A machine learning framework for auto classification of imaging system exams in hospital setting for utilization...

Area of Science:

• Medical Imaging
• Machine Learning in Healthcare
• Radiology Workflow Optimization

Background:

• Interventional X-Ray (IXR) systems are crucial in clinical settings for diverse procedures and anatomies.
• Accurate classification of IXR exams is vital for efficient scheduling, resource management, and revenue forecasting.
• Current methods struggle due to identical system settings used for different procedures, leading to underutilization.

Purpose of the Study:

• To develop a machine learning-based method for classifying IXR exams by anatomical type.
• To enhance the productivity and utilization of IXR systems in clinical environments.
• To provide hospital management with better data for forecasting and treatment planning.

Main Methods:

• Utilized machine learning techniques including Support Vector Machines (SVM), K-Nearest Neighbors (KNN), and decision trees.
• Applied these algorithms to analyze system log information from IXR machines.
• Focused on classifying exams based on the anatomical region being examined.

Main Results:

• Achieved a classification accuracy exceeding 90% for assigning IXR exams to their respective anatomical types.
• Demonstrated the effectiveness of machine learning in analyzing system logs for improved classification.
• The developed method shows promise for practical implementation in hospital settings.

Conclusions:

• Machine learning offers a robust solution for accurate IXR exam classification.
• This classification significantly enhances operational efficiency and financial planning in healthcare facilities.
• The study highlights the potential of leveraging system log data for optimizing medical imaging workflows.