I’m an electronic engineer by trade and a software engineer by training.  My cardiologist is not so different.  Our jobs are that of problem solvers.  To solve a problem:  1) Gather evidence,  2) Make an intelligent, educated guess,  3) Apply your educated guess.  The more empirical the evidence is the more accurate the guess is.  Don’t confuse engineering (the application of science) with science (the fact of knowing something).  In the cardiologists case it is obviously important to base your decisions on the best information possible.

In my last post, It’s My Heart I left off with my doctor having diagnosed me with vasovagal syncope.  Knowing this is wrong, I awoke at 4am and couldn’t sleep trying to figure out how to get a correct diagnosis.  I need to prove with empirical evidence to another engineer that my problem is rhythm related and not a reflex problem.  It comes down to being able to capture my issue on an ECG strip.  I either need to purchase or build an ECG machine.  My search revealed no consumer grade ECG machines which produce anything usable for less than $500.00, at least not that I could find.  I then searched for plans to build an ECG machine and I immediately came across this Scientific American Article.  It described a unit which amplifies the heart’s signal while not amplifying noise.  The output is a 1 volt peak-to-peak signal.  I then needed a way to capture this signal.  It didn’t take me long to remember that sound cards capture audio in a range that is compatible with a 1 volt peak-to-peak signal.

I spent some time on the internet and at the local electronics shop to purchase all of the parts needed.  I also needed some stick-on ECG electrodes to connect the leads to the skin.  Fortunately there are many sources for these on the internet.  For about $15.00 I purchased a pack of 50.

After building the ECG electronics and doing some limited testing I connected it up to my PC sound card and began to capture the data with the open source audio software Audacity.  This software is awesome software for novice and professional recording artists alike.  But the fact Audacity is free and graphs the audio signal were the overriding factors for me :-).  I hooked it all up and there is was.  An ECG:

Audacity ECG

Audacity ECG

At the time, I hadn’t realized how important using Audacity to capture the signal was.  It turns out that the filters available in Audacity are very important and allow me to cleanup any remaining noise in the signal.  Additionally I can capture hours of data then select only the sections that are important for review.

When I finally captured an SVT on the machine (4 months later) I needed to present this to the doctor.  My mom, a cardiac nurse, reviewed the data with me.  She found it difficult to review the data in Audacity because she had been trained as a cardiac nurse to read them on strips which have a graph paper background with specific timing.  Without this it is difficult to determine all of the important timing parameters of the ECG.  To present this data to my doctor I needed to convert it to strips he understands.

To convert from the raw audio data to a graphic strip I wrote a Perl script.  It converts 6 or 30 second audio captures into GIF files which are easy to print.  The script maps the audio timing onto the graph paper background that provide a reference for medical professionals.  All of the ECGs you see on this site were converted with this script.  Click Here For an Example

In the end the cardiologist acknowleged the SVT and we were back on track.  Still the diagnosis is “benign” SVTs.  While the word “benign” is great news they still don’t feel benign.  After time and time again being told they are benign I’m learning to live with them.  We are also trying to control them with meds.  More on the meds later.