Automated Web-based Software for CT Quality Control Testing of Low-contrast Detectability using Model Observers

Area of Science:

• Medical Imaging
• Radiology
• Image Analysis

Background:

• Channelized Hotelling observer (CHO) correlates with human performance in CT tasks.
• Widespread adoption in CT quality control is limited by a lack of accessible, validated software.
• Existing tools often require complex installation and are not user-friendly for routine use.

Purpose of the Study:

• To develop an automated, web-based software platform, CT image quality evaluation and Protocol Optimization (CTPro), for enhanced CT image quality assessment.
• To provide easy access to CHO analysis and traditional image quality metrics (e.g., modulation transfer function, noise power spectrum).
• To facilitate efficient and accurate image quality evaluation for both research and clinical communities without software installation.

Main Methods:

• Development of CTPro, a web-based platform integrating CHO with traditional image quality assessment tools.
• Utilized modulation transfer function and noise power spectrum for comprehensive image quality evaluation.
• Applied the platform to compare low-contrast detectability between a photon-counting-detector (PCD)-CT and an energy-integrating-detector (EID)-CT.

Main Results:

• CTPro provides a user-friendly interface for CHO analysis and other image quality metrics.
• The application demonstrated a 6.2% higher d' for UHR-T3D on PCD-CT compared to EID-CT with IR (p = 0.047).
• A 4.1% lower d' was observed for UHR-T3D on PCD-CT compared to EID-CT without IR (p = 0.122).

Conclusions:

• CTPro offers a validated, automated solution for CT image quality evaluation, improving accessibility to CHO analysis.
• The platform enables efficient and accurate performance assessment, supporting protocol optimization.
• The study highlights the potential of PCD-CT for improved low-contrast detectability, validated through CHO analysis via CTPro.