Register for Updates | Search | Contacts | Site Map | Member Login

ICRP: Free the Annals!

View Comment

Submitted by David Clunie, PixelMed
   Commenting as an individual
Document Diagnostic Reference Levels in Medical Imaging
Commenting individually but from my perspective as the editor of the DICOM Standard and member of the IHE Radiology Technical Committee and an author of open source software for dose-related information collection.

page 44 section 2.5. Data collection methods (para 98 and 99)

Though this section is (appropriately) brief, the emphasis and initial description should perhaps focus on RDSR rather than MPPS, if indeed MPPS needs to be mentioned at all.

RDSR is used in the IHE Radiation Exposure Monitoring (REM) profile [1], is used by the ACR CT Dose Index Registry (DIR), is the only mechanisms tested at IHE Connectathons [2], and is the only mechanism supported in the NEMA XR-29 CT standard [3] and the corresponding US regulation related to reduced reimbursement for obsolete equipment [4]. Though adoption of RDSR by modalities has been faster for CT, it is now seen in XA/XRF and mammography equipment.

The mention of MPPS should be relegated to a historical curiosity, since it was only rarely implemented (with dose information) and it is expected to be entirely replaced by DICOM RDSR; so if it is mentioned at all perhaps it should be with the caveat that it is a "dead end" and should be avoided if possible.

The need to extract technique information from DICOM image "headers" is perhaps worth mentioning, since that is far more often necessary than is the use of MPPS, especially for plain X-Ray applications. The slow adoption by X-Ray vendors of standard exposure index values [5], despite their availability in the DICOM standard, might be mentioned.

OCR of Dose Screen Images [6,7,8,9], is unfortunately necessary/useful for some of the older CT equipment that cannot be upgraded to support RDSR directly, but even then, the output of the OCR process is best encoded as RDSR to maximize interoperability with different data collection and registry systems.

A crucial data collection integrity issue worth highlighting is the difficulty of reliably identifying which procedures (exam type) were performed in a manner that allows cross-site comparison, and the need to perform site-specific mapping of local codes and descriptions to a standard agreed upon short list for data aggregation [10]. Usually overlooked is the matter of performing quality control of such mappings, particularly when the images are not collected and cannot be reviewed retrospectively.

On the subject of images, it might be mentioned that the ACR DIR has found it useful to collect localizer images to use for automated patient size estimate corrections [11].

DICOM has also recently completed a Radiopharmaceutical Radiation Dose Structured Report (R-RDSR) template [12] and this is being included in an updated IHE REM-NM Profile [13]. This will allow not only capture of the delivered dose from any nuclear medicine procedure, but particularly standardized recording of the radiopharmaceutical part of the dose of a PET/CT study.

Finally, though it may belong elsewhere in the report, one wonders about the selection bias that is implicit in choosing one data collection mechanism over another, particularly when the chosen mechanism is only available on newer (lower dose?) devices as opposed to a representative sample of the entire installed base. This may be less important if the desire is to produce DRLs that are at a desired lower level rather than being "representative" of the entire population being sampled.


[1] IHE. Radiation Exposure Monitoring (REM) profile. ""

[2] IHE. Connectathon Results Browser. ""

[3] NEMA. XR 29 - Standard Attributes on CT Equipment Related to Dose Optimization and Management. 2013. ""

[4] US Congress. H.R. 4302 Title II, Section 218, congress amended Section 1834 of the Social Security Act by adding a portion titled “Quality Incentives to Promote Patient Safety and Public Health in Computed Tomography. ""

[5]  IEC. IEC62494-1 Medical electrical equipment - Exposure index of digital X-ray imaging systems - Part 1: Definitions and requirements for general radiography. 2008. 

[6] Clunie, DA. Dose Matters. 2010.

[7] Li X, Zhang D, Liu B. Automated Extraction of Radiation Dose Information From CT Dose Report Images. American Journal of Roentgenology. 2011 Jun 1;196(6):W781–3. doi:10.2214/AJR.10.5718

[8] Cook TS, Zimmerman S, Maidment ADA, Kim W, Boonn WW. Automated Extraction of Radiation Dose Information for CT Examinations. Journal of the American College of Radiology. 2010 Nov;7(11):871–7. doi:10.1016/j.jacr.2010.06.026

[9] Warden GI, Sodickson A, Farkas C, Prevedello L, Andriole K, Khorasani R. Exposing Exposure: Automated Anatomy-specific CT Radiation Exposure Extraction from Existing Image Repositories. In 2011 [cited 2014 Aug 19]. Available from:

[10] Escalon JG, Chatfield MB, Sengupta D, Loftus ML. Dose Length Products for the 10 Most Commonly Ordered CT Examinations in Adults: Analysis of Three Years of the ACR Dose Index Registry. Journal of the American College of Radiology. 2015 Aug;12(8):815–23. doi:10.1016/j.jacr.2015.02.022

[11] Bhargavan-Chatfield M, Morin RL. The ACR Computed Tomography Dose Index Registry: The 5 Million Examination Update. Journal of the American College of Radiology. 2013 Dec;10(12):980–3. Available from:

[12] DICOM. Supplement 159. Radiopharmaceutical Radiation Dose Reporting. 2014. Available from:

[13] IHE. Radiation Exposure Monitoring for Nuclear Medicine. Available from: