Acquisition time reduction in pediatric 99m Tc-DMSA planar imaging using deep learning.

PURPOSE: Given the potential risk of motion artifacts, acquisition time reduction is desirable in pediatric 99m Tc-dimercaptosuccinic acid (DMSA) scintigraphy. The aim of this study was to evaluate the performance of predicted full-acquisition-time images from short-acquisition-time pediatric 99m Tc-DMSA planar images with only 1/5th acquisition time using deep learning in terms of image quality and quantitative renal uptake measurement accuracy. METHODS: One hundred and fifty-five cases that underwent pediatric 99m Tc-DMSA planar imaging as dynamic data for 10 min were retrospectively collected for the development of three deep learning models (DnCNN, Win5RB, and ResUnet), and the generation of full-time images from short-time images. We used the normalized mean squared error (NMSE), peak signal-to-noise ratio (PSNR), and structural similarity index metrics (SSIM) to evaluate the accuracy of the predicted full-time images. In addition, the renal uptake of 99m Tc-DMSA was calculated, and the difference in renal uptake from the reference full-time images was assessed using scatter plots with Pearson correlation and Bland-Altman plots. RESULTS: The predicted full-time images from the deep learning models showed a significant improvement in image quality compared to the short-time images with respect to the reference full-time images. In particular, the predicted full-time images obtained by ResUnet showed the lowest NMSE (0.4 [0.4-0.5] %) and the highest PSNR (55.4 [54.7-56.1] dB) and SSIM (0.997 [0.995-0.997]). For renal uptake, an extremely high correlation was achieved in all short-time and three predicted full-time images (R2  > 0.999 for all). The Bland-Altman plots showed the lowest bias (-0.10) of renal uptake in ResUnet, while short-time images showed the lowest variance (95% confidence interval: -0.14, 0.45) of renal uptake. CONCLUSIONS: Our proposed method is capable of producing images that are comparable to the original full-acquisition-time images, allowing for a reduction of acquisition time/injected dose in pediatric 99m Tc-DMSA planar imaging.

Vesicoureteral Reflux Detected with (99m)Tc-DTPA Renal Scintigraphy during Evaluation of Renal Function.

BACKGROUND: Radionuclide techniques, as direct radionuclide cystography and (99m)Tc-DMSA scintigraphy, have been used in evaluation of vesicoureteral reflux (VUR) and reflux nephropathy (RN) in children. Dynamic (99m)Tc-DTPA scintigraphy is reserved for evaluation of differential renal function and obstruction in children, where hydronephrosis is detected by ultrasonography (US) pre- or postnatally. CASE REPORT: Six year old boy was prenatally diagnosed with bilateral hydronephrosis. Postnatal, severe bilateral VUR was detected by voiding urethrocytography. US and (99m)Tc-DTPA scintigraphy performed in the first month of life showed small left kidney that participated with 2% in the global renal function. Bilateral cutaneous ureterostomy has been performed in order to obtain good renal drainage and promote optimal renal growth. Twelve months later, classic antireflux procedure was done. Control (99m)Tc-DTPA scintigraphy, 5 ys after antireflux surgery, revealed persisting radioactivity during the diuretic phase, in the left kidney that indicated antireflux procedure failure with VUR reappearance. CONCLUSION: (99m)Tc-DTPA scintigraphy is the first method of choice for long-term monitoring of individual kidney function in children with VUR and other congenital urinary tract anomalies. Additionally, it can be used as indirect radionuclide cystography when rising of radioactivity in the kidney region, during the diuretic phase can indicate presence of VUR.

Comparison of planar DMSA scan with an evaluation based on SPECT imaging in the split renal function assessment.

