of a single board when the subject extended his hand.

A maximum of 6 boards for recording different muscle groups can be used so that we are able to control more DOFs while we respect the system’s sampling frequency of 1 kHz.

Asif Rahman

Transcranial Direct Current Stimulation (tDCS) is investigated for a broad range of neuropsychiatric indications, various rehabilitation applications, and to modulate cognitive performance in diverse tasks. Specificity of tDCS refers broadly to the ability of tDCS to produce precise, as opposed to diffuse, changes in brain function. Practically, specificity of tDCS implies application-specific customization of protocols to maximize desired outcomes and minimize undesired effects. Especially given the simplicity of tDCS and the complexity of brain function, understanding the mechanisms leading to specificity is fundamental to the rational advancement of tDCS. We define the origins of specificity based on anatomical and functional factors. Anatomical specificity derives from guiding current to targeted brain structures. Functional specificity may derive from either activity-selectivity, where active neuronal networks are preferentially modulated by tDCS, or input-selectivity, where bias is applied to different synaptic inputs. Rational advancement of tDCS may require leveraging all forms of specificity.

Start by reading this: Bikson_Rahman_OriginsOfSpecificitytDCS_2013

Screen Shot 2013-10-01 at 6.20.30 PM

John M. Ettikkalayil


 A natural user interface is a system for human-computer interaction that the user operates through intuitive actions related to everyday, natural human behavior.  There are many devices and technologies currently out there in the market that can be considered a natural user interface, such as touch screen phones and motion sensing input devices like Microsoft’s Kinect.  In this video, we put forth the question, “what is the next natural user interface to change the way we can interact with computers and other devices?”  Some of us in neural engineering lab at the City College of New York (CCNY) believe that the next step in natural user interface is eye-control.  In the short demo video above, we explore an application for eye-control by piloting an AR.Drone with just our eyes.

Current Setup


  1. ITU Gaze Tracker (ITU Gaze Group)
  2. ARDroneForP5 (Shigeo Yoshida)
  3. Processing 2.0 revision 218
  4. UDP Processing Library


  1. Thorlabs Monochrome Camera (DCC1545M)
  2. Arecont Vision Varifocal Megapixel Lens 8-16mm (MPL8-16)
  3. Opteka HD² 37mm R72 720nm Infrared X-Ray IR Filter (OPTIR37)
  4. Giottos Mini Ball Head with Single Knob for Ball and Pan (MH1004)
  5. IR Lamp (IR010)
  6. Computer or Laptop with Windows 7
  7. Parrot AR.Drone version 1.0 (current setup doesn’t support version 2.0)

Custom Eye Tracker Cost*

No. Item Part Number Cost
1 Thorlabs High Resolution USB2.0 CMOS Camera, 1280 x 1024, Monochrome DCC1545M $345.00
2 Arecont Vision Varifocal Megapixel Lens (8-16mm) MPL8-16 $109.46
3 Opteka HD² 37mm R72 720nm Infrared X-Ray IR Filter OPTIR37 $14.95
4 Giottos Mini Ball Head with Single Knob for Ball and Pan MH1004 $12.99
5 Clover Electronics IR010 Night Vision IR Lights IR010 $34.89
TOTAL $517.29







*Does not include cost of AR.Drone

Preprint PDF: Rahman-Cellular-Effects-Acute-DCS-Somatic-Terminal

Asif Rahman

J Physiol. 2013 Mar 11.
Cellular Effects of Acute Direct Current Stimulation: Somatic and Synaptic Terminal Effects.

