PIAA: Pre-imaging all-round assistant for digital radiography.

Area of Science:

• Medical Imaging Technology
• Artificial Intelligence in Healthcare
• Radiography Optimization

Background:

• Suboptimal radiographer technique in radiography leads to image retakes.
• Image retakes increase patient exposure to ionizing radiation.
• Reducing retakes is essential for patient safety and resource conservation.

Purpose of the Study:

• To introduce a Digital Radiography (DR) Pre-imaging All-round Assistant (PIAA) system.
• To leverage Artificial Intelligence (AI) for enhancing traditional DR procedures.
• To minimize unnecessary radiation exposure and improve workflow efficiency.

Main Methods:

• The PIAA system integrates an RGB-Depth multi-camera array and an embedded computing platform.
• Key modules include Adaptive RGB-D Image Acquisition (ARDIA) and 2.5D Selective Skeletal Keypoints Estimation (2.5D-SSKE).
• A Domain expertise embedded Full-body Exposure Parameter Estimation (DFEPE) module utilizes 2.5D-SSKE and domain expertise for accurate parameter estimation.

Main Results:

• The PIAA system optimizes the DR workflow, enhancing operational efficiency.
• Average patient positioning and exposure parameter preparation time reduced from 73 seconds to 8 seconds.
• Demonstrated significant improvements in workflow speed and potentially reduced retake rates.

Conclusions:

• The PIAA system shows significant promise for improving digital radiography.
• The technology is well-suited for extension to full-body examinations.
• AI-driven solutions can enhance patient safety and efficiency in medical imaging.