some questions about ECG

[electronnne]

Hi, I am an engineering student in telecommunications looking for some answers about ECG. Actually, in my annual project i am asked to find an algorithm that enables a simple person (not a doctor) to judge whether the ECG record is correct or not (not correct due to misplacement of the electrodes or stress or high temperature ...).
The problem is that i can't even distinguish between a correct ECG and a non correct one, because simply (as i deduced by myself) we can have a record which looks wrong (e.g inversed QRS complex) but in fact it is a correct one for a sick person ( an anomaly which is represented by the inversion of the QRS complex).
Some silly questions :$ :
1) Are there any hints that help to conclude if the ECG is of a "sick" patient or it is just an artifact?
2) the 12 leads of an electrocardiograph used in a "healthy" person in perfect conditions give the same features? i mean the regular known form P QRS T waves are similar in all leads records?
3)the ECG can be recorded as sound directly from the electrocardiogram?

Please I am looking for an answer as soon as possible.Thank you

---

Members don't see this ad.

 

[nlax30]

I would actually recommend you picking up an ECG book and learn some of theory behind it as well as exactly how it's measure and what it's showing you.

I recently had a long conversation with a family member working on a similar computer science grad project involving artificial intelligence and interpreting an ECG. Well, after a lengthy discussion of how it actually works and taking him through the basics if how one interprets it, he admitted be had been thinking about it all wrong and had to seriously rethink his project.

So I'd say get a book that starts from the beginning and takes you through the theory, describes each lead, each wave, and takes you through how to analyze one.

So it's not just a simple of saying whether an ECG is "correct" or not. If you're just trying to determine if it's a valid tracing then you can look into how having the limb leads reversed would affect the tracing or what "artifact" looks like, but some of that is still fairly subjective.

And tracings can vary quite a bit between two relatively healthy patients, so it's not just a matter of having one correct kind of tracing. A book will take you through the typical normal characteristics of each portion of the ECG.

Not trying to discourage you or anything, but this is almost the same discussion I had a fee months ago and the other person realized it wasn't nearly as cut and dry as he originally believed.

 

[electronnne]

Thank you nlax30 for your answer. You know, it is only 6 a.m in Tunisia ... Your post made me feel totally lost. I hope it won't be a bad day. No kidding

The problem is that it is a one month project which will end in 19th April 2011. My supervisor told me not get in deep with ECG, only peaks form concerns an engineer but here i am confused.

Thank you once again