[Evaluation of postoperative hepatobiliary function by deconvolution analysis of hepatobiliary image data by...

Dynamic hepatobiliary image data were analyzed by the mathematical deconvolution to derive the transfer function (TF) which represents the impulse response function of the liver following direct bolus injection of a tracer into the liver. Biliary flow was evaluated by TF in patients with previous hepatobiliary surgery to detect abnormal flow causing problems such as cholangitis. Serial image data were collected for 60 min after intravenous injection of 37-75 MBq (1-2 mCi) of 99mTc-N-pyridoxyl-5-methyltryptophan (99mTc-PMT). TF was obtained by the matrix algorithm, using regional hepatogram as output and time-activity curve over the heart as input function. Minimum, mean and maximum transit time (Max. TT) were determined from TF. The functional image for each of the transit times was constructed by displaying the distribution of the values for matrix elements with a color scale. In this study, a critical Max. TT of 60 min was used to detect abnormal biliary flow. Of 30 cases, 9 positive cases were discriminated from 21 negative cases on the basis of the finding of presence or absence of areas with Max. TT of more than 60 min anywhere in the liver. Bowel activity overlapping the liver was excluded in the judgment. The validity of this method was evaluated in comparison with the clinical courses after surgery with (8 cases) or without (22 cases) problems. The results were sensitivity of 88%, specificity of 91%, accuracy of 90% and positive predictive value of 78%.(ABSTRACT TRUNCATED AT 250 WORDS)