Medical Imaging Cardiac CT Imaging Medical Image Processing Computer Assisted Surgery Radiation Therapy
Osteoporosis Research Dose Assessment/Dose Reduction

Dr. Martin Hupfer
Dr. Daniel Kolditz
Prof. Dr. Willi Kalender


Dose Assessment - Dose Reduction


X-ray computed tomography (CT) is one of the diagnostic procedures that can cause relatively high radiation exposure to the patient. In the last two decades the increasing use of CT in clinical practice has made CT a significant contributor to the total collective dose from medical x-ray examinations. With the introduction of multi-slice CT (MSCT) scanners a decade ago, the number of clinical applications of CT procedures has continued to increase due to the greater volume coverage (e.g. 64 slices) and faster rotation times providing significantly higher performance. CT was estimated to contribute around 67% of the total collective medical dose in the USA and about 40% in the UK and Germany, although CT examinations represent only about 11% and 4% of all radiology examinations in the USA, the UK and Germany, respectively. For this reason, CT is subject to the particular scrutiny of radiation protection measures; there are demands to limit or reduce dose levels arising from the use of CT. The Institute of Medical Physics (IMP) is active in the field of dose determination and dose reduction. The key aspects of these activities are listed below.


Dose assessment

Monte Carlo calculations

Three-dimensional dose distributions for arbitrary CT scanners and scan parameters can be calculated by using Monte Carlo (MC) techniques. In addition to scanner-specific parameters like the x-ray spectrum and shaped filters, patient specific parameters such as size, weight and anatomy can be taken into account as well. See [Deak 2006, Schmidt 2002, Schmidt 2001] for a detailed description of the Monte Carlo methods in use at the IMP.








Screenshot of ImpactMC: A PC program to estimate three-dimensional dose for arbitrary scanners and scanning protocols.


Organ and effective dose estimates

While MC techniques can determine dose levels to within a few percent they are time consuming and such high accuracy is not always necessary. The usage of pre-calculated dose tables based on standard voxelized phantoms can provide instant organ and effective dose estimates. ImpactDose is such a program aiming at the estimation of gender and age-specific organ and effective dose for arbitrary scan parameters and anatomical ranges, see [Kalender 1999b].







Screenshot of ImpactDose.
A PC program for estimation of organ and effective doses.


In vitro dose measurements

For validating the Monte Carlo simulations, dose is also measured on the IMPís CT scanners. For the measurement of absolute dose, ionization chambers of different lengths and sizes are used. For the determination of dose profiles and dose distributions, we use our own thermoluminescence dosimeters (TLDs) and readout laboratory (see figure below for an example).






Comparison of Monte Carlo simulations with in vitro measurements of z-profiles at the centre of a CTDI body phantom (tube voltage 80 kV, beam collimation 10 mm).



Dose reduction

Optimization of tube-voltage and/or x-ray spectrum

CT scanners can be operated with a variety of x-ray tube voltage and thus x-ray spectra. Changing the x-ray spectrum influences both image quality and patient dose. Simulations were performed at the IMP in order to determine the relationship between image contrast and image noise as a function of the x-ray spectrum. The dose was estimated by using Monte Carlo techniques. For this purpose, several phantoms were investigated, including pediatric sized phantoms. The optimal spectrum was determined for several applications (both pediatric and adult CT) and various diagnostic questions (including contrast-enhanced studies, see figure below) [Kalender 2006]. The optimization of spectra will continue for various applications and include the investigation of different filter materials.

a) b) c)

Spectrum optimization for paediatric CT. (a) Simulated CT scan of a paediatric chest phantom (10x15 cm2) with iodine insert. (b) Calculated dose distribution. (c) Contrast-to-noise ratio (in contrast-enhanced CT studies) normalized to the square root of dose as a function of tube voltage. In this case, dose can be considerably reduced by using a low x-ray tube voltage.


