Risk predictions of surgical wound complications based on a machine learning algorithm: A systematic review
Hui Zhang1, Junde Zhao2, Ramyar Farzan3
1The Second Clinical Medical School, Lanzhou University, Lanzhou, China.
International Wound Journal
|January 25, 2024
Summary
Machine learning (ML) shows promise in predicting and managing surgical wounds, improving patient outcomes. This AI application aids in evaluating surgical site infections and wound classification across diverse procedures.
Area of Science:
- Medical Informatics
- Artificial Intelligence in Surgery
- Wound Healing Research
Background:
- Surgical wounds can lead to severe complications, prolonged hospitalization, and adverse clinical outcomes.
- Machine learning (ML), a subset of artificial intelligence (AI), is increasingly applied in healthcare for prediction and diagnosis.
- Effective surgical wound management is crucial for patient recovery and reducing healthcare burdens.
Approach:
- A systematic review following PRISMA guidelines was conducted.
- Searches included terms like 'machine learning', 'surgical', and 'wound' in electronic databases.
- Studies from 1990 to present focusing on ML in surgical wound evaluation were analyzed, excluding reviews and guidelines.
Key Points:
- Nine articles utilizing ML for surgical wound management were identified.
- ML applications included surgical site infection (SSI) evaluation (7 studies) and burn-grade diagnosis/wound classification (2 studies).
- Support Vector Machine (SVM) and Convolutional Neural Network (CNN) were the most common ML algorithms, with high reported accuracies.
Conclusions:
- ML algorithms demonstrate significant potential in enhancing surgical wound management.
- These AI tools can improve postoperative care and lead to better clinical outcomes.
- Further development of ML in surgical wound assessment is warranted for advanced patient care techniques.
