For patients with colon cancer, surgery is the only curative option, and is associated with a high rate of complications and readmissions. Up to 20% of surgeries result in complications, many of which are severe and lead to readmission and a second surgery . However, if detected early enough, this can potentially be avoided. Moreover, early diagnosis of complications and adverse events is preferred as it decreases pain, discomfort, length of stay, readmission rate and potentially improves quality of life. 

In this project we will focus on prediction of postoperative complications such as surgical site infection, postoperative delirium and anastomosis leakage using the Electronic Health Records from the Department of Gastrointestinal Surgery at UNN. These EHRs contain both unstructured and structured data in the form of free text (admittance summaries, doctor and nurses notes, descriptions of surgical procedures, discharge summaries), blood samples and vital signs.  The current team has already demonstrated the potential of this project in several different scientific publications. 

Back to Projects