N/AClass of 2026
- How can an AI program process lactate levels, C-reactive protein levels, and Interleukin-6 levels to figure out whether the patient needs further tests or already has sepsis? with mentor Lauren (Feb. 26, 2022)
How can an AI program process lactate levels, C-reactive protein levels, and Interleukin-6 levels to figure out whether the patient needs further tests or already has sepsis?Started Oct. 29, 2021
Abstract or project description
Abstract: Every year 1.7 million people in the US are diagnosed with sepsis, a bacterial infection that leads to death in roughly 250,000 patients¹. Sepsis is a leading cause of death in the US, and the most common death cause in hospitals1. Sepsis is hard to diagnose because its symptoms, such as shortness of breath, fever, high heart rate, discolored skin, and dizziness, are very similar to other diseases2-3. In this project we use AI technology to create a computer program that, given patient’s lactate levels4-5, Interleukin-6 (IL-6) levels6-8, C-reactive protein (CRP) levels9-11, and symptoms, can identify whether the patient has sepsis or not. The main goal is to create a computer program that takes symptoms inputted by a doctor and outputs the next steps and tests to be done in order to determine whether the patient indeed has sepsis. The program will assist doctors by minimizing the number of tests as well as the time to finalize the diagnosis. In the future this program should be able to help diagnose other diseases as well. Introduction: Sepsis is a life-threatening organ dysfunction caused by a bacterial infection. If not recognized early, it could lead to septic shock, which is multiple organ failures, which finally leads to death. 1.7 million people in the US are diagnosed with sepsis, where 75,000 of the cases are children12. Sepsis is hard to diagnose in the beginning since symptoms are similar to many other diseases13. Symptoms include; shortness of breath, fever, high heart rate, discolored skin, and dizziness14-15. A survey of more than 1000 hospitals stated that 86% of intensive care specialists misattributed the symptoms and 45% missed a diagnosis of sepsis16. Diagnosis isn’t easy and require multiple tests to fully diagnose a patient. Biomarkers, specifically lactate, CRP, and IL-6, were recently proposed to evaluate sepsis patients’ prognosis. Our program will use these biomarkers to provide an easy and quick diagnosis for sepsis.