Biomedical Data Science Lab

Pre-Lab: Introduction to Python, working environment, and DSP

Heart beat estimation, signal denoising and mobile health

In this lab experience we will be dealing with the problem of heart rate estimation from an electrocardiogram (ECG). For a healthy individual, and under ideal conditions, this task might be easy to accomplish. However, in a noisy setting the estimation task becomes more challenging. We will study how prior information about the signal and the noise can be exploited in order to improve the estimation task under less ideal conditions. In particular, we will implement a simplified version of the Pan-Tompkins algorithm starting with and ideal toy example and ending up with a real scenario where the signal has been acquired using a mobile device.

• QRS-complex detection algorithm (Pan-Tompkins, 1985)
• ECG background notes
• Datasets

Pre-Lab: Intro. to optimization via gradient descent with PyTorch

Cardiomyocite classification based on action potential

In this lab experience we will be looking at the problem of classifying cardiac cells by looking at their Action Potentials (APs). For the purpose of this task we will be using synthetically generated APs following the models in (Nygren et al., 1998) and (O'Hara et al., 2011).

• Adult action potential dataset