μblog: engineering from the trenches

As a final follow up, here my paper that was accepted for a slide presentation at the IEEE EMBS meeting:

nchernyy-embs2008i

I am working on a more thorough paper which also covers harmonic/non-linear analysis and does statistical validation of the results, so hopefully that will be forthcoming in the next year. Any questions or comments are always welcome.

As for the image, this advertisement was on many of the bus stops in Vancouver. I guess Intel lost their trademark in Canada.

del.icio.us | digg

I was very impressed with the IEEE EMBC 2008 as it was really obvious that the conference organizers were able to share cutting edge science while maintaining an engineering atmosphere. My presentation went smoothly! The over-all theme of the conference was to push the limits of current bio-medical technologies to make a better patient experience focusing on better diagnostics, ease of use and instrumentation intelligence. Vancouver ended up being a pretty nice city as well, after it stopped raining that is.

Most of the diagnostic improvements were in the form of better signal processing and data analysis. The key motivation was to get more use from already-available instruments without substantial additional investment. The star of this show was definitely breast cancer detection where the leading edge technology focused on using both elastography and magnetic resonance imaging. The key idea is that an ultrasound actuator can send elastic waves through the breast tissue that ultimately displace (cyclically) charged particles and this displacement can be detected on MR in 3D. Since this displacement depends on tissue properties, this is effectively a measurement of both electrical and mechanical tissue impedance where changes in this impedance can highlight potential tumor areas.

The focus of the ease of use designs was overrun by wireless applications. Everything from neural recordings, to patient vital sign measurement, to device configuration was being moved to a wireless network without any serious security considerations. There were various stimulation devices proposed, both for neural and cardiac applications, that used standard bluetooth or zigbee type protocols with no encryption or authentication whatsoever. The typical response I received from the presenters was that they were developing a proof-of-concept and would add the security features at a later point, however, my experience with network engineering leads me to believe that it is best to security considerations from the start. It is easy to tell when an authentication system was cobbled on as an afterthought and such a system either decreases performance or has some flaws at the interface. It would take a pretty sick person to launch a denial of service attack on a pacemaker, but I don’t doubt that such people exist.

Device intelligence was mainly about making devices that require less intervention and can be operated by unskilled technicians and patients. Patient operation stood out as there seems to be a substantial positive psychological effect when they are allowed to have some input on their therapies. The increased device intelligence would then attempt to control the patient input to prevent any harmful reactions. Furthermore, it was good to see at least a few medical devices that would initiate some sequence of events based on monitored patient activity as compared to strictly open-loop programs.

del.icio.us | digg

I am leaving for Vancouver, Canada tomorrow to attend the 2008 IEEE Engineering in Medecine and Biology conference. I am presenting my paper on Friday at 15:00 titled: multi-taper transfer function estimation for stimulation artifact removal from neural recordings. Anyone who wants to meet up or come to my talk, just show up, leave a comment, or send an email.

del.icio.us | digg

I am spending the first half of this week in Cleveland, Ohio at the 2008 Neural Interfaces conference.

The first day was spent mostly discussing neural stimulation as an interface paradigm. Judging from the presentations, deep brain stimulation (DBS) is the most popular in both the research and medical communities. There are about five devices on the market that are FDA approved for DBS so many clinicians are looking to use these stimulators in various off-label applications to stimulate many targets within the brain as well as the vagus nerve in the periphery. The primary success story for DBS is continuous stimulation in the sub-thalamic nucleus and globus pallidus to alleviate effects of Parkinson’s disease. The same techniques are being applied with marginal success as a treatment for epilepsy, depression and obsessive compulsive disorder. It is still tough to show significant improvement for some of these latter diseases. Furthermore, it is equally tough to prove that the turning on the stimulation has any beneficial effects for those diseases. In some cases, implanting the stimulating electrode is good enough.

The second day has had some excellent coverage of the state-of-the-art prosthetic devices. One of the novel mechanisms that I was not aware of is nerve re-direction in the case of amputation. For example, there are surgical techniques to re-route nerve bundles that control arm movement in the case of amputation near the shoulder. The muscle groups on the chest become the targets and in some cases, the amputee can control those muscles by attempting arm movement and feel arm sensations through mechanical stimulation of their chest. If this type of surgery is successful, there are examples of external prosthetic arms which are controlled by electrically tracking muscle contraction on the chest and provide basic tactile feedback through mechanical stimulation.

