Validation of automated spectrofluorimetry for measurement of regional organ perfusion using fluorescent microspheres

Citation
C. Schimmel et al., Validation of automated spectrofluorimetry for measurement of regional organ perfusion using fluorescent microspheres, COMPUT M PR, 62(2), 2000, pp. 115-125
Citations number
15
Categorie Soggetti
Multidisciplinary
Journal title
COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE
ISSN journal
01692607 → ACNP
Volume
62
Issue
2
Year of publication
2000
Pages
115 - 125
Database
ISI
SICI code
0169-2607(200006)62:2<115:VOASFM>2.0.ZU;2-E
Abstract
The fluorescent microsphere (FM) method for determination of regional organ blood Row is labor intensive, requiring processing of 100 s to 1000 s of s amples per organ. To save time and reduce handling errors, we developed an automated fluorescence analysis system by interfacing a commercially availa ble spectrofluorimeter and sample delivery unit to a PC, operated by a Wind ows95(R)-based program, WINFAC. WINFAC allows versatile analysis setup and instrument control for fluorescent intensity acquisition at fixed wavelengt hs or by synchronous scanning. Data is presented on screen, in real-time, a nd stored in text format. Reference analyses are performed at specified int ervals and the coefficient of variation is continuously updated to monitor instrument performance. The automated system was validated against radioact ive microspheres (RM) for lung perfusion in a pig and sheep and evaluated f or analysis reproducibility. Fluorescent and radioactive microspheres were delivered by simultaneous injection into a femoral vein. Lungs were excised , flushed, dried at total lung capacity, and cubed into approximate to 2cm( 3) pieces (n = 833 and 1560, pig and sheep, respectively). Radioactive coun t rates were determined for each lung piece (corrected for decay, backgroun d and spillover). Fluorescent dyes were extracted in solvent and intensitie s were determined at fixed wavelengths, using the automated spectrofluorime rer (corrected for background and spillover). Multi-color reference solutio ns were measured every 50 samples to monitor instrument reproducibility. Bl ood flow estimates for each piece determined by RM and FM methods were high ly correlated: R-2 = 0.98 +/- 0.017, Slopes = 1.00 +/- 0.007 and Intercepts = 0.00 +/- 0.006 (mean +/- SD). The CV of repeat reference analyses was 0. 71% +/- 0.16, a 30% to 50% reduction relative to manual analysis. Automated spectrofluorimetry reduces measurement errors and is a reliable and time s aving advancement. With this technology, use of FM to measure regional lung perfusion approaches the ease and accuracy of the RM method. (C) 2000 Else vier Science Ireland Ltd. All rights reserved.