Abstract
Objectives
To develop and evaluate an artificial intelligence (AI) system that can automatically calculate the glomerular filtration rate (GFR) from dynamic renal imaging without manually delineating the regions of interest (ROIs) of kidneys and the corresponding background.
Methods
This study was a single-center retrospective analysis of the data of 14,634 patients who underwent 99mTc-DTPA dynamic renal imaging. Two systems based on convolutional neural networks (CNN) were developed and evaluated: sGFRa predicts the radioactive counts of ROIs and calculates GFR using the Gates equation and sGFRb directly predicts GFR from dynamic renal imaging without using other information. The root-mean-square error (RMSE), mean absolute error (MAE), mean absolute percentage error (MAPE), and R2 were used to evaluate the performance of our approach.
Results
sGFRa achieved an RMSE of 5.05, MAE of 4.03, MAPE of 6.07%, and R2 of 0.93 for total GFR while sGFRb achieved an RMSE of 7.61, MAE of 5.92, MAPE of 8.92%, and R2 of 0.85 for total GFR. The accuracy of sGFRa and sGFRb in determining the stage of chronic kidney disease was 87.41% and 82.44%, respectively.
Conclusions
The findings of sGFRa show that automatic GFR calculation based on CNN and using dynamic renal imaging is feasible and efficient and, additionally, can aid clinical diagnosis. Furthermore, the promising results of sGFRb demonstrate that CNN can predict GFR from dynamic renal imaging without additional information.
Key Points
• Our CNN-based AI systems can automatically calculate GFR from dynamic renal imaging without manually delineating the ROIs of kidneys and the corresponding background.
• sGFR a accurately predicted the radioactive counts of ROIs and calculated GFR using the Gates method.
• sGFR b -predicted GFR directly without any parameters related to the Gates equation.
Similar content being viewed by others
Abbreviations
- AI :
-
Artificial intelligence
- CKD :
-
Chronic kidney disease
- CNN :
-
Convolutional neural network
- DICOM :
-
Digital imaging and communications in medicine
- GFR :
-
Glomerular filtration rate
- MAE :
-
Mean absolute error
- MAPE :
-
Mean absolute percentage error
- RMSE :
-
Root-mean-square error
- ROIs :
-
Regions of interest
References
Finco DR (1997) Kidney function. In: Clinical biochemistry of domestic animals. Elsevier, pp 441–484. https://doi.org/10.1016/B978-012396305-5/50018-X
Porrini E, Ruggenenti P, Luis-Lima S et al (2019) Estimated GFR: time for a critical appraisal. Nat Rev Nephrol. 15(3):177–190. https://doi.org/10.1038/s4158
Gaspari F, Perico N, Remuzzi G (1997) Measurement of glomerular filtration rate. Kidney Int Suppl. 63:S151–S154
Levey AS, Coresh J, Tighiouart H, Greene T, Inker LA (2019) Strengths and limitations of estimated and measured GFR. Nat Rev Nephrol. 15(12):784–784. https://doi.org/10.1038/s41581-019-0213-9
Price M (1972) Comparison of creatinine clearance to inulin clearance in the determination of glomerular filtration rate. J Urol. 107(3):339–340. https://doi.org/10.1016/S0022-5347(17)61021-8
GATES GF (1983) Split renal function testing using Tc-99m DTPA. Clin Nucl Med. 8(9):400–407. https://doi.org/10.1097/00003072-198309000-00003
Zubal IG, Caride VJ (1992) The technetium-99m-DTPA renal uptake-plasma volume product: a quantitative estimation of glomerular filtration rate. J Nucl Med. 33(9):1712–1716
Blaufox MD, De Palma D, Taylor A et al (2018) The SNMMI and EANM practice guideline for renal scintigraphy in adults. Eur J Nucl Med Mol Imaging. 45(12):2218–2228. https://doi.org/10.1007/s00259-018-4129-6
Ziessman HA, O'Malley JP, Thrall JH (2013) Nuclear medicine: the requisites e-book. Elsevier Health Sciences
