Initial Survey Results on the Accuracy of Dose Data in the DICOM Objects

Transcription

1 Initial Survey Results on the Accuracy of Dose Data in the DICOM Objects Renato Padovani, Annalisa Trianni University Hospital S. Maria della Misericordia Udine, Italy

2 INTRODUCTION The EC MED has forced standardisation bodies and manufacturers to provide patient dose information to the operator. DICOM has developed standards aiming to store and distribute dose data. Today, dose to patients and population are usually assessed on the base of dosimetric indicators provided by the equipment and exported to databases. These advancements are increasing awareness on radiation doses and are supporting efforts on procedure optimisation

3 ACCURACY OF DOSE INDICATORS Limiting accuracy of dose indicators are established by IEC norms: Interventional: 50% for KAP and CK (CEI EN ; 2002) CT: 20% for CTDI w (CEI EN ; 2007); No indication for DLP Mammography: No indication for AGD

4 DICOM Standard DICOM standard provides encoding definitions for a number of dosimetric indicators: CR-DR Exposure Index Mammography AGD, ESAK Fluoroscopy and Interventional KAP and CK CT DLP, CTDI vol and Effective Dose Sources of variability and inaccuracy: Flexibility allowed by DICOM Different estimation models and internal calibrations used Calibration/maintenance protocols

5 DICOM Standard Dosimetric information could be archived together with images 3 solutions: DICOM header DICOM Modality Performed Procedure Step (MPPS) Radiation Dose Structured Report (RDSR)

6 DICOM Header limits Mostly Exposure details (for interpretation) No image (fluoro, deleted) no dose Lacks complete dose details See Conformance statement to know data provided

7 MPPS limits Modality Performed Procedure Steps (MPPS) Status messages back to RIS / PACS Transient Information designed for workflow; not persistent archiving Lacks of complete dose details

8 Radiation Dose Structured Report (RDRS) An irradiation object be stored for each irradiation event, irrespective of the storage of the images produced by that irradiation. The irradiation objects, along with other information, is stored in a Radiation Dose Structured Report (RDSR). The RDSR could be archived in the RIS, or PACS, or transferred to a Radiation Safety Reporting System (RSRS).

9 IHE Radiation Exposure Monitoring Profile (REM)

10 QUALITY ASSURANCE Dosimetric information are given to patients and sent to hospital/national databases. Up to date no Quality Assurance protocol has been established to verify completness and accuracy of the data registered in the DICOM datasets.

11 AIM The purpose of the work was to explore the accuracy of the dosimetric information provided in the DICOM data objects once the equipment is declared calibrated by the manufacturer and to underline the needs of a Quality Assurance protocol for dose data records. DICOM dose data have been compared with data measured with calibrated dose meters.

12 RESULTS Computed Tomography Model Dose index Console Measured Diff (%) GE Lightspeed VCT GE Lightspeed 750HD nctdiw (mgy/mas) 120 kv Head Body Head Body

13 RESULTS Interventional Model KAP Console (mgycm 2 ) AGD Measured* (mgycm 2 ) Diff (%) Siemens Artis Zee Siemens Artis Zee Philips Allura

14 DISCUSSION Interventional Differences due to: Energy dependence showed by the KAP meters Different calibration procedures used by the manufacturers.

15 RESULTS Mammography Model AGD Consolle (mgy) AGD Measured* (mgy) Diff (%) Fuji Amulet GE Senographe essential Hologic Lorad Selenia Hologic Lorad Dimension Siemens Mammomat Inspiration Philips Mammodiagnost Calculated using method proposed by Dance et al. (Phys. Med. Biol. 2009)

16 DISCUSSION Mammography Large differences (up to 23%) because AGD is calculated using different dose calculation algorithms: Dance Wu & Barnes US Method No information concerning the methodology used is reported in the DICOM objects

17 CONCLUSION Accuracy requested to dosimetric indicators registered in DICOM structures is established in few cases. When present, limits are really large and methods of measurement not agreed. In case of calculated dose indicators a variability of methods could be used and are not reported in the DICOM objects

18 OPEN ISSUES Even if in the newest equipments dosimetric data recorded seem to be enough accurate, further developments and rules should be introduced concerning this concepts. Medical physicists have to assure the quality of the dose information provided and distributed.

19 OPEN ISSUES On going works DICOM WG-28 White paper containing additional doserelated information, to extend existing RDSR and taking into account accuracy needed. IEC PT and DICOM CP 1232 include more than one KAP meters calibration coefficients in the standard MITA X-Ray Interventional Physics Mode. KAP meters calibration factors by users IAEA CRP to extend TRS 457 to DICOM dosimetry AAPM Task Group No Determination of a Patient Size Metric for CT Size Specific Dose Estimation