Research Project Title:
Characterizing and Predicting Sepsis Treatment Outcomes Using Dynamic Features from Vital Sign Time Series
abstract:Sepsis is a serious health condition that has a high morbidity and mortality rate. It results from infection and causes organ dysfunction, which leads to death in approximately 25% of the cases. Sepsis can have numerous symptoms making early diagnosis more complex. If the signs of sepsis are not recognized early on, this could lead to septic shock which significantly increases the likelihood of mortality. The objective of my project is to predict the outcome of sepsis treatments using machine learning and signal processing techniques. Application of this work will aim to provide clinicians with a tool to characterize patients' response to treatments and stratify risks for possible adverse outcomes for sepsis patients in intensive care units.
"I decided to participate in SuperUROP to apply the knowledge I have gained during university to real-world research problems. I chose this project as I have experience applying recently conceived adaptive signal processing techniques to analyze EEG/ECG data. I would like to learn to apply machine learning techniques to biological data. I am keen to apply my knowledge to acquire useful results and if possible, publish the findings in a paper."