On a side note, there was a luncheon focusing on commercialization of neurological devices. Translating an implantable device from academia to “mass product” seems to be a process that will test the designer’s patience and humility. The general feeling on the commercial side is that the academic designer should already have a working design that is undergoing human clinical trials when they approach a company for commercialization. Even then, the commercialization company will evaluate their potential for revenue generation and may provide a 0.25-0.5% royalty for the academic designer to take over the mass production of their product. This is while keeping in mind that the path from specifications to clinical trials is at least ten years. This may be the reason that many researchers try to create start-up companies and commercialize their own products to see more fruits of their labors. This approach, however, is perilous and may end up being unrewarding. This is clearly seen with the bankruptcy of Cyber-Kinetics, the company that has been receiving media attention for years for allowing several paraplegics control of an on-screen cursor by thought alone. (Their website has not been updated recently to disclose this.)

del.icio.us | digg

As I have mentioned, the semester is winding down and projects are piling up. To help move things along, I have written a pair of documents that outline how to determine the conductivity of a plane (and volume) with a conducting disk (and sphere) of arbitrary size in the middle. I solved Laplace’s equations in both polar and spherical coordinate systems, then used boundary conditions to determine the electric potential and then determined the ratio of applied field to current density to determine the conductivity in the presence of the suspended object. I have checked these early drafts over a few times, however, there may still be some mistakes remaining, so please be warned. Also, feel free to post questions and I will make an attempt to answer them.

( disk-efield.pdf )

( sphere-efield.pdf )

P.S. The photo-op was staged during my last vacation, I would never use Classical Electrodynamics as a coaster.

del.icio.us | digg

I have been working on spectral analysis of time series data and I am finding that (duh!) it is often easier to use someone else’s wheel than to invent a new one myself. That is, if their wheel has all of the necessary features and meets your performance criteria. I have been working with multi-taper spectral estimates and have found that the Chronux code (MATLAB scripts), available from Partha Mitra’s lab, to be fairly easy to use and effective for some continuous time-series analysis. The code also handles point process analysis, however, I don’t use that feature much.

The main reason that I like to use the multi-taper methods based on Slepian tapers, is that each taper gives a independent measurement of the signal spectra which then allows for computation of variances and therefore confidence intervals. This gives a certainty measurement to the analysis which can either make a strong case for its significance or will identify deficiencies in the data. Basically, these people put some years into writing this code and if you need to do coherence or spectral analysis in MATLAB, might as well give it a try.

del.icio.us | digg

A little off topic, but some clever researchers at UCLA have created the Gene2Music project where they mapped classical music to genetic sequences and created some MIDI files (as well as sheet music). The image above is from one of the sheet music sections.

del.icio.us | digg

Adding references, or data from, publications that are over a hundred years old seems to be a popular trend among scientific presentations these days. Sometimes it is to give a false sense of scholarship, however, it is often to remind us that some ‘new’ scientific breakthroughs may be simple re-interpretations of old discoveries. I try to note the references and look them up, when time permits. Here is the first paragraph from the preface of ‘Epilepsy and Its Treatments‘ (1904) by Spratling:

 The great progress made in the knowledge of epilepsy and its treatments during the past decade and a half, and in fact that no complete treatise on the subject has appeared in the United States since Echeverria’s work was published thirty-three years ago, was the chief reason that lead to the preparation of this volume.

With the exception of Manuel Echeverria (On Epilepsy: Anatomo-Pathological and Clinical Notes (1870)), the sentence can still be used in a modern book/review of Epilepsy without much alteration. The reason that the 1904 book was cited was to show that, a hundred years ago, physicians were aware that, on very rare occasions, were non-clinical. For example, it was noted that verbal interactions were sometimes enough to bring people out of seizure, something that researchers who seek alternative epilepsy treatments are rediscovering. (On a slight side note, there is an interesting personal account by Feydor Dostoyevsky starting at the bottom of page 466 where he links a pre-seizure state to a state of mental enlightenment.)

This long winded introduction was to present a pair of review articles from the early 1900s that covered what the authors thought were the highest achievements in physics and applied math of the previous century.

