Premium
TU‐A‐12A‐09: Absolute Blood Flow Measurement in a Cardiac Phantom Using Low Dose CT
Author(s) -
Ziemer B,
Hubbard L,
Lipinski J,
Molloi S
Publication year - 2014
Publication title -
medical physics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.473
H-Index - 180
eISSN - 2473-4209
pISSN - 0094-2405
DOI - 10.1118/1.4889255
Subject(s) - imaging phantom , perfusion , nuclear medicine , hounsfield scale , blood flow , volume (thermodynamics) , biomedical engineering , scanner , pulsatile flow , mathematics , medicine , physics , radiology , computed tomography , optics , quantum mechanics , cardiology
Purpose: To investigate a first pass analysis technique to measure absolute flow from low dose CT images in a cardiac phantom. This technique can be combined with a myocardial mass assignment to yield absolute perfusion using only two volume scans and reduce the radiation dose to the patient. Methods: A four‐chamber cardiac phantom and perfusion chamber were constructed from poly‐acrylic and connected with tubing to approximate anatomical features. The system was connected to a pulsatile pump, input/output reservoirs and power contrast injector. Flow was varied in the range of 1‐2.67 mL/s with the pump operating at 60 beats/min. The system was imaged once a second for 14 seconds with a 320‐row scanner (Toshiba Medical Systems) using a contrast‐enhanced, prospective‐gated cardiac perfusion protocol. Flow was calculated by the following steps: subsequent images of the perfusion volume were subtracted to find the contrast entering the volume; this was normalized by an upstream, known volume region to convert Hounsfield (HU) values to concentration; this was divided by the subtracted images time difference. The technique requires a relatively stable input contrast concentration and no contrast can leave the perfusion volume before the flow measurement is completed. Results: The flow calculated from the images showed an excellent correlation with the known rates. The data was fit to a linear function with slope 1.03, intercept 0.02 and an R 2 value of 0.99. The average root mean square (RMS) error was 0.15 mL/s and the average standard deviation was 0.14 mL/s. The flow rate was stable within 7.7% across the full scan and served to validate model assumptions. Conclusion: Accurate, absolute flow rates were measured from CT images using a conservation of mass model. Measurements can be made using two volume scans which can substantially reduce the radiation dose compared with current dynamic perfusion techniques.