Keywords:
Radioprotection / Radiation dose, Thorax, Respiratory system, CAD, CT, Computer Applications-Detection, diagnosis, Dosimetry, Cancer, Multidisciplinary cancer care
Authors:
A. Christe1, L. Leidolt1, A. Huber1, P. Steiger1, Z. Szucs-Farkas2, J. Roos3, J. Heverhagen1, L. Ebner1; 1Bern/CH, 2Biel/ Bienne/CH, 3Winterthur/CH
DOI:
10.1594/ecr2013/C-2614
Purpose
The purpose of this study was the evaluation of two radiologists and three different CAD-systems as first and second readers for lung cancer screening at various CT dose levels including low kilo-voltage settings,
to find the best reader pairing at the lowest acceptable radiation exposure level. Sensitivity of CT for lung nodules depends on many things like nodule size,
nodule density,
nodule location,
image quality,
reader and post-processing tools [1-7].
Most of these given variables may not be changed,
except for the last ones.
Therefore,
it makes sense to improve these variables in order to increase detectability by a second reader,
either with a second radiologist or computer assisted detection system.
Wormanns et al.
reported a 11 to 16% higher sensitivity,
when a second reading was performed by a different radiologist.
Previous investigators stated the usefulness of a computer assisted detection software (CAD) [6-10].
An increase of sensitivity between 2 and 22% was reported,
when an additional CAD was used [6,
9,
11,
12].
However,
Lee et al.
[13] showed that the sensitivity of a CAD system as standalone tool (81%) did not significantly differ from that of radiologists (85%).
Radiologists were more sensitive at detecting nodules attached to other structures,
whereas CAD was better at detecting isolated nodules and those that were ≤ 5 mm in diameter [13].
But until now there is little results available on how accurate a CAD is working at low dose levels in combination with radiologists or with a second CAD system: Hein et al.
described a feasibility of low dose CAD at 5mAs/120kVp and Das at 10 mAs/120kVp.
Furthermore,
previous studies have shown,
that diagnostic image quality does not suffer at lower dose levels [1,3,14].
To maintain diagnostic quality in the detection of lung nodules tube currents can be reduced to well below 100 mAs at a constant voltage of 120 kVp [2,3,14,15,16].
Some authors even proposed threshold tube currents of 20 mAs (at 120 kVp) to detect ground glass nodules,
alveolar consolidation and lung nodules [3,16].
To decrease tube voltage is another strategy to reduce radiation dose in CT.
There are still high voltages (100-140 kVp) widely applied in radiology,
although results from vascular studies suggest that a reduction to 80 kVp is possible [17-20].
There is very little data on the combination of low kVp and mAs for lung nodule detection.
In screening CT the prevalence of part solid or ground glass nodules is reported in the literature between 13% to 40% [21-23].
Furthermore,
peripheral pulmonary adenocarcinoma represents up to 35 % of all lung cancer and usually demonstrates as GGN in the beginning [24,
25].
Therefore it is important to intersperse ground glass nodules (GGN) to simulate a screening setting.