Eingabe löschen

Kopfbereich

Schnellnavigation

Hauptnavigation

Phantom- and Model-based Optimisation of Medical Imaging Modalities

Auf einen Blick

Beschreibung

Optimisation of CT protocols is important to ensure a sufficient image quality with a radiation exposure as low as possible. Recently, image quality and radiation exposure (especially dose-length product DLP) are simultaneously evaluated by the KSA group using the ZHAW elliptical phantom. Complementary to this information, patient exposure data (CTDI, DLP) from CT units can be collected automatically by dedicated software such as DIDT or GE DoseWatch. The calibration of the dose indications of the CT unit can be checked by a CTDI phantom. In consequence, DLP and CTDI are available for patients and phantom measurements for a large variety of clinical protocols and CT units. For the same protocols, image quality data is only available for the phantom measurements. To elaborate a base line for of the CT protocols concerning image quality and dose, the collection of intrinsic image quality of patient scans is required to complete the information needed for a quantitative basis form optimisation. It is hard to measure image quality parameters such as SNR, CNR or MTF in patient images due to homogeneity reasons. Several approaches have been investigated in past. If only a comparison of image quality is requested, an estimate of these parameters may be sufficient. Such estimates should indicate clinical relevant aspects of image quality and must be robust. Using morphological knowledge of a standard anatomy and contour based image transformation, ROI’s with a defined gradient of the mean CT numbers can be identified. By subtraction of the background inhomogeneity, SNR (and for 2 ROI’s the CNR) can be evaluated. The aim of this project is a feasibility study of the practicability and robustness of different approaches by introducing novel physical and numerical phantoms combined with dedicated algorithm for image quality assessment. Dosimetric aspects of Radiation exposure will be evaluated by Monte-Carlo (MC) simulations.


Publikationen