BACKGROUND: Validation of the 99mTc-DMSA planar scintigraphy accuracy for split renal function assessment and comparison with evaluation based on SPECT imaging both with and without CT attenuation correction. MATERIALS AND METHODS: For split renal function assessment two methods were used: A) planar scintigraphy based on anterior and posterior projections using correction for kidney depth calculated by the geometric mean; B) semi-quantitative evaluation based on SPECT (B1) and attenuation-corrected SPECT/CT (B2) images using locally developed software for kidney segmentation and voxel-based analysis. All three methods were performed with a phantom simulating body including pair of kidneys. For patient study methods A and B1 were applied on a group of 140 children and adolescents with various renal diseases. Renal function ratios were compared both mutually and with physically measured activity ratios in the phantom. RESULTS: Method A provided results which were closest to measured reference values (average absolute difference of 0.9 percentage points [pp]). Method B1 was noticeably worse (2.1pp), whereas attenuation correction (B2) improved tomography results considerably (1.3 pp). The superiority of planar imaging could be caused among others by differences in creation of planar range of interest compared to tomographic volume of interest. However all the differences were under the threshold of any clinical importance. The comparison between method A and B1 based on patient study also showed differences mostly of none clinical importance. CONCLUSION: Routine evaluation of split renal function using planar technique with correction of the kidney depth is at least equivalent to tomographic evaluation, and there is no need to update the established clinical practice.

Función renal relativa normalizada por volumen en la gamagrafía con 99mTc-DMSA: estudio basado en una población pediátrica / Volume normalized relative renal function in 99mTc-DMSA scintigraphy: study in a paediatric population

Objetivo Evaluar las diferentes características diagnósticas de la función renal relativa (FRR) y de la función renal relativa normalizada por volumen renal relativo (FRRN), con el fin de analizar la posible mejora diagnóstica que implicaría su uso conjunto. Métodos Hemos estudiado con 99mTc-DMSA 952 unidades renales (riñones individuales), que pertenecían a 476 pacientes pediátricos (269 niñas) de edades comprendidas entre 0 y 11 años. De ellas, 355 unidades renales se consideraron patológicas a partir del análisis visual gammagráfico. Se determinaron los puntos de corte más apropiados para establecer los valores normales de las funciones FRR y FRRN, considerando como patrón de referencia al análisis visual gammagráfico. Se realizaron curvas ROC y se calculó el índice “J de Youden” y la exactitud para diferentes puntos de corte desde 42 por ciento hasta 50 por ciento, tanto para la FRR como para la FRRN. Se realizó también un análisis bayesiano (global y parcial clasificando las unidades renales por indicaciones clínicas) de la FRR y la FRRN. Resultados Para la FRR el mejor punto de corte fue 45 por ciento y para la FRRN fue 47 por ciento. El análisis bayesiano global mostró mejores resultados de los diferentes parámetros estadísticos para la FRRN. El análisis bayesiano parcial indicaba un importante incremento de la sensibilidad en pacientes con patología de dilatación de vía: de 10 por ciento (FRR) a 74.6 por ciento (FRRN); y con pielonefritis aguda: de 29.2 por ciento (FRR) a 66.2 por ciento (FRRN) mientras que en pacientes con nefropatías crónicas la sensibilidad fue similar (70.1 por ciento). Conclusión El cálculo de la FRRN (punto de corte 47 por ciento) proporciona información útil y complementaria a la FRR, ya que cuantifica la calidad relativa del riñón independientemente del volumen renal y refleja mejor los hallazgos gammagráficos.

Purpose To assess the different diagnostic features of relative renal function (RRF) and volume normalized relative renal function (NRRF) with the aim of analysing the possible diagnostic improvement that would imply their joint use. Methods We studied 952 kidneys with 99mTc-DMSA, belonging to 476 paediatric patients (269 girls) aged 0-11 years. 355 out of total were considered pathologic (visual analysis). The most appropriate cut-off points for establishing normality of RRF and NRRF were determined, considering as gold standard the visual scintigraphic analysis. R.O.C. curves were performed, J Youden index and accuracy were calculated for the different cut-off points from 42 percent to 50 percent for RRF and NRRF. A bayesian analysis (global and partial according to clinical indications) of RRF and NRRF was also performed. Results For RRF the better cut-off point was 45 percent and for NRRF 47 percent. The global bayesian analysis showed better values of the different statistical parameters for NRRF. Partial bayesian analysis indicated an important increment of sensitivity in patients with dilated pyelocalicial system: from 10 percent (RRF) to 74.6 percent (NRRF); and acute pyelonephritis: from 29.2 percent (RRF) to 66.2 percent (NRRF) while the sensitivity in patients with chronic nephropathy remained stable (70.1 percent). Conclusion NRRF calculation (cut-off 47 percent) provides useful and additional information to RRF, as it quantifies the relative quality of kidney tissue regardless of renal volume and better reflects better the scintigraphic findings.