CT derived Patlak images of the human kidney

This study aimed to produce Patlak images of the kidney from dynamic CT data and to determine whether such images are substantially affected by fluid movement between renal tubular segments. Renal permeability was measured in 31 kidneys by applying Patlak analysis to time-density data from kidney and aorta during dynamic CT. Permeability parameters were correlated against plasma urea. The renal region (cortex or medulla) with the greatest permeability was determined from parametric images generated using pixel by pixel analysis. The mean value for whole kidney permeability was 517.5 mu l min(-1) ml(-1). A correlation was found between whole kidney permeability and plasma urea (p<0.01). Permeability values were highest in the renal medulla in 24 (77%) kidneys. The higher medullary values of permeability are artefactual, resulting from movement of fluid and contrast medium between cortex and medulla. Although Patlak images do not reflect, true intrarenal permeability values, the apparent medullary permeability may provide diagnostically useful information about the concentrating ability of the kidney. CT measurements of whole kidney permeability reflect filtration function but the apparent intrarenal variations in permeability will result in measurement errors dependent upon the relative amounts of renal cortex and medulla included in the CT slice studied.