1905barus-the-progress-of-phsyics-in-the-nineteenth-century.pdf

1900woodward-the-centurys-progress-in-applied-mathematics.pdf

The image is from the IEEE and is of the Georgetown, CO steam/hydro powerplant.

del.icio.us | digg

The last two days of the meeting started to blend together as the 8am to 8pm days started taking their toll. Sunday was mostly taken up by an investigators’ workshop dealing with basic principles behind epilepsy. Jack Parent had some interesting words to say about a protein called reelin. This protein has been implicated in limiting the process by which neurons form new connections, as though it somehow helped regulate “good” networking. The it seemed that, in animal models, seizures were associated with reduced reelin concentration and increased neuronal proliferation which led to more seizures.

Monday was divided between visiting Brian Litt’s and Marc Dichter’s labs at U Penn, presenting my poster, and more lectures. Visiting the labs made it apparent that while we have more space at Penn State, the U Penn labs have a more strategically located near civilization. The poster presentation went fairly smoothly with the toughest questions being about my choice of multi-taper spectral estimation approach over something like wavelet transform. The final lectures that I attended were covering new definitions of seizures. The basic idea was that clinicians had access to new equipment that could record more EEG channels at higher frequency as well as higher density electrode grids/arrays. The result was that neurological signals could be recorded with finer spatial and temporal resolution. These newer systems could pick up the huge electrographic seizures that have been recorded for decades, as well as shorter discharges termed “microseizures”. This could certainly give medical professionals better tools to diagnose neurological disorders and localize seizure foci, however, I think that the definition of a clinical seizure will consequently be broadened in the upcoming years to include some of these smaller discharges. This may lead to a broader definition of the disease and would lead to more people becoming clinically epileptic. This, coupled with an increased epilepsy risk as a result of traumatic brain injury,  will lead to a great jump in the number of reported clinical epilepsy cases of epilepsy in the upcoming decades which the analysts will probably attribute to things like mobile phones, wireless internet, and video games.

During the breaks, I had a chance to walk around the convention center visiting places like the city hall and the reading terminal market. In addition to the pictures of the city and the market, I am including pictures of the exhibitor section of the American Epilepsy Society meeting. Since the target audiences of the meeting are medical doctors, the drug companies really spend money on marketing and give away cell phone chargers, thumb drives, usb battery chargers and laser pointer pens which include usb flash drives which can also recharged through the usb connector.

del.icio.us | digg

This day was split into three sections: Merritt-Putnam lectures, Engineering special interest group, and a set of lectures on high-frequency EEG recordings.

The morning was dedicated to the Merritt-Putnam lecture series, with this one focusing on traumatic brain injury (TBI) and the resulting possibility of epilepsy. The key theme was that there was a correlation between TBI, such as getting hit with a baseball, and developing epileptic seizures down the road (epileptogenesis). There are some clinical measures which can sort the severity of the TBI into mild, moderate and severe categories, however, the specifics were not presented. Various mechanisms, ranging from molecular to neural network organization, were implicated as potential targets for epileptogenesis, however, the key point that was reiterated by every presenter was that there was concern that troops coming back from Iraq were looking forward to epilepsy in their later years. Part of the reason for this is that explosives play a much larger role in modern warfare, as compared to bullets, which have a higher propensity to generate shock waves and thereby contribute to TBI. According to the statistics presented, there are over 2,000 “official” cases of severe TBI as of October 2007 with an estimated 10,000-20,000 more at risk.

The main focus of the SIG was to demonstrate to try to define what a seizure was. Several groups were presenting data from tiny electrodes embedded in standard electrode arrays to help get better spatial resolution of EEG measurement for seizure focus detection. The most entertaining talk, however, came at the very end and provided an overview of stimulation efficacy for seizures. The point was that in many cases, simply implanting the electrodes was enough to stop seizures. It didn’t statistically matter if the stimulator was turned on or not, providing that the seizure focus was located correctly.

The last session was titled “Broadband EEG”, but the range of talks was very wide. Gyorgy Buzsaki gave the first talk part of which dealt with qualitative energetics. The idea can be explained with this analogy: when people clap at the end of a musical performance, they clap randomly at first, then, a pattern develops and people clap rhythmically. When people clap in unison, they typically clap slower than they did before, expending less energy. The same can be seen on functional MRI (something that measures metabolism) where things like the hippocampus, a popular seizure focus, shows decreased energy usage during seizure activity.

del.icio.us | digg