Validation of Tikhonov adaptively regularized gamma variate fitting with 24-h plasma clearance in cirrhotic patients with ascites

Purpose

The aim was to compare late-time extrapolation of plasma clearance (CL) from Tikhonov adaptively regularized gamma variate fitting (Tk-GV) and from mono-exponential (E1) fitting.

Methods

Ten ⁵¹Cr-ethylenediaminetetraacetic acid bolus IV studies in adults—8 with ascites—assessed for liver transplantation, with 12–16 plasma samples drawn from 5-min to 24-h, were fit with Tk-GV and E1 models and CL results were compared using Passing-Bablok fitting.

Results

The 24-h CL(Tk-GV) values ranged from 11.4 to 79.7 ml/min. Linear regression of 4- versus 24-h CL(Tk-GV) yielded no significant departure from a slope of 1, whereas the 4- versus 24-h CL(E1) slope, 1.56, was significantly increased. For CL(Tk-GV-24-h) versus CL(E1-24-h), there was a biased slope and intercept (0.85, 5.97 ml/min). Moreover, the quality of fitting of 24-h data was significantly better for Tk-GV than for E1, as follows. For 10 logarithm of concentration curves, higher r values were obtained for each Tk-GV fit (median 0.998) than for its corresponding E1 fit (median 0.965), with p < 0.0001 (paired t-test of z-statistics from Fisher r-z transformations). The E1 fit quality degraded with increasing V/W [volume of distribution (l) per kg body weight, p = 0.003]. However, Tk-GV fit quality versus V/W was uncorrelated (p = 0.8).

Conclusion

CL(E1) values were dependent on sample time and the quality of fit was poor and degraded with increasing ascites, consistent with current opinion that CL(E1) is contraindicated in ascitic patients. CL(Tk-GV) was relatively more accurate and the good quality of fit was unaffected by ascites. CL(Tk-GV) was the preferred method for the accurate calculation of CL and was useful despite liver failure and ascites.

Tikhonov adaptively regularized gamma variate fitting to assess plasma clearance of inert renal markers

The Tk-GV model fits Gamma Variates (GV) to data by Tikhonov regularization (Tk) with shrinkage constant, λ, chosen to minimize the relative error in plasma clearance, CL (ml/min). Using (169)Yb-DTPA and (99m)Tc-DTPA (n = 46, 8-9 samples, 5-240 min) bolus-dilution curves, results were obtained for fit methods: (1) Ordinary Least Squares (OLS) one and two exponential term (E₁ and E₂), (2) OLS-GV and (3) Tk-GV. Four tests examined the fit results for: (1) physicality of ranges of model parameters, (2) effects on parameter values when different data subsets are fit, (3) characterization of residuals, and (4) extrapolative error and agreement with published correction factors. Test 1 showed physical Tk-GV results, where OLS-GV fits sometimes-produced nonphysical CL. Test 2 showed the Tk-GV model produced good results with 4 or more samples drawn between 10 and 240 min. Test 3 showed that E₁ and E₂ failed goodness-of-fit testing whereas GV fits for t > 20 min were acceptably good. Test 4 showed CL(Tk-GV) clearance values agreed with published CL corrections with the general result that CL(E1) > CL(E2) > CL(Tk-GV) and finally that CL(Tk-GV) were considerably more robust, precise and accurate than CL(E2), and should replace the use of CL(E2) for these renal markers.

An improved method for determining renal sufficiency using volume of distribution and weight from bolus 99mTc-DTPA, two blood sample, paediatric data

OBJECTIVES

To find an improved method of determining renal sufficiency by exploring power functions for estimating normal value, E(arg), single compartment glomerular filtration rate (G1), rate constant (gamma) and renal sufficiency index, RSI = gamma/E(gamma) = G1/E(G1), using compartment volume (V), patient mass ( W), patient age (A), patient height (H), and sex (S). To present the best estimator of normal, E(G1) = f(V, W).

