This page has details of the task we'd like you to complete for Idoven ML engineer job offering.
The electrocardiogram (ECG) is a non-invasive representation of the electrical activity of the heart from electrodes placed on the surface of the torso. The standard 12-lead ECG has been widely used to diagnose a variety of cardiac abnormalities such as cardiac arrhythmias, and predicts cardiovascular morbidity and mortality. The early and correct diagnosis of cardiac abnormalities can increase the chances of successful treatments. However, manual interpretation of the electrocardiogram is time-consuming, and requires skilled personnel with a high degree of training.
Automatic detection and classification of cardiac abnormalities can assist physicians in the diagnosis of the growing number of ECGs recorded. Over the last decade, there have been increasing numbers of attempts to stimulate 12-lead ECG classification. Many of these algorithms seem to have the potential for accurate identification of cardiac abnormalities. However, most of these methods have only been tested or developed in single, small, or relatively homogeneous datasets.
We've tried to keep this task as similar to working here as possible. With the data provided we would like to ask you to implement a ML classification model and provide the code necessary to train it from scratch. We are not asking for a model with any specific capabilities or performance, it is up to you to explore the problem and assess what could be reasonable and interesting within the given time constraints, taking design decisions accordingly. We'd like you to analyse the problem and give us some insights. The insights should be useful and actionable in some way.
We ask data scientist do want to join Idoven to work with anonymised patient data, and on the basis of this data be able to:
- Read the ECG files and corresponding annotations
- Show how they will work on the signal and plot the signal in appropriate manner to be read by a doctor
As a result we expect a github project with and extructure that will include:
- Reference documentation used
- Jupyter Notebook, in an running environment, Colab, Docker.
- An explanation of the work done and lessons learned.
It would be great if you could have this done within a week. If that's not doable for you, let us know early.
Also, we don't know how long this should take you, but we're not looking to reward the person that spends the most time on it. We believe in working smarter not harder.
In case it's helpful, here's some other tips for you:
You can ask questions. This isn't a "bonus points if they ask questions" thing, just that we'll answer what we can if you need us. Like we would when working together.
We like to have a real work example work flow, we wencorage you to do a pull request and send the pull request for evaluation.
You can request more information/data, but we'd rather you didn't. If you really need more, let us know, but there'd need to be a compelling reason for it.
We want to see what it's like to work with you, and the quality of work you'd produce. This is a chance for both sides to see how that is.
We will be making a decision based on these tests, so do give it your best.
Thanks for giving this a go, we can't wait to see what you come up with.