Tonnesen KH Influence on the radiorenogram of variation in skin to kidney distance and the clinical importance hereof. Radionucl Nephrol. Published online 1975:79-86. http://ci.nii.ac.jp/naid/10008538216/en/. Accessed 25 Sept 2021
He K, Zhang X, Ren S, Sun J (2016) Identity mappings in deep residual networks. In: Lecture notes in computer science (Including Subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). 9908 LNCS :630-645. https://doi.org/10.1007/978-3-319-46493-0_38
Kingma DP, Ba J Adam: A method for stochastic optimization. Published online December 22, 2014:103. http://arxiv.org/abs/1412.6980
Paszke A, Chanan G, Lin Z et al (2017) Automatic differentiation in PyTorch. Adv Neural Inf Process Syst 30(Nips):1–4
Virtanen P, Gommers R, Oliphant TE et al (2020) SciPy 1.0: fundamental algorithms for scientific computing in Python. Nat Methods. 17(3):261–272. https://doi.org/10.1038/s41592-019-0686-2
Garcia EV, Folks R, Pak S, Taylor A (2010) Totally automatic definition of renal regions of interest from 99mTc-MAG3 renograms: validation in patients with normal kidneys and in patients with suspected renal obstruction. Nucl Med Commun. 31(5):366–374
Lin K-J, Huang J-Y, Chen Y-S (2011) Fully automatic region of interest selection in glomerular filtration rate estimation from 99mTc-DTPA renogram. J Digit Imaging. 24(6):1010–1023
Tian C, Zheng X, Han Y, Sun X, Chen K, Huang Q (2013) A semi-automated region of interest detection method in the scintigraphic glomerular filtration rate determination for patients with abnormal low renal function. Clin Nucl Med. 38(11):855–862
dos Santos WHS, Ataky STM, Silva AC, de Paiva AC, Gattass M (2017) Automatic method for quantitative automatic evaluation in dynamic renal scintilography images. Multimed Tools Appl. 76(18):19291–19315
Zheng X, Wei W, Huang Q, Song S, Huang G (2019) Automated region of interest detection method in scintigraphic glomerular filtration rate estimation. IEEE J Biomed Heal Informatics. 23(2):787–794. https://doi.org/10.1109/JBHI.2018.2845879
Li M, Zheng X, Liu K Application of distance regularized chan-vese method for kidney segmentation in renal scintigraphy. ACM Int Conf Proceeding Ser. Published online 2019:192-196
Ma Y-C et al (2007) Comparison of 99mTc-DTPA renal dynamic imaging with modified MDRD equation for glomerular filtration rate estimation in Chinese patients in different stages of chronic kidney disease. Nephrol Dial Transplant 22(2):417–423
Funding
This study has received funding by the National Major Science and Technology Projects of China under Grant 2018AAA0100201, Sichuan Science and Technology Program under Grant 2020JDRC0042, and 1.3.5 project for disciplines of excellence, West China Hospital, Sichuan University (ZYGD18016&2021HXFH033).
Author information
Authors and Affiliations
Corresponding authors
Ethics declarations
Guarantor
The scientific guarantor of this publication is Huawei Cai.
Conflict of interest
The authors of this manuscript declare no relationships with any companies, whose products or services may be related to the subject matter of the article.
Statistics and biometry
No complex statistical methods were necessary for this paper.
Informed consent
Written informed consent was waived by the Institutional Review Board.
Ethical approval
Institutional Review Board approval was obtained.
Methodology
• retrospective
• diagnostic or prognostic study
• performed at one institution
Additional information
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
ESM 1
(PDF 1453 kb)
Rights and permissions
About this article
Cite this article
Pi, Y., Zhao, Z., Yang, P. et al. Deep regression using 99mTc-DTPA dynamic renal imaging for automatic calculation of the glomerular filtration rate. Eur Radiol 33, 34–42 (2023). https://doi.org/10.1007/s00330-022-08970-6
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00330-022-08970-6