Accurate and reproducible reporting of lung scintigraphy is predicated on a sound knowledge of the segmental anatomy of the lungs. A limited amount of hard data exists about the true segmental anatomy of the lungs. A virtual model of human lungs was created using a CT-based dataset and a Monte Carlo simulation technique to examine the optimal projections for the visualization of each segment in the lungs. Methods: Segmental anatomy of the lungs was modeled using CT, cadaveric lungs and standard anatomical texts. The emission, scatter and attenuation of photons was modeled within these virtual lungs and the surrounding tissues. Single segmental lesions were created in eight projections and submitted for blinded reporting to four experienced nuclear medicine physicians to obtain the best views for each segment. Results: The anterior end posterior oblique projections yielded the best views for 10 of 18 segments, with the laterals contributing four views, the anterior contributing two views and the posterior contributing one view. The majority of basal segments (six of nine) were best seen in the anterior and posterior oblique projections. Conclusion: This model overcomes the major problems associated with experimentation in the normal human and has the potential to provide answers to the major problems of scatter, attenuation and 'shine-through' in lung scintigraphy.
|Number of pages||5|
|Journal||Journal of Nuclear Medicine|
|Publication status||Published - Dec 1997|
- Lung scintigraphy
- Monte Carlo simulation
- Pulmonary embolism
- Segmental anatomy