HCR.20.080 – Combining e-Health and artificial intelligence to improve surgical wound care management.

Route: Health care research, sickness prevention and treatment

Cluster question: 105 How can Big Data and technological innovation (e-health) contribute to health care?

Regardless of the type of surgical intervention, all invasive modalities carry a risk of surgical site infection (SSI). Despite the improvement of surgical technique, (prophylactic) antibiotic therapy and other post-operative care, SSI still results in increased morbidity, mortality and healthcare costs. Depending on the severity these costs vary from several hundreds to many thousands of dollars. SSI after an ambulatory surgical procedure result in readmissions, acute care visits, increased length of in hospital stay, reduction in quality of life and a multi-billion dollars cost worldwide.
Early detection of SSIs is paramount in order to prevent or reduce the severity and negative associated effects. Patients often lack the ability to self-diagnose these infections. Most SSI occur after discharge and follow-up is needed to detect these wound infections along with feedback for quality of care.
For obvious reasons there is a mismatch in occurrence of SSI and standard planned follow-ups. Additionally, healthcare workers often lack the skills to recognize signs of SSI. This results in the unnecessary prescription of antibiotics or undertreatment with an exacerbation of the infection resulting in a hospital admission with its related expenses and complications. Early detection and adequate treatment of SSI is needed, thereby preventing exacerbation of the infection. The limited effectiveness of face-to-face appointments to detect SSIs is a call for a more effective monitoring tool.
In this project, we propose to combine telemedicine, a safe and effective method for self-monitoring tool, and artificial intelligence (AI), a technique already capable of outperforming clinicians in the detection of melanomas and pathological nodules on mammography.
We therefore hypothesize that detection and prediction of SSIs with AI, trained with standardized surgical site photos, can be an equivalent or superior to physicians for detecting post-operative wound infections. Thereby reducing both under- and overtreatment of SSIs.

Keywords

Artificial Intelligence, Big Data, e-health, overtreatment, Surgical Site Infection, undertreatment

Submitter

Organisation Radboudumc (RUMC)
Name Prof. dr. R.H.M.A. (Ronald) Bartels
E-mail Ronald.Bartels@radboudumc.nl