Rahman A, Reato D, Arlotti M, Gasca F, Datta A, Parra LC, Bikson M.
The City College of New York, CUNY;

Transcranial Direct Current Stimulation (tDCS) is a non-invasive brain stimulation technique to modulate cortical excitability. Although increased/decreased excitability under the anode/cathode electrode is nominally associated with membrane depolarization/hyperpolarization, which cellular compartments (somas, dendrites, axons and their terminals) mediate changes in cortical excitability remains unaddressed. Here we consider the acute effects of direct current stimulation on excitatory synaptic efficacy. Using multi-scale computational models and rat cortical brain slices we show: 1) Typical tDCS montages produce predominantly tangential (relative to the cortical surface) direction currents (4-12 times radial direction currents), even directly under electrodes. 2) Radial current flow (parallel to the somatodendritic axis) modulates synaptic efficacy consistent with somatic polarization, with depolarization facilitating synaptic efficacy. 3) Tangential current flow (perpendicular to the somatodendritic axis) modulates synaptic efficacy acutely (during stimulation) in an afferent pathway specific manner that is consistent with terminal polarization, with hyperpolarization facilitating synaptic efficacy. 4) Maximal polarization during uniform direct current stimulation is expected at distal (branch length is >3 times the membrane length constant) synaptic terminals, independent of and 2-3 times more susceptible than pyramidal neuron somas. We conclude that during acute direct current stimulation the cellular targets responsible for modulation of synaptic efficacy are concurrently somata and axon terminals, with the direction of cortical current flow determining the relative influence.



Figure 1. Multi-scale methods and outcome measures of uniform electric field directionality and effects. A1, Gyri-precise finite element models of current flow during tDCS indicate a uniform voltage gradient in cortical grey matter (GM) directly under the anode. A2, The induced electric field in the cortex can be decomposed into a radial component ( Ex ) that is parallel to the somatodendritic axis and a tangential component ( Ey ) that is orthogonal to the somatodendritic axis. A3, We quantified the relative occurrence of radial and tangential fields in cortical GM expressed as the ratio of the average of the field magnitude in the tangential direction to the average of the field magnitude in the radial direction ( Ey / Ex ). B1-2, The brain slice preparation was used to study the change in synaptic efficacy during a uniform radial or tangential field by recording evoked field potentials. The voltage gradient between parallel Ag/AgCl wires is superimposed on a schematic of a sagittal slice of the rat primary motor cortex. From the macroscopic to the mesoscopic scale we can approximate a uniform electric field along the length of a neuron (compare voltage gradients in A1 and B1). B3, The field EPSP provides a measure of synaptic efficacy through facilitation or inhibition of the response amplitude. C1, Compartment model simulations of morphologically reconstructed neocortical pyramidal neurons were used to provide a description of axon terminal polarization in a uniform electric field. C2, The polarization profile of a layer 5 pyramidal neuron in a radially directed uniform electric field indicates soma depolarization (red) corresponds to apical dendrite hyperpolarization (blue). Layer 2/3 neurons have a more complex polarization profile with long processes reaching maximal depolarization independent of the neuronal body. C3, Neurons in a uniform electric field directed tangential to the somatodendritic axis preferentially affects processes that are oriented along the tangential field.



Figure 2. Forward model of tDCS and HD-tDCS quantifying electric field direction metrics. During tDCS, current may be dominantly tangential (along the cortical surface) rather than radial, even in brain regions directly under the electrodes. A, MRI-derived finite element models of current distribution in a gyri-precise head model are used to quantify the relative occurrence of radial (normal to the cortical surface) and tangential (along the cortical surface) components of the electric field. Both conventional (top) and high definition (HD, bottom) tDCS montages produce radial ( Ex , normal to the cortical surface) and tangential current ( Ey , along the cortical surface) indicated by the global electric field distribution (V/m) across the head. In the HD-tDCS montage, current is focalized within the ring configuration (inset) with radial currents under the center electrode and tangential currents between the surround electrodes. Qualitative comparison of the electric field components indicate greater radial field magnitudes in the gyral wall and greater tangential field magnitudes in the gyral crown (compare insets). B, Regionally, the distribution of field component magnitudes indicate prevailing tangential currents under the anode, cathode and between electrodes as described by the ratio of tangential to radial field magnitude ( Ey / Ex ratio, see methods). However, most elements have both radial and tangential components, and the isolated highest electric fields are radial. The tangential and radial component for individual elements is shown for each sub-region in false color density plots, which show relative occurrence (relative density from absent (green) to maximal (red)). Axis histograms show relative distribution of elements with a given tangential or radial component electric field. Inset histograms describe the distribution of the % of elements in a region as a function of the normalized component magnitude (such that 1 indicates elements with dominant radial or tangential component). C, Cortical folding further influences the distribution of the electric field, therefore, sub-regional field component distributions are indicated for a gyral crown and wall. Tangential fields are dominant in magnitude in the gyral crown but are weaker in the walls where radial magnitudes are stronger, as observed in A.