METHODS

One hundred and thirty 99mTc-diethylenetriaminepentaacetic acid (99mTc-DTPA) combined imaging and G1 studies in 97 children were screened by findings and history to obtain 44 normal studies of patients 1.46-18.5 years of age with blood samples at 109 (90-214) and 152 (120-246) min. Normal studies were used to generate predictive formulae.

RESULTS

For power functions, [equation: see text] the statistically acceptable formulae in descending order of adjusted R2 were f(V, W), f(V,A), f(V,H), f(V), f(W,A,H), f( W), f(H) and f(A). Relationships of the body surface area type, f(W, H, S) and f(W, H), were statistically unwarranted. Kleiber's law, E(G1) proportional to W(3/4), with E(G1) = 6.9190W(0.7618) found here, allowed confirmation of GFRinulin approximately 0.87G1. The best estimator is f(V, W) = 10.998V(0.64717 W(0.20185), and may relate to a volumetric measure of body habitus. To verify methods, Monte Carlo simulation of the glomerular filtration rate (GFR) and f(V, W) was performed and yielded less than 5% precision error, 98% of the time. Normal RSI from f(V, W) had the smallest standard deviation, 11.3%, no regression bias over a six-fold range on a Bland-Altman ratio plot, p = 0.4, and good agreement with clinical classification at 95% specificity (RSI > 0.8589, Cohen's Kappa 0.70+/-0.062 (mean+/-bootstrap standard error).

CONCLUSIONS

The best RSI from f(V, W) is RSI = 90.927gammaV(0.35283)W(-0.20185) and should detect mildly (14.1%) reduced renal sufficiency.

Spectral restoration and energy resolution improvement of electron energy-loss spectra by Pixon reconstruction: I. Principle and test examples

We explore the use of the Pixon method, one of the highest performance image reconstruction methods available today, for the analysis of electron energy loss spectra (EELS). The method takes particular advantage of two-dimensional (2-D) CCD data, in which the data are blurred by a PRF (point-response-function as measured by the low-loss spectrum) both in the direction of energy dispersion and in the cross-dispersion direction. The Pixon method is used to simultaneously (i) deconvolve the PRF, increasing the spectral resolution, and (ii) fit the data in the cross-dispersion direction which efficiently combines the signal into a single parent spectrum. Relative to 1-D methods, we demonstrate that our 2-D treatment (i) more effectively detects weak features and (ii) allows EELS methods to be applied to much lower signal-to-noise ratio data. Both of these advantages are critical to the future development of EELS analysis.

Spectral restoration and energy resolution improvement of electron energy-loss spectra by Pixon reconstruction: II. Application to practical ELNES analysis of low SNR

We applied Pixon deconvolution as introduced in Part I to several practical, examples of low signal-to-noise ratio (SNR), electron energy-loss spectra with a goal toward restoring their fine spectral features and/or improving the energy resolution. We demonstrate that by directly fitting the two-dimensional spectral data recorded on the CCD; the method enables us to reveal fine spectral structures. Consequently, Pixon reconstruction extends the ability to probe electronic states in very spatially localized areas, a capability currently unique to our method.

Improved lesion detection from spatially adaptive, minimally complex, Pixon® reconstruction of planar scintigraphic images

Pixon noise reduction was applied to 18 planar images, six each from 99mTc-methylene diphosphonate (99mTc-MDP), 67Ga citrate (67Ga), and (123)I-metaiodobenzylguanidine ((123)I-MIBG) studies. Pixon processing increased patient signal-to-noise ratio, 6.8-11.8 fold. Three specialists preferred processed images 44 of 54 times with good agreement (87%). Most (9/10, p<0.02) of the null and negative preferences were from (123)I-MIBG studies. Inter-rater association was shown for 1-4 scale rated artifact p<0.1, noise p<0.01 and lesion detection p<0.05. Pixon images had superior lesion detection ability, p<0.02, and noise levels, p<0.02 and no statistically significant change in artifacts.