What is it about?
We used machine learning to accurately detect catheter tip positions in neonatal lung x-ray reports. We benchmarked different machine learning classifiers to identify if the tip locations are in the appropriate place or not. Further, we have classified 12 different locations the tip positions are in using the radiology reports.
Featured Image
Photo by Sharon McCutcheon on Unsplash
Why is it important?
In critically ill infants, the position of a peripherally inserted central catheter (PICC) must be confirmed frequently and the tip may move from its original position. We used natural language processing (NLP) and supervised machine learning (ML) techniques to predict PICC tip position based primarily on text analysis of radiology reports of patients in a neonatal intensive care unit (NICU). Our results shows that ML classifiers can automatically extract the anatomical location of PICC tips from radiology reports with high accuracy. Implementing these algorithms in a NICU as a clinical decision support system may help clinicians address PICC line positions.
Perspectives
In this paper we attempted to utilize natural language processing (NLP) and supervised machine learning (ML) techniques for predicting catheter positions in a NICU setting. Clinical implementation of these AI algorithms can augment decision support systems which can aid clinicians to avoid the risk of hyperosmolar vascular damage or extravasation into surrounding spaces for wrongly placed catheter lines.
Surya Prasath
Cincinnati Children's Hospital Medical Center
Read the Original
This page is a summary of: Machine Learning for Detection of Correct Peripherally Inserted Central Catheter Tip Position from Radiology Reports in Infants, Applied Clinical Informatics, August 2021, Thieme Publishing Group,
DOI: 10.1055/s-0041-1735178.
You can read the full text:
Resources
Contributors
The following have contributed to this page
