Database of dynamic renal scintigraphy
Original database of dynamic renal scintigraphy 2012
(data set "drsprg" containing data of 107 adult patients)
1 Introduction
Database of dynamic renal scintigraphy is a collection of test data for
(1) development and testing of analytic methods and computer programs in renal nuclear medicine,
(2) standardization and quality assurance of nuclear medicine software dedicated to analysis of renal studies, and
(3) comparative clinical trials including national and international audits and accreditation.
A need of a large set of test data covering wide range of normal variants and abnormal cases for development and testing of analytic methods is self-evident. It can be witnessed by anybody ever involved in such a business.
Software testing is a complex subject and there are many different approaches to it. Use of test data to simulate the conditions under which the software is expected to work has always been an important aspect of testing. In nuclear medicine, both real (patient) and simulated data has been used for test purpose. Besides traditional mathematical and physical phantoms, other types of phantoms, namely software and hybrid phantoms have been developed under the European project COST B2 [1-4]. More recently, similar phantom for static renal studies has been proposed [5].
By definition, software phantoms are sets of gamma camera images obtained from patients with known diagnosis which means that the raw clinical image data are validated [1]. In a strict sense, data in this database differ from the software phantoms as they were defined by COST B2 in that their clinical validation has not been confirmed by a panel of experts. With respect to this database, clinical documentation (rather than validation) includes demographic, clinical, and laboratory data from the patients.
A need of standard data sets for standardization of analytic methods and performing comparative clinical trials, especially in renal nuclear medicine, is documented by poor results of national and international audits and comparative studies. Large scale UK audits on GFR measurement [6] and quantitative analysis of dynamic renal scintigraphy [7-10] demonstrated wide variations in the values of reported diagnostic parameters especially in the patients with abnormal kidney function and also reporting technically demandning parameters like renal transit times and description of parenchymal time-activity curves. Worldwide comparative study by Austrian authors has shown warning differences in the measurement of split renal function in children [11,12] - the simplest parameter frequently used by the urologist to decide on kidney surgery. Urgent need for better standardisation in assessment of renal drainage has been declared by the authors of inter-observer reproducibility study in children with hydronephrosis [13]. Somewhat better results have been reported with static renal studies [14,15]. A need for clinical audit in nuclear medicine due to its complex technical aspects has been discussed for a long time [16-18].
In effort to approach the above requirements, clinically documented image data of dynamic renal scintigraphy collected under the research project supported by the Czech Science Foundation grant no. 303/07/0950, are offered to the researches and nuclear medicine professionals with the aim to improve standardisation and quality of nuclear medicine services in nephrology and urology.
It is assumed that, in the future, the database will be extended by additional data sets from other interested individuals and institutions. In case of such interest, the webmaster of this webpage or its medical custodian can be contacted.
2 Data in the database
The data included in the database had been recorded in 2009 as part of nuclear medicine services of the General University Hospital in Prague under the research project of the First Faculty of Medicine, Charles University Prague, supported by the Czech Science Foundation grant no. 303/07/0950. Image data are available in DICOM and INTERFILE formats. Supporting data are available in the text, pdf, and Microsoft Excel formats. All patients had been examined for medical reasons, all examinations were indicated by clinical nephrologists and urologists, and all data are strictly anonymized. The project and the database has complied with local ethical committee requirements and are presented with permission of the General University Hospital. The database and its contact web page has been created with financial support of M.G.P. s.r.o. Zlin, Czech Republic, under the agreement between M.G.P. and the General University Hospital in Prague.
3 Demographic data
At the moment (March/April 2012), the database contains raw image data of 107 dynamic renal studies performed in 45 men and 62 women in the age of 15 - 86 years.
patients | n | mean | st.dev. | min | max |
---|---|---|---|---|---|
age [years] | 107 | 53.1 | 18.4 | 15 | 86 |
weight [kg] | 107 | 80.0 | 19.0 | 45 | 143 |
height [cm] | 107 | 168.4 | 9.7 | 145 | 192 |
BMI [kg m-2] | 107 | 27.4 | 5.8 | 17 | 52 |
BSA-Haycock [m2] | 107 | 1.917 | 0.270 | 1.367 | 2.689 |
BSA-Dubois [m2] | 107 | 1.873 | 0.239 | 1.370 | 2.562 |
Most patients suffered from various stages of chronic renal disease, diabetic and ischaemic nephropathy, recurrent infections, chronic pyelonephritis, nephrosclerosis, polycystic kidneys, carcinoma of urinary bladder, and other pathological conditions specified for individual patients in the tables attached to every set of downloaded data.
The studies have been recorded in routine clinical practice using standard clinical protocol of the department of nuclear medicine extended by recording simultaneous anterior projection of the kidneys for geometric mean calculations and by a short record of transmission data for measured attenuation correction. In some patients, simultaneous examination of GFR using 51Cr-EDTA and additional blood samples have been collected in case of medical indication.
