I had developed a software in cooperation with my student that can predict arrhythmias from ECGs (rabbits). I am planning a further development of the said software but more human-like data, apes, are very welcomed.
It would be great if we find a common denominator in our researches.
In any case, your research is very interesting because, in the future, apes in captivity can be compared to apes in the wilderness. In the way, we can pinpoint the real causes of civilisation diseases.
I am not sure what you mean but the answer goes for the following question: "does the age of the subject matter?"
The answer: " It is age independent. All that I need is based on ECG recordings and similar input data."
The software is to a certain extent species and data independent. :-)
Are you using voltage changes (LVH), ST T wave changes, shift in axis, HR variability, abberancy, interval changes , etc... or is it some pattern algorithm?
thank you for your reply on it. so just want to have 'images' little bit
Perhaps I lost something in translation but there are no algorithms that can predict arrhythmia. There are those that can identify early changes that MAY be predictive, but changes do not 100% result in true arrhythmia (e.g. QT prolongation does not always produce Torsades). As for your statement that the software will be species independent - that is true for its processing capability - however, species vary markedly in predominant ion channels that form the ECG so they will have different markers related to early arrhythmic changes. The current ECG pattern analysis software (DSI, EMKA) are capable of analyzing thousands of beats to determine trends - is your software capable of that?? Given the limited data that could be collected from apes, what could your software provide that a qualified cardiologist could not?
The software that is developed, based on advanced tools from complex systems, is capable of assessing trends in any species, different species will have different features, and is capable to predict arrhythmias up to one hour before their onset just by taking 10 minutes of one lead ECG.
Specificity and sensitivity of arrhythmia prediction is 93% on a rabbit model with artificially generated arrhythmias.
As far as I know there exist yet another approach which goes 89% sensitivity and specificity on human ECGs taken from the intensive care unit.
So, our software is the best predicting but not on human ECGs.
Currently, it would be great to test the software on others ECGs from humans and animals. It could be done in cooperation.
The question is whether these predictive values found in experiments and taken from the intensive care unit will be the same under natural conditions. Generally, if the abnormalities in the electrocardiogram are not with large deviations from the norm, they are considered pathological only when associated with the clinical signs. I will follow your project with interest.
Dear Atanas,
Thank you for your best wishes, it can help many people to live a better life.
It is clearly explained in my project. I am 100% positive from my 30+ years long experience with VTs that I can feel their onset (for sure before the heart transplantation as after it, it is not sure into which extend the heart innervation is reconnected).
From the preprint of the review paper that I published on researchgate recently, it is visible that complex systems tools get enough mature to be applied in medical praxis. It will get accelerated in the coming decades. I recommend you to read it as it is digestion of all the major achievements of CSs applications in medicine. It is worth several years of full studying of this research. It definitely saves a lot of your time and helps us to get on the common ground.
The secret is hidden in the database storing all ECG curves necessary to evaluate classifiers. We must build this database first. I would like to initiate a project that will accomplish this. It will be an international, multi-institutional project with the sole purpose: building a heavily tested reliable testbed for all who does research on arrhythmias. Why the aerospace industry has a higher level of normalization & safety measures than medicine? We must rebuild it from the bottom.
All research input data in this database must be open to everyone without restriction. Everyone will get the credit for it.
I am not hiding the information that there is a huge potential for development of the completely new level of medicine based on analysis of biosignals using the concept of entropy (that asses information content in the signal), complex systems, machine learning, and data mining.
It is really a powerful toolkit and there are appearing quite of few very good research outputs.
My own research is confirming the results of the research group (nature paper) which predicts the onset of arrhythmias in patients on ICU using deep learning. My approach goes a completely different way. I will make the research available asap.
Currently, I am too slow because of serious disabling neuropathy due to the use of statin and immunosuppressive drugs in combination with other drugs during and after the heart transplantation.
Thank you for your patience. I have to mention the pressures of a certain group of people, medical & pharmaceutical sphere, who do not wish my continuation in this research. I do not like it and I do continue despite the threatening.
Therefore, my research on this subject, the life-saving part and not sport related part is for free. That is absolutely fair to those diseased people who are waiting on the output of this research. I am absolutely against making money on diseases and suffering of others. Science must keep high esteem of morality.
All the best,
Jiri
PS I wished to cite your name but still cannot figure out how to do it. I tried @AtanasPankov but it does not work.
- Badges
- Science topic
- Similar topics
- Phytochemistry
- Extracts