Related Experiment Video
Updated: May 28, 2026
08:22
Application of Optical Coherence Tomography to a Mouse Model of Retinopathy
Published on: January 12, 2022
Automated assessment of low contrast sensitivity for CT systems using a model observer
I Hernandez-Giron1, J Geleijns, A Calzado
1Física Mèdica, Facultat de Medicina i Ciències de la Salut, Universitat Rovira i Virgili, 43201 Reus, Spain. irene.debroglie@gmail.com
Medical Physics
|October 8, 2011
Summary
This study introduces an automated software using a model observer to objectively measure CT scanner low contrast detectability. The software provides reliable results consistent with human observers, aiding in image quality assessment.
Area of Science:
- Medical Imaging
- Radiology
- Image Quality Assessment
Background:
- Current CT scanner low contrast sensitivity assessment relies on subjective scoring of phantom images.
- Subjective scoring can be biased due to fixed patterns of low contrast objects in phantoms.
- Objective and automated methods are needed for reliable low contrast detectability evaluation.
Purpose of the Study:
- To develop and validate software for automated, objective assessment of low contrast detectability in CT images.
- To utilize a model observer for quantifying low contrast object visibility.
- To establish a reliable method for CT image quality evaluation.
Main Methods:
- Employed the Catphan 600 phantom's low contrast module for software evaluation.
- Utilized a non-prewhitening matched filter with an eye (NPWE) model observer for automated analysis.
- Calculated a discrimination index (d') transformed into proportion correct (PC) values, with PC ≥ 75% indicating visibility.
Main Results:
- Automated method showed expected trends: increased low contrast detectability with higher kVp and mAs, and with soft reconstruction filters.
- High contrast objects (1%) were consistently detected above the 75% threshold for diameters >2 mm.
- Lower contrast objects (0.3%) detection was dependent on object size and radiation dose (mAs).
Conclusions:
- An automated, objective method for assessing CT image quality using a NPWE model observer and the Catphan phantom was successfully developed.
- The software accurately evaluated low contrast detectability across various acquisition and reconstruction parameters.
- This method holds potential for evaluating image reconstruction algorithms, dose reduction strategies, and new CT technologies.