Simulation of low-dose CT scans

Automatic exposure control in CT aims at keeping the image quality constant when scanning patients of different sizes. Ideally the user specifies only image quality parameters, in particular 3D spatial resolution and noise level. It is then the task of the CT system to choose the scan parameters which guarantee that the respective image quality is provided at minimal patient dose. It is therefore necessary to achieve a consensus on the image noise level and 3D spatial resolution required for a given examination type.

The IMP has developed a software tool, called "DoseTutor", which can demonstrate the relationship between the scan parameters, image quality and patient dose for real patient data [Leidecker, 2004, Van Straten 2007].





Screenshot of DoseTutor, a PC program to simulate CT scans with changing dose or changing reconstruction kernel. When changing one of these parameters, the program interactively shows the effect on the dose and image quality.

In the figure below, four images are shown for different dose levels and spatial resolutions. The original image (a CT scan of the abdomen administered with intravenous contrast agent) is shown in the upper right corner. This scan was made using routine dose levels and reconstructed with a sharp reconstruction kernel. The other three images were generated with DoseTutor.

They differ from the original with respect to patient dose and/or spatial resolution. These results show that a reduction of dose (lower right image) at constant spatial resolution affects the visibility of the vessels in the liver, due to the increase of noise. The visibility of these low contrast structures can (partly) be restored by decreasing the spatial resolution (lower left image), while keeping the dose constant.







Relationship between spatial resolution, image noise and dose.


Angular tube current modulation

Angular tube current modulation is a technique to reduce the exposure without reducing image quality [Gies 1999, Kalender 1999a]. We investigated these dose reduction techniques by simulation and experiment. The potential of reduction was also determined and validated in clinical tests (e.g. anatomy-adapted tube current modulation) on the IMP's own MSCT scanner. The effect of angular tube current modulation on dose distribution was investigated as well [Van Straten 2006a].








Angular tube current modulation leads to both an exposure and a dose reduction without affecting the image quality.

At the IMP, the tube current modulation concept was adapted to include so called partial scan techniques at which the x-ray tube has virtually no output in the anterior-posterior direction. This might be advantageous in a CT scan of the female thorax [Vollmar 2006]. Our evaluation (see figure below) addressed both the dose to the patient and the image quality. In addition, an evaluation of approaches to dedicated CT of the breast is in process and will be continued in the future.

a) b) c)
Dose reduction to the female breast in chest CT. (a) Simulated CT scan of an anthropomorphic chest phantom. (b) Dose distribution in case of bismuth shielding. (c) Dose distribution in case of partial scanning. In both cases the dose to the breasts is reduced. Image quality, however, is deteriorated significantly by bismuth shielding (not shown).


Automatic exposure control (AEC)

The human anatomy changes with the z-position for a given patient and the size of a patient differs strongly for various patients. Therefore, the exposure has to be adapted if a constant level of image quality is to be achieved (for all z-positions and patient sizes). The automatic adaptation of the exposure is readily available in state-of-the-art CT scanners. Angular tube current modulation reduces dose without affecting the image noise level (see above), z-modulation or automatic exposure control increases/decreases the dose in order to keep a (almost) constant image noise level. It will depend on prior practice, whether AEC will lead to a dose reduction or not. At the IMP the influence of AEC on patient-specific dose distributions and organ doses is investigated with the aid of Monte Carlo techniques and voxelized patient models of varying sizes [Van Straten 2006b].




References and selected publications

Kyriakou Y, Kolditz D, Langner O, Krause J, Kalender WA. Digitale Volumentomografie (DVT) und Mehrschicht-Spiral-CT (MSCT): eine objektive Untersuchung von Dosis und Bildqualität. Digital Volume Tomography (DVT) and Multislice Spiral CT (MSCT): an Objective Examination of Dose and Image Quality. Fortschr. Röntgenstr. 2011; 183: 144-153

Deak P, Smal Y, Kalender WA. Multisection CT Protocols: Sex- and Age-specific Conversion Factors Used to Determine Effective Dose from Dose-Length Product. Radiology 2010; 257: 158-166