4 Image data
Dynamic renal scintigraphy was recorded in well hydrated patients after injection of 145 - 301 (198 ± 26) MBq 99mTc-MAG3 in supine position with double-head gamma camera Infinia or Infinia-Hawkeye (GE Healthcare) equipped with LEHR collimator. In 54 patients, up to 3 MBq of 51Cr-EDTA had been administered simultaneously. Data acquisition started before injection. Field of view was set to include the kidneys and the heart. Images in both posterior and anterior projections were recorded in 128x128 matrix in 10s frame intervals for 30 minutes. Patients voided shortly before the study and then again after the study was finished. Post-voiding images had been recorded in most patients. In 27 patients, 40 mg of furosemide was applied in 13 - 22 min. In 76 patients, transmission data were recorded using 57Co flood source before the dynamic renal study was started. The flood source was placed between the patient table and lower (posterior) detector and its posterior and anterior projections had been recorded for 1 minute into 256x256 matrix. Additional radiation dose to the patient was below 5 µSv. Anonymized image data in the database are stored in DICOM and INTERFILE image formats.
Minimum image data available in all patients in the database are both posterior and anterior projections of dynamic renal study. Additional data available for most but not all the patients are posterior and anterior projections of the flood source (transmission data) and post-voiding images in posterior and anterior projections. Minimum number of image data sets for individual patient is 4 (dynamic renal study and either post-voiding images or transmission data). Maximum number of image data sets in individual patient is 6 (dynamic renal study and both post-voiding images and transmission data).
Transmission data can be used for accurate attenuation correction in geometric mean of posterior and anterior projections. Attenuation factor of the patient table is 1.106.
5 Laboratory data
In all patients, a blood sample had been withdrawn from contralateral arm in 36 - 68 (48 ± 5) min and its activity measured 3-times in a well counter. In 54 patients in whom 51Cr-EDTA was also applied, 2 additional blood samples were withdrawn in 97 - 169 (135 ± 10) min and in 221 - 249 (239 ± 5) min. Average sample counts were corrected for background and decay. The database contains individual sampling times and average counts of all individual blood samples corrected for decay and background.
In 60 patients, serum creatinine and serum urea levels were available. Serum creatinine was 0.54 - 11.74 (2.44 ± 2.02) mg/100 ml or 48 - 1038 (216 ± 179) µmol/l. Serum urea was 16 - 286 (88 ± 58) mg/100 ml or 2.7 - 47.1 (14.5 ± 9.5) mmol/l. In 56 patients, creatinine clearance was available in the range of 3 - 177 (48.1 ± 41.2) ml/min or 0.05 - 2.95 (0.80 ± 0.69) ml/s. Of 56 patients in which it was available, creatinine clearance was below 15 ml/min in 13 patients, between 15-29 ml/min in 10, 30-59 ml/min in 18, 60-89 ml/min in 6, and greater than 89 ml/min in 9 patients.
In order to classify the patients into approximate chronic kidney disease (CKD) stages, GFR was estimated by regression from EDTA clearance available in some patients and MAG3 clearance available in all patients. Spectrum of approximate CKD stages is given in the following table.
GFR estimated | ≤ 15 ml/min | 16 - 30 ml/min | 31 - 60 ml/min | 61 - 90 ml/min | > 90 ml/min | total number |
---|---|---|---|---|---|---|
CKD stage (approximately) | 5 | 4 | 3 | 2 | 1 | 1-5 |
number of patients | 12 | 19 | 29 | 19 | 28 | 107 |
Relative function of the left kidney in 107 patients was 0 - 91 (49 ± 20) %.
6 Download
Downloading the data is subject to registration and licence agreement. Registration is required only for recording and administratition of the database and will not be shared with any third party. The users of data may or may not wish to be informed on future extensions, additions, and corrections of the database. Simple licence specifies conditions and responsibility related to the data download, use, distribution, and potential publication of data analysis.
The data to download can be selected from the list that appears after pushing one of the buttons "Browse database", or "Browse now" on the database homepage. The list of dynamic renal data includes
- name of the study (= name of the patient folder including several image data sets)
- patient gender (F female / M male)
- patient age in years
- CKD = approximate stage of chronic kidney disease (GFR estimated by regression from EDTA clearance available in some patients and single sample MAG3 clearance available in all patients)
- LK = approximate relative function of the left kidney (the values may somewhat differ using different methods of measurement and background subtraction especially in the patients with low renal function)
In all the following items, green mark = yes (available) / red mark = no (not available)
- SCR = availability of serum creatinine level
- CRC = availability of creatinine clearance
- MAG3 = availability of more than 1 blood sample with MAG3 measurement
- EDTA = availability of EDTA plasma clearance measured from 2 blood samples
- FLOOD = availability of transmission measurement with 57Co flood source
Using the patient's gender, age, CKD and LK, the list of patients can be sorted in either ascending or descending way. Alternatively, the database can be searched for specific values of these variables using the search tool above the list of patients.
Compressed data file for download includes patient image data and the tables with available clinical information (in all patients).
