| vol) of no more than 3 mGy in a standard-size patient (height, 170 cm; weight, 70 kg), as per ACR recommendations for screening CT (8).
A number of dose reduction techniques, including dose modulation and iterative reconstruction, may be used (13). It is important that a similar technique be used to perform the follow-up examination to minimize interscan variability, with section thickness and reconstruction filter being the most important parameters in this respect. For these guidelines, manual nodule measurements should be based on the average of long- and short-axis diameters, both of which should be obtained on the same transverse, coronal, or sagittal reconstructed images. Whichever image reveals the greatest dimensions is the image that should be used. Measurements should be made with electronic calipers or semiautomated methods and should be recorded to the nearest whole millimeter (grade 1C; strong recommendation, low- or very-low-quality evidence). Although several screening trials have used the maximum diameter of nodules on transverse sections to estimate size, others (iELCAP) have used the average of long- and short-axis diameters measured by using lung windows (4,5,7,14,15). Prediction models used to estimate malignancy yield better results with the average diameter than with the maximum transverse diameter (16). The Fleischner Society has recommended use of the average diameter since 2005, as the average of long and short axes more accurately reflects three-dimensional tumor volume (1). For larger nodules and for masses larger than 10 mm, it is generally appropriate to record both long- and short-axis dimensions, with the long-axis dimension being used to determine the T factor in lung cancer staging and being a criterion for tumor response to treatment. Measurements should be rounded to the nearest millimeter. Fractional millimeter measurements are not recommended, as their use implies a greater degree of accuracy than that which can be achieved in practice. Thus, the size threshold (<6 mm) corresponds to a rounded measurement of 5 mm or less in these guidelines. As an alternative to manual linear measurements, automated or semiautomated volumetric measurements can be used, and they have the advantage of being more reproducible than manual techniques (17). Volume thresholds of 100 and 250 mm3 are used for volumetry instead of the 6- and 8-mm thresholds used for linear measurements. However, volumetry is substantially dependent on the specific software used (18,19). For this reason, volumetric measurements to assess nodule growth should be performed with identical software versions. More comprehensive recommendations on nodule measurements, including a full discussion of technical and observer-related factors, will be provided in a separate White Paper from the Fleischner Society that is currently in preparation. |