Figure 3. Electrophysiology of direction-specific uniform DC electric fields in synaptic pathways of the rat motor cortex. A, Schematic of electrophysiology setup where uniform extracellular electric fields were generated in all experiments by passing constant current across parallel Ag-AgCl wires positioned in the bath across the slice. Activity was monitored in layer II/III or layer V with a glass microelectrode. An additional field electrode (REF) was positioned in an iso-potential to remove the uniform field artifact. Activity was evoked with a bipolar stimulating electrode (S1-S4) positioned 500 μm from the recording electrode in either layer II/III or layer V targeting one of four distinct synaptic pathways corresponding to different orientations of afferent axons: posterior horizontal layer II/III (S1), anterior horizontal layer II/III (S2), posterior horizontal layer V (S3), and vertical layer V to II/III (S4). B, Diagram summarizing the primary synaptic circuits in this study. Line thickness and diameter of the filled circles, which represent synapses, are correlated with the strength of the synaptic input. C, Schematic of the expected polarization in distinct synaptic pathways exposed to radial and tangential fields. Somas, dendrites, axons and axon terminals are depolarized (red), hyperpolarized (blue), or not affected (black) by DC fields. D, Characteristic fEPSP and field spike waveforms from the layer V pathway. The fEPSPs, but not earlier field-spike, were suppressed by the non-NMDA receptor antagonist DNQX.



Figure 4. Modulation of pathway-specific synaptic efficacy by radial and tangential DC fields. Application of DC currents in cortical slice demonstrates that tangential current are as effective in modulating pathway-specific synaptic efficacy as radial currents, though pathway-average effects result only for radial electric fields. A, Input-output curve of fEPSP response amplitude and peak latency in the horizontal layer V pathway. Horizontal grey bars indicates 25th and 75th percentile of fEPSP peak latency. B, Relative fEPSP amplitude in the vertical layer V to II/III pathway at different radially oriented electric field intensities (correlation coefficient R2 of linear fit=0.96). The fEPSP waveform inset shows a characteristic change in fEPSP amplitude with positive (+8 V/m, red) and negative radial fields (-8 V/m, blue) from control (no field, black). C, fEPSP responses are significantly (P < 0.05, *) facilitated with +8 V/m fields (left) and reduced with -8 V/m (right) in three pathways. In each pathway, individual slice averages are indicated with colored circles. Grouped average fEPSP amplitudes across all synaptic pathways indicate a 7% polarity-specific modulation of synaptic efficacy with 8 V/m radial fields. Circles in the grouped average represent the across slice means of distinct pathways (blue, red, green, and yellow are S1, S2, S3, and S4 pathways, respectively). D, fEPSP amplitude was significantly modulated by tangential electric fields in all three horizontal pathways but with direction sensitivity and not in the vertical pathway, all consistent with terminal polarization. Although tangential fields affected individual pathways, grouped average of fEPSP amplitudes across all pathways was not significant.


Figure 5. Terminal polarization by uniform DC electric fields using neuron compartment model and analytical/hybrid approximations. Maximum terminal polarization ( Vt ) depends on the length ( l ) of the last axonal branch and becomes uncoupled from the bend point at distant terminals ( l > 3λ ), however for short branches the terminal membrane potential is coupled with the membrane potential at the bend (V0 ). In all cases, numerical simulations applied 1 V/m electric fields. A, For a typical cortical pyramidal neuron, the maximum terminal polarization (Vt ) is a plotted with the corresponding optimal polarization angle ( θ ) of the branch relative to the electric field and the length ( l ) of terminating axon branch. B1-2, Relative terminal polarization (Vt normalized by the axonal length constant and by the electric field) as a function of branch electrotonic length and angle (circle color). B3, Considering the optimal polarization angle the relative polarization asymptotically approaches magnitude 1 with branch length (equivalently, terminal polarization reaches the maximal polarization Eλ for increasing lengths). C1-2, Schematic of a branched and straight axon in a uniform electric field with analytical solutions (see Methods). The straight axon is a special case of the branched axon with infinite final branch length. For long branches, where l >> λ , the terminal membrane potential becomes independent of the branch point and approaches Eλ cosθ . C3, The branched axon model approaches maximal terminal polarization Eλ for l > 3λ . D, Error of approximations (analytical vs. numerical estimates) for branched (blue) and straight (red) axons.