To download the data, please follow instructions on the screen. In case of problems please contact the webmaster of this webpage or its medical custodian (cf contacts).
7 Acknowledgement
The authors acknowledge with thanks a support by the following subjects
- Czech Science Foundation grant no. 303/07/0950
- General University Hospital in Prague
- First Faculty of Medicine, Charles University Prague
- M.G.P.s.r.o. Zlin, Czech Republic
8 Disclaimer
The data is provided "as is" without any cost. The provider and the authors disclaims all warranties, expressed or implied, including, without limitation, the warranties of merchantability and of fitness for any purpose, or noninfringement. Neither the provider, the author, nor anyone else who has been involved in the creation, production or delivery of the data assumes any liability for damages, direct or consequential, which may result from the use of the data. The entire risk arising out of use of the data remains with a user who shall defend, indemnify and hold the provider and the authors harmless from and against all liabilities, demands, damages, expenses, and losses, including attorney's fees, death, personal injury, illness, or property damage arising out of the use of the data. In no event the author or the provider be liable for any damages whatsoever (including, without limitation, damages for loss of business profits, business interruption, loss of business information, or any other pecuniary loss) arising out of the use of or inability to use the data, even if the provider and the author have been advised of the possibility of such damages.
The further details are given in licence agreement.
9 References
[1] Cosgriff PS (editor). COST B2 final report. Quality assurance of nuclear medicine software. EUR 17916 EN. European Communities, Directorate-General Science, Research and Development, Luxembourg, 1997 (ISBN 92-828-1501-3).
[2] Britton KE, Vauramo E. The quality assurance of nuclear medicine software. Eur J Nucl Med 1993;20:815-816.
[3] Bergmann H, Buseman-Sokole E, Horton PW. Quality assurance and harmonisation of nuclear medicine investigations in Europe. Eur J Nucl Med 1995;22:477-480.
[4] Samal M, Bergmann H. Hybrid phantoms for testing the measurement of regional dynamics in dynamic renal scintigraphy. Nucl Med Commun 1998;19:161-171.
[5] Peel SA, Dawson AH, Fleming JS, Hoffmann SM, Papaspyrou L. A technique for the simulation of planar radionuclide images of the kidney. Nucl Med Commun 2007;28:305-313.
[6] Cosgriff PS, Fleming JS, Jarritt PH, Skrypniuk J, Bailey D, Whalley D, Houston A, Burniston M. UK audit of glomerular filtration rate measurement in 2001. Nucl Med Commun 2008;29:511-520.
[7] Houston AS, Whalley DR, Skrypniuk JV, Jarritt PH, Fleming JS, Cosgriff PS. UK audit and analysis of quantitative parameters obtained from gamma camera renography. Nucl Med Commun 2001;22:559-566.
[8] Fleming J. Software audit in nuclear medicine. Nucl Med Commun 2000;21:206 (abstract).
[9] Nijran KS, Fleming JS, Skrypniuk JV, Whalley DR, Jarritt PH, Houston AS. UK audit of quantitative parameters obtained from a physical phantom generated renography. Nucl Med Commun 2005;26:280 (abstract).
[10] Nimmon CC, Fleming JS, Samal M. Probable range for whole kidney mean transit time values determined by reexamination of UK audit studies. Nucl Med Commun 2008;29:1006-1014.
[11] Staudenherz A, Dobrozemsky G, Krall Ch, Samal M, Schaffarich PM, Sinzinger H. Reproducibility of renal split function. Presented at the 14th International Symposium Radionuclides in Nephrourology, Mikulov, Czech Republic, 11-14 May 2010.
[12] Samal M. The 14th international symposium on radionuclides in nephrourology. Semin Nucl Med 2011;41:3-10.
[13] Tondeur M, De Palma D, Roca I, Piepsz A, Ham H. Inter-observer reproducibility in reporting on renal drainage in children with hydronephrosis: a large collaborative study. Eur J Nucl Med Mol Imaging 2008;35:644-654.
[14] Fleming JS, Cosgriff PS, Houston AS, Jarritt PH, Skrypniuk JV, Whalley DR. UK audit of relative renal function measurement using DMSA scintigraphy. Nucl Med Commun 1998;19:989-997.
[15] Tondeur M, Melis K, De Sadeleer C, Verelst J, Van Espen M-B, Ham H, Piepsz A. Inter-observer reproducibility of relative 99mTc-DMSA uptake. Nucl Med Commun 2000;21:449-453.
[16] Peters AM, Bomanji J, Costa DC, Ell PJ, Gordon I, Henderson BL, Hilson AJW. Clinical audit in nuclear medicine. Nucl Med Commun 2004;25:97-103.
[17] Jarritt PH, Perkins AC, Woods SD. Audit of nuclear medicine scientific and technical standards. Nucl Med Commun 2004;25:771-775.
[18] Mirzaei S, Maffioli L, Hilson A. Clinical audit in nuclear medicine. Eur J Nucl Med Mol Imaging 2011;38:3-4.
version 1.0, Prague and Zlin, 17 March 2012