Kolditz D, Kyriakou Y, Kalender WA. Volume-of-interest (VOI) imaging in C-arm flat-detector CT for high image quality at reduced dose. Med. Phys. 2010; 37(6): 2719-2730

Vollmar S, Kalender WA. Reduction of dose to the female breast as a result of spectral
optimization for high-contrast thoracic CT imaging: A phantom study. Brit J of Radiology 2009; 82:920-929

Deak P, Langner O, Lell M, Kalender WA. Effects of Adaptive Slice Collimation on Patient Dose in Multi-Slice Spiral Computed Tomography. Radiology 2009; 252(1): 140-147

Deak P, van Straten M, Shrimpton PC, Zankl M, Kalender WA. Validation of a Monte Carlo tool for patient-specific dose simulations in multi-slice computed tomography. Eur Radiol 2008; 18: 759-772

Deak P. Patient-Specific Dose Calculations in X-ray Computed Tomography using Monte Carlo Methods. Thesis, Institute of Medical Physics, University of Erlangen-Nuremberg 2006.

Deak P, Van Straten M, Kalender WA . Validation of a Monte Carlo (MC) based tool for patient-specific dose calculations in x-ray computed tomography. European Congress of Radiology, March 9-13 2007, Vienna , Austria .

Gies M, Suess C, Wolf H, Kalender WA, Madsen MT. (1999). Dose reduction in CT by anatomically adapted tube current modulation: I. Simulation studies. Medical Physics 26 (11): 2235-2247

Kalender WA, Wolf H, Suess C, Gies M, Bautz WA. (1999a). Dose reduction in CT by anatomically adapted tube current modulation: II. Phantom measurements. Medical Physics 26 (11): 2248-2253

Kalender WA, Schmidt B, Zankl M, Schmidt M. (1999b). A PC program for estimating organ dose and effective dose values in computed tomography. European Radiology 9: 555-562

Kalender WA . Computed Tomography; Fundamentals, System Technology, Image Quality, Applications. Publicis Corporate Publishing, Erlangen , second revised edition, 2005.

Kalender WA , Buchenau S. Optimal X-ray Energies in Pediatric CT with Regard to Contrast, Noise, and Dose. Radiological Society of North America, Scientific Assembly and Annual Meeting Program 2006, page 228.

Leidecker C, Kachelriess M, Kalender WA. A Dose Tutor to help optimize clinical scan protocols. European Radiology, Suppl. 1(14):274, 2004.

Schmidt B. Dosisberechnungen für die Computertomographie. Berichte aus dem Institut für Medizinische Physik, ed. W.A. Kalender. Vol. 7. 2001, Aachen: Shaker Verlag.

Schmidt B, Kalender WA. A fast voxel-based Monte Carlo method for scanner- and patient-specific dose caclulations in computed tomography. Physica Medica, 2002; XVIII (2) April-June: 43-53

Van Straten M, Deak P, Kalender WA. Automatic Exposure Control: How to Estimate Patient Organ Doses When Detailed Information on the Modulated Tube Current Is Not Available. Radiological Society of North America, Scientific Assembly and Annual Meeting Program 2006a, page 924.

Van Straten M, Deak P, Kalender WA. Patient-Specific Organ Dose Determination for Patients of Varying Size on CT Scanners with Automatic Exposure Control (AEC). Radiological Society of North America, Scientific Assembly and Annual Meeting Program 2006b, page 553.

Van Straten M, Staatz G, Kalender WA . Optimization of paediatric computed tomography (CT) examinations by simulation of low-dose CT scans with a dose tutor program. European Congress of Radiology, March 9-13 2007, Vienna , Austria .

Vollmar S, Deak P, Kalender WA. Reduction of Dose to the Female Breast in Chest CT : A Comparison of Standard Protocols, Tube Current Modulated, Partial, and Bismuth-shielded Scans. Radiological Society of North America, Scientific Assembly and Annual Meeting Program 2006, page 595.