Compton scatter correction in dual isotope (/sup 81/Kr/sup m/ and /sup 99/Tc/sup m/) lung SPECT

Ventilation/perfusion lung scintigraphy using inhalation of ⁸¹Kr^m and ⁹⁹Tc^m-labeled macro-aggregated albumin (MAA) respectively, is commonly used for the diagnosis of lung embolism. A simultaneous acquisition protocol poses the additional problem of cross-talk of ⁹⁹Tc^m by the scattered photons of ⁸¹Kr^m. Through a study of this contamination we propose and validate a correction methodology. Three tomographic acquisitions were carried out successively on 10 healthy volunteers, namely ventilation with ⁸¹Kr^m (1), perfusion with ⁹⁹Tc^m-MAA (2), followed by a simultaneous acquisition of ⁸¹Kr^m/⁹⁹Tc ^m-MAA (3). Data was acquired with 128 views over 360deg using a dual headed camera (E.CAM Siemens) equipped with medium energy collimators. Six energy windows were used in each of the acquisitions; W _iTc=119keVplusmn3.5%, W_pTc=140keVplusmn10%, W _sTc=158keVplusmn4%, W_iKr=166keVplusmn4%, W_pKr=190keVplusmn10%, W_sKr=216keVplusmn3%. Considering the projections of the ventilation only scan, the contamination of the three ⁹⁹Tc^m windows (W_iTc, W_pTc, W_sTc) by the scattered photons of ⁸¹Kr^m were determined as function of projection angle and body surface area. These functions in combination with the projections corresponding to window W_iKr where used to correct the combined ⁸¹Kr^m/⁹⁹Tc^m projections for the presence of ⁸¹Kr^m. A comparison between these corrected ⁸¹Kr^m/⁹⁹Tc^m projections and the ⁹⁹Tc^m only projections demonstrated not significant differences indicating the good accuracy of the implemented correction methodology