For ventilation imaging, several alternatives exist. ⁵ These include the following:

• 
  

  Inert radioactive gases such as krypton-81 m (^81mKr) and xenon-133 (¹³³Xe).

  
  
• 
  

  Radiolabeled aerosols, most commonly ^99mTc-diethylenetriaminepentaacetic acid (^99mTc-DTPA) but occasionally sulfur colloid or albumen.

  
  
• 
  

  Technegas (^99mTc ultrafine carbon suspension) (Cyclopharm, Sydney, Australia).

Each has its advantages and disadvantages (► Table 6.1); however, most centers performing SPECT would use either Technegas (preferable, especially if underlying airways disease ⁶ )or ^99mTc-DTPA aerosol. ^81mKr is also an ideal ventilation agent for SPECT; however, it is expensive, requires continuous administration, it is no longer available in the United States, and its use is limited to some European centers. Although still used in the United States for planar imaging, ¹³³Xe is unsuitable for SPECT due to its low-energy photons leading to poor resolution and high scatter.

^99mTc-macroaggregated albumin (^99mTc-MAA) is generally used to assess perfusion. ⁷

The dose of ^99mTc-MAA used is dependent on the ventilation agent used.

If a radioactive gas is used, the dose of perfusion agent is typically lower than if a technetium-based ventilation agent is used as the signal from the radioactive gas can be separated from that of the perfusion agent based on the energy level of the emitted photons.

If a technetium-based agent is used for both ventilation and perfusion imaging, a greater dose of perfusion agent (resulting in a perfusion-ventilation dose ratio of at least 4:1) is required to “drown out” the underlying ventilation signal. ⁷

The administered activity of ^99mTc-MAA is typically 2.5 to 6 mCi (100-250 MBq).5,7

For pregnant patients, dose reduction is usually implemented. This can be achieved by reducing the administered dose of both the ventilation and perfusion agents, usually by half. ⁸ This will necessitate a longer acquisition time to maintain images of good quality. Some centers advocate omitting the ventilation scan; however, the radiation savings from this approach are minimal, and diagnostic accuracy may be adversely impacted. ⁹ In centers performing SPECT/CT imaging, consideration can be given to omitting the CT scan to reduce breast radiation exposure.

Multihead gamma cameras (either dual or triple head) are required for SPECT imaging. ⁵ Acquisition times become prohibitive for standard clinical practice if a single-head camera is used.

A typical protocol that uses a multihead camera requires 15 to 25 minutes of total acquisition time for a V/Qdata set. ⁵ , ¹⁰ This is faster than traditional six- or eight-view planar imaging. ¹⁰ , ¹¹

If V/Q can be performed simultaneously (e.g., when ^81mKr is used for ventilation), the acquisition time is halved. ¹²

Representative acquisition and processing parameters are shown in ► Table 6.2. Note that if ^81mKr is used for ventilation, a medium energy collimator is typically used.

SPECT/CT imaging is acquired on a hybrid SPECT/CT scanner.

While most hybrid scanners can perform diagnostic quality CT, they can also be operated solely for attenuation correction and anatomical localization using “low-dose” parameters. ¹³

For lung scanning, the CT acquisition is typically acquired immediately after the perfusion SPECT. Intravenous contrast is not required, and a reduced beam current of the order of 20 to 80 mA will suffice. ⁴ This results in an absorbed dose of the order of 1 to 2 mSv. ⁷ , ⁹ , ¹² , ¹⁴ This compares favorably with the 2 to 2.5 mSv from the V/Q scan itself and is well below the levels received from a diagnostic CTPA. ⁹ , ¹⁵ , ¹⁶

Due to the duration of acquisition time for SPECT imaging, breath holding is not feasible, and these are therefore performed during free breathing. To better match the CT images to SPECT acquisitions, it has been suggested that they be performed using a mid inspiratory breath hold, or during shallow breathing, rather than at maximal inspiration. ¹⁷

The CT acquisition time is rapid (< 1 minute), and combined with the setup time, the procedure adds only 1 to 2 minutes to a V/QSPECT study.

Image review should occur following coregistration of the V/QSPECT data sets (as well as the CT data sets in the case of SPECT/Cr).

Images are best viewed simultaneously in transverse, coronal, and sagittal planes on a workstation (► Fig. 6.3). This allows the reporter to interactively examine and triangulate the linked V/Q SPECT studies as well as the CT in each of the three orthogonal imaging planes and to adjust the relative image intensities, especially of fused images (► Fig. 6.4).

Review of images on a workstation also facilitates the viewing of CT data in different windows so that lungs, soft tissue, and bones can all be reviewed, as appropriate.

Segmental lung anatomy charts may be of help for reporting specialists, particularly for reporters who are inexperienced (► Fig. 6.5). ⁸

V/Q SPECT studies are usually reported using the European Association of Nuclear Medicine (EANM) reporting guidelines. These guidelines recommend that studies be reported as positive for PE if there is V/Q mismatch of at least one segment or two subsegments that conform to pulmonary vascular anatomy (► Table 6.3). ⁷ Probabilistic reporting (as used for planar scanning) is not recommended, and has not been validated, for V/Q SPECT. ¹¹ , ¹⁸ While the EANM reporting guidelines do not specifically address hybrid imaging, these guidelines are also generally used to report V/ Q SPECT/CT studies. Indeed, the addition of the CT component is likely to help classify the V/Q SPECT pattern more appropriately. The CT will provide patient-specific anatomical information, including the lung and segment borders, fissures, and the location of major vessels, as well as the presence of parenchymal disease (► Fig. 6.6).

Further demonstration of V/Q abnormalities can be made by creating a quotient or ratio image. In these images, the ratio of counts is usually derived by the formula (q-v₂),where V ₁ are counts at time of ventilation acquisition and V₂ decay corrected counts at time of perfusion acquisition. There are novel ways of displaying V:Q quotient data to assist image reporting. Palmer et al have described a technique where these images can be presented as either 3D surface-shaded images or as tomographic sections in each of the orthogonal planes. ¹⁹ These “quotient images” can be helpful in facilitating image reporting and are a useful way of demonstrating the location and extent of mismatched defects (► Fig. 6.7).

Planar-like images can be generated from SPECT data using several approaches. While Bailey et al have described a technique using reprojection, ²⁰ many of the commercial vendors offer a simpler approach using an “angular summing” technique. ¹¹ With this approach, images are generated by summing several consecutive projections from the SPECT acquisition. This can blur small defects as data are acquired over an arc; however, the images produced approximate true planar images (► Fig. 6.8). ²¹ These “pseudoplanar” images give a familiar and rapid view of the lungs for quick evaluation and may be of particular value during the transition phase from planar imaging to SPECT imaging. ⁸ As the images are generated from the SPECT data, no additional image acquisition time is required.