Davide Reato

Department of Physics, Universita’ degli Studi di Padova, 2007

The purpose of this purely modeling study was to analyze subthreshold voltage oscillations (SVO) in pyramidal neurons and to understand if those oscillations can be important for a fast synchronization of action potentials among neurons. A MATLAB GUI was implemented to change the neuronal activity “online”.
The model is a simple 2 compartments conductance-based model.
In the first compartment persisten sodium channels, “normal” potassium channels and a leackage current are present.
The action potentials generation regard the second compartment in which normal sodium and potassium channels are simulated. The subthreshold activity current was considered as an input current for the second compartment.


Some references:

  1. Desmaisons D, Vincent J, Lledo P (1999), Control of action potential timing by intrinsic subthreshold oscillations in olfactory bulb output neurons, J Neurosci. 19(24):10727-37
  2. Gutfreund Y, Yarom Y, Segev I (1995), Subthreshold oscillations and resonant frequency in guinea-pig cortical neurons: physiology and modeling, J Physiol. 483 ( Pt 3):621-40

Yuzhuo Su

Cancerous tissue exhibits altered metabolite concentrations as compared to normal brain tissue. Magnetic resonance spectroscopy imaging (MRSI) reveals such abnormalities in altered spectral profiles. Although the relations between spectral profiles and histological findings are well established, the significant variability of in vivo spectra, which is due to the heterogeneity of tumor tissues, large voxel sizes, and the mixture of normal brain tissues with infiltrative tumors (partial volume effect), often limits their diagnostic potential. This variability complicates tumor diagnosis and grading, as well as the determination of tumor spatial extend. Different spectral analysis methods are being developed to address this problem.

Previously we proposed an algorithm called non-negative matrix factorization (NMF) that extracts constituent spectra associated with different tissue types by simultaneously analyzing all voxel spectra. In principle this method solves the partial volume effect as it determines also the proportion with which each constituent spectrum contributes to an individual voxel spectrum. The algorithm was shown to extract spectral profiles and their spatial distributions consistent with normal and cancerous tissue.

The goal of our research is to demonstrate the physiological relevance of this decomposition for routine clinical brain tumor scans. To do this we analyzed the extracted spectra and showed an improved correlation of choline (Cho) and N-acetyl aspartate (NAA) peak areas with tumor grade compared to conventional method. To validate the physical interpretation of abundances as volume fraction we compared in a phantom study the extracted abundance values with the expected values following the geometry of the phantom and the result showed a good match.

These results indicate that MRSI in combination with the proposed spectrum separation method can improve MRSI in the diagnosis of brain tumors, especially in defining tumor margins for treatment planning of radiation therapy or surgical resection.

Some references:

1. Su Y, Thakur SB, Karimi S, Du S, Sajda P, Huang W, Parra LC. Spectrum Separation Resolves Partial Volume Effect of MRSI as Demonstrated on Brain Tumor Scans. NMR in Biomedicine, 2008, 21 (10): 1030-1042.PDF

2. Sajda P, Du S, Brown TR, Stoyanova R, Shungu DC, Mao X, Parra LC. Nonnegative matrix factorization for rapid recovery of constituent spectra in magnetic resonance chemical shift imaging of the brain. IEEE Trans Med Imaging, 2004,23(12):1453-1465. PDF

3. Lee JS, Lee DD, Choi S, Lee DS. Application of nonnegative matrix factorization to dynamic positron emission tomography. Third International Conference on Independent Component Analysis and Blind Signal Separation. San Diego, CA; 2001. p 556–562. PDF

