Eur J Nucl Med (1987) 13:213 220 European Nuclear Journal of Medicine © Springer-Verlag 1987 Diagnostic accuracy of rest-exercise first pass ventriculography with a fast single crystal gamma camera in detecting coronary artery disease Study in a group of male subjects without previous myocardial infarction Raffaele Giubbini, Marco Metra, Paolo Guerra, Giovanni La Canna, Giovanni Bissoli, Giovanni Arosio, Luigi Niccoli, Odoardo Visioli, and Maurizio Bestagno Servizio di Medicina Nucleare, Servizio di Policardiografia, Divisione di Cardiologia degli Spedali Civili di Brescia e Cattedra di Cardiologia dell' Universitfi di Brescia, Italy Abstract. Rest and exercise radionuclide ventriculograms were performed in 61 non infarcted, male, patients who underwent cardiac catheterization for chest pain and in 16 normal control subjects. Studies were performed using the first pass method with a fast single crystal gamma camera, which allowed a count rate of 140_+ 19 Kcounts/sec to be reached during left ventricular filling; the count integral on left ventricular area was 10.8 _+1.6 Kcounts and the max- imum count/pixel 155-+ 16. We analyzed sensitivity, speci- ficity, positive and negative predictive value of global ejec- tion fraction (EF) and of the regional wall motion in identi- fying ventricular function abnormalities due to obstructive coronary artery disease. The regional wall motion was eval- uated with four functional images: regional ejection frac- tion (REF), amplitude (A) and phase (PH) from Fourier analysis and systolic transit times (TT). Sensitivity was near 90 % for EF, REF, A and TT, while PH was less sensitive (80%); all functional images were more specific (nearly 90 %) than EF (80 %). Both sensitivity and specificity were lower for the exercise EKG (59 % and 63 %, respectively) in this patient group. Significant differences between single vessel and multiple vessel disease were also observed either for the EF increase/decrease (-1.34+7.4 and -7.82-+9.96; P<0.05) or for the number of segments which developed wall motion abnormalities during exercise (1.22-+ 0.73 and 2.15 _+1.04; P < 0.02). In conclusion, with our method, a fast single crystal gamma camera is suitable for obtaining optimal first pass radionuclide ventriculo- grams with a count density sufficient either for global or regional left ventricular function evaluation. First pass ra- dionuclide ventriculography seems to provide very high di- agnostic accuracy in the detection of coronary artery dis- ease in non infarcted, male, subjects. Key words: First pass ventriculography - Coronary artery disease - 99mTc-DTPA - Cardiac functional images - Four- ier analysis Exercise radionuclide ventriculography (RNV) has been shown to reliably detect abnormalities of left ventricular (LV) systolic function induced by myocardial ischemia. Re- Offprint requests to." Dr. Raffaele Giubbini, Servizio di Medicina Nucleare Spedali Civili, Piazza Spedali Civili, 1, 1-25100 Brescia, Italia cently two main techniques have been developed and suc- cessfully correlated with the results of contrast ventriculog- raphy (CVG): first pass RNV (FP-RNV) (Schelbert et al. 1975; Schad 1977, 1978) and gated equilibrium RNV (Borer et al. 1977, 1979). FP-RNV offers many theoretical advantages: firstly, the acquisition time is very short (a few seconds) and so this examination can be easily performed at the end of maximal exercise while the patient is comfortably still; secondly, mo- tion of the diaphragm can be eliminated by making the patient hold his breath during the LV transit of the radioac- tive bolus; and lastly, FP-RNV allows the use of the right anterior oblique (RAO) view by which the LV can be more thoroughly visualized along its long axis (Fig. 1). These characteristics make FP-RNV the examination of choice when performed during isotonic exercise, which, as demon- strated by Bodenheimer et al. (1983), gives more accurate results in the identification of coronary artery disease (CAD) than isometric tests. Until recently, however, the use of FP-RNV has been limited by the requirement for a multicrystal gamma cam- era, which could reach the high count rate needed for the FP-RNV acquisition, but, on the other hand, is very expen- sive and almost exclusively dedicated. The recently introduced, fast single crystal gamma cam- era seems to overcome the main limitation of traditional equipment, that is its low count rate performance while, at the same time, it is less expensive and more versatile than the multicrystal camera. Thus it could allow FP-RNV to be performed more widely and in less specialized centers. The aim of this study was to verify the accuracy of FP-RNV performed with a fast single crystal gamma cam- era in the detection of the abnormalities of LV systolic function caused by exercise induced myocardial ischemia in patients with CAD. Materials and methods Data collection and injection technique. The acquisition pro- tocol for FP-RNV was standardized in order to obtain the highest count rate and the best reproducibility. We used a fast digital gamma camera system, LFOV 1, capable of measuring 200000 cps with 25% loss at 20% window. It was equipped with a high sensitivity parallel hole collimator (6.5 x 10 4 FWHM:15 ram); the energy 1 Elscint Apex 415