Deep learning models reshape healthy vital signs to mimic clinical conditions
Representing Clinical Conditions on Vital Signs from Healthy Individuals using Latent Modeling
Machine Learning
Summary
It can be hard to get enough medical data to train computer programs that spot health problems from vital signs like heart rate or blood pressure. The authors made a machine learning tool that changes healthy people's vital signs to look like they belong to sick patients, using patterns learned from intensive care unit data. This approach helps create new data that mimics specific illnesses without needing real patient samples. They also developed a way to measure how well the changed data matches the disease patterns compared to other methods.
What this means in practice
- •For clinical data teams: Generate synthetic patient vital signs to support training of diagnostic algorithms when access to real patient data is limited.
- •For healthcare ai developers: Create realistic variations of healthy vital signs representing specific clinical conditions to improve machine learning model robustness.
Authors
Rafael Pina, Varuna De Silva, Mindula Illeperuma
Abstract
Machine learning can be crucial to help scale complex signal processing applications in scenarios such as healthcare. However, these machine learning models need rich datasets to be trained and there are often cases where it is not possible to access representative datasets. In this paper, we propose a deep generative model based on conditional variational autoencoders with the objective of augmenting the vital signs of healthy individuals in a way that mimics the patterns of a certain clinical condition. More specifically, we use a publicly available ICU (Intensive Care Unit) dataset to train our model and then evaluate it using the vital data that we have collected from healthy individuals. Our results demonstrate that the proposed model can not only learn the underlying dynamics of the ICU data but, more importantly, can reshape our collected data from healthy individuals in a way that is aligned with the vital signs of a certain clinical condition. We propose a distance metric that shows how our model can generate samples that are more aligned with the intended clinical labels when compared to the tested baselines.