4. Lee DD, Seung HS. Learning the parts of objects by non-negative matrix factorization. Nature 1999;401(6755):788-791. PDF

Davide Reato

I recently received an awesome gift: a running watch that can measure speed and heart rate (CW Kalenji 500 SD PC, in Europe you can find it at Decathlon). Since the data can be recorded and then transferred to a normal pc, I would like analyze them to see if there is a simple but clear quantitative signature of improved performances in speed, heart rate or a certain combination of the two (of course, assuming I go on running regularly). In other words, my question is simply: am I improving running?

With the Geonaute software it is possible to save the recorded data as .fitlog files and then open them with whatever spreadsheet-type of program. Finally, normal .txt files can be saved with incremental time, space and the values of the heart rate.

All the analysis can then be done in Matlab.

In the following analysis, I grouped together runs in different locations, different times and without controlling for the altitude (no gps in the device I use). While all these variables obviously play a role, I simply want to see if there is some metrics solid enough to quantify the performances. Right now I have only 17 runs for about 120 km total. I will constantly update the graphs on this page.

The sampling rate of the watch is 0.2 Hz (one sample every 5 seconds). Speed can be determined immediately by considering the space intervals and then dividing it by time (the canonical definition of speed). Since I often run on the street, I have sometimes to almost stop (for example for the traffic lights). To remove these “artifacts” I simply consider all the data points where the speed is within 2 standard deviations.

The following plot represents speed (blue) and heart rate (red) in a representative run (about 9 km total). In this specific case I ran 3 laps around the St Nick park.


The three bumps represent a pretty steep uphill (3 laps as i mentioned). Clearly the speed decreases and the heart rate increases.

Averaging across all the runs, I get this plot:

where the average speed is in blue and the average heart rate in red (n = 17). Shading indicates the standard deviation for each point.

The first thing I noticed is an increase in both speed and heart rate with time (note that the heart rate jumps immediately when the run starts). The change can be estimated by fitting speed and heart rate over time for all the runs (a robust fit). The following plot represents the estimated slope of the linear fits:

Even if there are two outliers where the slopes are negative (for one the slope is not significant, while for the other the p-value is way higher than in all the other cases), the slopes are significantly positive (p < 10-4), showing that both speed and heart rate increase during the run. Interestingly, the slopes for speed and heart rate do not significantly correlate (p > 0.25).

I then examined how speed and heart rate ratio changes in different runs. The following plot reports all the ratios. The magenta line connect all the average values, while the green one connects the medians. The plot was created by using the wonderful function notBoxPlot by Rob Campbell (way better than Matlab boxplot).

I was expected to see a decrease in this ratio, considering that I was imagining that with more training, the same speed can be achieved by keeping a lower heart rate, but that is not the case, at least with the current data set. More data are clearly necessary before making any claim about this point.

I then focused on the relationship between speed and heart rate in a 2d representation. The plot shows the average speed and heart rate for the different runs (error bars indicate standard deviation).

The clear two clusters are easily explainable. I ran with someone else. It is interesting to note that while this effect was not visible in the average values (previous figure), it is very evident with this representation. This suggests that speed and heart rate probably correlates (so for different speeds, heart rate keeps in track the change).

To check for that, I grouped together all the points from all the runs in a heart-rate/speed diagram (~9000 data points).

As you can see, there is a strong significant linear relationship between heart rate and speed. Something I find extremely interesting is that the intercept of the fit, ~65 beats/min well represents my heart rate when I simply sit on chair (so at speed = 0 km/h, 58-70 beats/min but I have to measure it more systematically).

The following plot shows the density of points in the heart rate/speed space.

Plotting the density of points, it is possible to still see clearly the linear trend and, at the same time, the two clusters corresponding to running alone or running with someone else.

I then tried to look at the normalized cross-correlation with a sliding temporal window between speed and heart rate. The following plot has been generate using this function by Tim Streeter and it is relative to all the runs. I did not label the axis to do not reduce the space for the plot itself. The time lag (min) is on the y-axis, while time (min) on the x-axis. The color goes from 0.5 to 1 (normalized cross-correlation). The time window is 5 minutes.

