Diagnosis of pulmonary embolism

Dr Joachim Kropp from Dresden has developed a simple system using a thermistor in the nostril of a patient to sense the breathing cycle and using the computer's cardiac gating routine to create 8 sets of images of his V/Q lungs. He presented his work at a 'Technegas Colloquium' at the EANM meeting in Berlin in 1998.  Here with the abstract and one of the more SPECTacular examples of the resultant images.

E-mail
Department of Nuclear Medicine
University Hospital of the TU, Dresden
Fetscherstr. 74 01307, Dresden, Germany

Planar pulmonary scintigraphy is limited to resolve sub-segmental PE, and is improved by 'normal' SPECT (nSPECT) acquisition, but still hampered by movement artifacts. A Gated SPECT (gSPECT) protocol using a dual head, 64x64 matrix, 64 views, 8 time bins, 15 breath cycles per view was devised and tested first in a phantom. In patients, gSPECT was realised by registration of the temperature difference of the in- and ex-haled breath to derive the gating.

In 101 patients with DVT proven by duplex sonography, V/Q scintigraphy was performed using Tc-99m-labelled MAA and Technegas with both n and g SPECT immediately before and 10 days after initialisation of two different therapy regimes (standard [group 1] or low molecular weight heparin [group 2]; early [group 2], versus late [group 1] mobilisation). All scans were read 'blind' by two experienced Physicians. Interobserver agreement was 82% and 94% for nSPECT and gSPECT respectively. Initially 59.6% of patients had PE with no difference between the two groups. 66.1%[group 1] and 61%[group 2] patients showed no change. In the two groups with different therapies, improvement of PE was demonstrated in both in 17% of the patients whereas in 16.9%[group 1] and 22%[group 2] new PE was diagnosed (p=0.46). We conclude that by gated pulmonary SPECT there is an improvement in spatial resolution and contrast which leads to a more reliable diagnosis of PE in V/Q scintigraphy.