Hyperspectral Imaging Accurately Detects Renal Malperfusion Due to High Intrarenal Pressure

Abstract

Background and objective: High intrarenal pressure (IRP) is a significant concern in both endoscopic procedures and acute hydronephrosis, and may cause renal parenchymal damage, forniceal rupture, and long-term impaired renal function. Its pathomechanism and effect on renal perfusion patterns remain undetermined. This study investigates the impact of elevated IRP on renal perfusion and oxygen saturation (StO₂) using hyperspectral imaging (HSI).

Methods: In vivo experiments were conducted on porcine models establishing hydronephrosis on specific IRP levels (30, 50, 70, and 90 mmHg) by pressure-controlled infusion of crystalloid solution into the ureter after distal ureteral clamping. HSI data were recorded at baseline, during IRP application, and after release to measure hydronephrosis-induced changes in reflectance and perfusion in a total of 501 recordings. The results were compared with spectral patterns of renal malperfusion states from previous internal studies. In total, data of 73 pigs and 1744 HSI recordings were included.

Key findings and limitations: Elevated IRP significantly affected renal perfusion and oxygenation. StO₂ decreased from 70.3% ± 10.9% (physiological) to 39.9% ± 9.5% in hydronephrotic kidneys. Perfusion values in hydronephrosis decreased significantly at the renal poles (6.5% ± 4.0%) compared with physiological values (34.8% ± 7.5%). A principal component analysis and machine learning classification confirmed distinct malperfusion states, with hydronephrosis resembling ischemic conditions.

Conclusions and clinical implications: HSI revealed that high IRP reduces renal oxygenation and perfusion, with the poles being disproportionately affected. The results from this study provide quantitative evidence of perfusion restriction and ischemic conditions as the pathomechanism behind hydronephrosis-induced kidney damage. These findings underscore the importance of monitoring IRP during endourological procedures to mitigate renal damage and associated complications.

Patient summary: High pressure in the kidneys during surgery or kidney disease can severely reduce blood flow and oxygen, causing damage. This study used a special camera to show this damage, especially at the end of the kidney. These findings highlight the importance of monitoring kidney pressure carefully during procedures to prevent damage to the kidney.

Keywords: Animal study; Hydronephrosis; Hyperspectral imaging; Intrarenal pressure; Porcine model; Urology.

Fig. 1

Recording protocol: (A) experimental setup, (B) validation setup, and (C) overview over animal numbers of baseline groups: physiological (I = 60, n = 777), avascular (I = 18, n = 171), arterial ischemia (I = 17, n = 201), venous congestion (I = 16, n = 94), hydronephrosis (70 and 90 mmHg with time points 15, 30, 45, and 60 min; I = 12, n = 147). (D) Flow chart of the main experimental groups with measurements for hydronephrosis reference data in the baseline groups marked with a green dashed box. I = number of individuals; n = number of independent measurements; ROI = region of interest.

Fig. 2

Baseline kidney data. HSI color-index pictures and respective spectra for porcine kidney: (A) physiological (I = 54, n = 757); (B) avascular (I = 17, n = 162); (C) arterial ischemia (I = 14, n = 156); (D) venous congestion (I = 15, n = 91); (E) hydronephrosis (hydronephrosis reference data; I = 12, n = 147), (F) quantification of HSI index values for StO₂, NIR perfusion, THI, and TWI; and (G) comparison of L1-normalized reflectance spectra. White scale bar equals 5 cm. HSI = hyperspectral imaging; I = number of individuals; n = number of independent measurements; NIR = near-infrared perfusion; StO₂ = oxygen saturation; THI = tissue hemoglobin index; TWI = tissue water index.

Fig. 3

Comparison of baseline kidney data: (A) principal component analysis (PCA) of the baseline perfusion groups (see previous publication [9]), (B) PCA including the additional hydronephrosis group in green (hydronephrosis reference data) including example images, (C) previously developed machine learning model for differentiating renal malperfusion states, and (D) results from applying the model to hydronephrosis images. Measurements from the hydronephrotic kidney that were classified by the model as “physiological” are indicated with a red frame, while the ones classified as “combined compromised inflow” are indicated with a white frame. Mean and standard deviation of oxygenation values are provided as boxplots.

Fig. 4

Pole demarcation in hydronephrotic kidney: (A) HSI color-index pictures of ROI1 and ROI2 of the hydronephrotic kidney (hydronephrosis reference data); (B) quantification of HSI index values for StO₂, NIR perfusion, and TWI; (C) comparison of L1-normalized reflectance spectra; and (D) PCA of the physiological kidney and ROI1 versus ROI2. HSI = hyperspectral imaging; NIR = near-infrared perfusion; PCA = principal component analysis; ROI = region of interest; StO₂ = oxygen saturation; TWI = tissue water index.

Fig. 5

Hydronephrosis stratified by pressure. HSI color-index pictures and respective spectra for porcine kidney: (A) 30 mmHg (I = 6, n = 71), (B) 50 mmHg (I = 6, n = 74), (C) 70 mmHg (I = 6, n = 73), (D) 90 mmHg (I = 6, n = 74), (E) quantification of HSI index values for StO₂, NIR perfusion, THI, and TWI, and (F) comparison of L1-normalized reflectance spectra. White scale bar equals 5 cm. HSI = hyperspectral imaging; I = number of individuals; n = number of independent measurements; NIR = near-infrared perfusion; PCA = principal component analysis; StO₂ = oxygen saturation; THI = tissue hemoglobin index; TWI = tissue water index.