Kailasnath Purushothaman

Volumetric flow estimation in vivo and in vitro using pulsed-Doppler ultrasound.

The measurement of volumetric blood flow in small vessels in vitro and in vivo poses a significant technological challenge. In this study, two pulsatile flow models were developed, one with a 3.2-mm lumen diameter and one with a 12.7-mm lumen diameter, to assess the accuracy of volumetric flow estimation of two pulsed-Doppler devices, a Crystal Biotech VF1 20-MHz system with either a cuff-mounted or a needle-mounted probe and an Advanced Technology Laboratories Ultramark 9 High Definition Imaging system with a 5-MHz linear array transducer. The VF1 volumetric flow error was measured in the 3.2-mm phantom over a variety of pulsatile and continuous waveforms. The accuracy of the VF1 was also tested in porcine femoral and renal arteries. VF1 volumetric flow error ranged from 4.8% to 54.3% in the in vivo studies. The ATL demonstrated similar volumetric flow errors in the porcine femoral artery (approximately 3.2 mm diameter), but these errors were reduced to < or = 17.4% in the 12.7-mm-diameter in vitro flow model.

Effects of adenosine and a selective A2A adenosine receptor agonist on hemodynamic and thallium-201 and technetium-99m-sestaMIBI biodistribution and kinetics.

OBJECTIVES The purpose of this study was to compare a selective A(2A) adenosine receptor agonist (regadenoson) with adenosine in clinically relevant canine models with regard to effects on hemodynamics and thallium-201 ((201)Tl) and technetium-99m ((99m)Tc)-sestaMIBI biodistribution and kinetics. BACKGROUND The clinical application of vasodilator stress for perfusion imaging requires consideration of the effects of these vasodilating agents on systemic hemodynamics, coronary flow, and radiotracer uptake and clearance kinetics. METHODS Sequential imaging and arterial blood sampling was performed on control, anesthetized closed-chest canines (n = 7) to evaluate radiotracer biodistribution and kinetics after either a bolus administration of regadenoson (2.5 microg/kg) or 4.5-min infusion of adenosine (280 microg/kg). The effects of regadenoson on coronary flow and myocardial radiotracer uptake were then evaluated in an open-chest canine model of a critical stenosis (n = 7). Results from ex vivo single-photon emission computed tomography were compared with tissue well-counting. RESULTS The use of regadenoson compared favorably with adenosine in regard to the duration and magnitude of the hemodynamic effects and the effect on (201)Tl and (99m)Tc-sestaMIBI biodistribution and kinetics. The arterial blood clearance half-time was significantly faster for (99m)Tc-sestaMIBI (regadenoson: 1.4 +/- 0.03 min; adenosine: 1.5 +/- 0.08 min) than for (201)Tl (regadenoson: 2.5 +/- 0.16 min, p < 0.01; adenosine: 2.7 +/- 0.04 min, p < 0.01) for both vasodilator stressors. The relative microsphere flow deficit (0.34 +/- 0.02%) during regadenoson stress was significantly greater than the relative perfusion defect with (99m)Tc-sestaMIBI (0.69 +/- 0.03%, p < 0.001) or (201)Tl (0.53 +/- 0.02%, p < 0.001), although (201)Tl tracked the flow deficit within the ischemic region better than (99m)Tc-sestaMIBI. The perfusion defect score was larger with (201)Tl (22 +/- 2.8% left ventricular) than with (99m)Tc-sestaMIBI (17 +/- 1.7% left ventricular, p < 0.05) on ex vivo single-photon emission computed tomography images. CONCLUSIONS The bolus administration of regadenoson produced a hyperemic response comparable to a standard infusion of adenosine. The biodistribution and clearance of both (201)Tl and (99m)Tc-sestaMIBI during regadenoson were similar to adenosine vasodilation. Ex vivo perfusion images under the most ideal conditions permitted detection of a critical stenosis, although (201)Tl offered significant advantages over (99m)Tc-sestaMIBI for perfusion imaging during regadenoson vasodilator stress.

Motion analysis of 3D ultrasound texture patterns

We model the process of imaging soft tissues with a 3D ultrasound probe using a linear convolution model, and obtain analytical expressions of both the ultrasound image and its spectrum. We use this model to study the ultrasound decorrelation caused by tissue motion both in the spatial domain and spectral domain. Finally, we propose a spectral-feature-based algorithm to analyze tissue motion. The comparison with intensity-based algorithm shows promising results.

Support Vector Machine Density Estimator as a Generalized Parzen Windows Estimator for Mutual Information based Image Registration.

Mutual Information is perhaps the most widely used multimodality image registration method. A crucial step in mutual information is the estimation of the probability density function (pdf). In most cases, the Parzen window estimator is employed for this purpose which results in an excessive computational cost. In this paper we demonstrate that replacing the Parzen density estimator with a Support Vector Machine (SVM) density estimation will result in a significant reduction of the computational time. We verified this by registering 2D portal images to DRRs (digitally reconstructed radiographs) projected from 3D CT volumetric data.

Influence of position errors on tumor control in external beam therapy of prostate cancer: concepts of TCP maps and probability densities

Purpose/Objective: The overall objective of this study was to test the hypothesis that position uncertainty significantly affects tumor control probability (TCP) in some of the current theoretical models of tumor biology. The purpose of this work was to systematically explore the relevance of position errors to TCP using a TCP model, which accounts for the variability of radiosensitivity in the clonogen population.

Corrigendum to “Constrained non-rigid registration for use in image-guided adaptive radiotherapy” [Medical Image Analysis 13 (2009) 809–817]

Corrigendum Corrigendum to ‘‘Constrained non-rigid registration for use in image-guided adaptive radiotherapy” [Medical Image Analysis 13 (2009) 809–817] W.H. Greene *, S. Chelikani , K. Purushothaman , J.P.S. Knisely , Z. Chen , X. Papademetris , L.H. Staib , J.S. Duncan a,b Department of Biomedical Engineering, Yale University, New Haven, CT 06520, USA Department of Diagnostic Radiology, Yale University, New Haven, CT 06520, USA Department of Therapeutic Radiology, Yale University, New Haven, CT 06520, USA