While the single plots are clearly noisy, it looks like in many of the runs, there is an increase in the cross-correlation between speed and heart rate. Averaging all the runs I get this plot (where the time variable has been cut to the shortest run).

The average, restricted to all the runs longer than 30 minutes (n = 14), clearly shows that trend. The cross-correlation between heart rate and speed increases with time, with time delays within minutes. Note again, that the cross-correlation is normalized, so higher values do not simply reflect the increase in speed or heart rate during the run (shown in the first figures).

So, for now, I do not have any way yet to see if I am running better or not but I am having fun 😀

Thomas Radman

I am using whole cell patch clamp to study the electrop hysiological response of single cortical cells to uniform electric field stimulation.  Cell type and morphology is reconstructed through the use of biocytin stain injection throught the recording electrode.  The long-term goal of the lab is to create an accurate in vitro model of non-invasive electrical stimulation therapies such as tDCS and TMS.

Some references:

  • Tranchina, D. and C. Nicholson, A model for the polarization of neurons by extrinsically applied electric fields. Biophys J, 1986. 50(6): p. 1139-56.
  • Ranck, J.B., Jr., Which elements are excited in electrical stimulation of mammalian central nervous system: a review. Brain Res, 1975. 98(3): p. 417-40.
  • Esser, S.K., S.L. Hill, and G. Tononi, Modeling the effects of transcranial magnetic stimulation on cortical circuits. J Neurophysiol, 2005. 94(1): p. 622-39.
  • Maccabee, P.J., et al., Magnetic coil stimulation of straight and bent amphibian and mammalian peripheral nerve in vitro: locus of excitation. J Physiol, 1993. 460: p. 201-19.


Clinical Trials at Memorial Sloan Kettering Cancer Center are ongoing with support from NIH-NCI

Next generation devices under development


Varun Bansal


We developed a working prototype of reflective pulse oximeter that enables surgeons to obtain real time feedback on local tissue oxygen saturation (SpO₂) during operative procedures.

This handheld wireless pulse oximeter is suitable for the intra-operative measurement of tissue SpO₂ during bowel surgery. The device adapts principles and technology developed for non-invasive pulse oximetry, and introduces tissue interface, physician’s tool, and signal processing algorithms for intra-operative applications. The handheld device includes local display of SpO₂  levels (<1s refresh) at the contacted tissue, and signals the operative on degraded signal quality/fault. An onboard microcontroller digitizes and processes signals transduced through a control LED array. Signal processing and display parameters were optimized for operating room conditions. A disposable functionally transparent cover provides both device and tissue protection.  SpO₂ and pulse signals can be processed on a PC or operating room VI. The incorporations of a pressure sensor to increase accuracy and robustness is explored.

Varun BansalJorge Vega


The Multi – Channel Stimulation Interface device is designed to be used as an interface device between an isolated 2-channel current controlled DC stimulator and 5 stimulation leads. The Interface device is developed to enhance the DC stimulation targeting capabilities withhout making any changes in 2-Channel stimulator device. The multi channel Stimulation Interface device can be connected to any kind of 2-channel stimulator available in market by just using the appropriate input cables. The Multi Channel Stimulation Interface device can be connected with 2- 5 output electrod es depending on the treatment and targeted area of the stimulation.

The Multi Channel Stimulation Interface device has the functionality for pre-stimulation accessing lead quality and potential faults. Though these functions are intended to facilitate multi channel stimulation, in PASS MODE this functionality is disengaged, allowing the current to pass through without any interruption through multiple electrodes. The combined output of the multiple output leads is thus equivalent to the single input.

An inbuilt TICKLE features activates a transient <100 µA pulse, intended to “regulate” lead resistance. When the Multi Channel Stimulation Interface is powered OFF the device disconnects 2-channel input from the 5-channel output. When turned OFF, the output of the 4X1-C2 is zero even if the 2-channel stimulator is active.

Refer to Device Manual for more information.