Seitaro Oda MD, Presenter: Nothing to Disclose

Kazuo Awai MD, Abstract Co-Author: Nothing to Disclose

Funama Yoshinori, Abstract Co-Author: Nothing to Disclose

Takeshi Nakaura MD, Abstract Co-Author: Nothing to Disclose

Yumi Yanaga MD, Abstract Co-Author: Nothing to Disclose

Yasuyuki Yamashita MD, Abstract Co-Author: Nothing to Disclose

Shinichi Nakamura MD, PhD, Abstract Co-Author: Nothing to Disclose

Koichi Kawanaka MD, Abstract Co-Author: Nothing to Disclose

00030490-DMT et al, Abstract Co-Author: Nothing to Disclose

We investigated the effect of nodule density and location on the diagnostic performance of radiologists detecting pulmonary nodules on dual-energy subtraction chest radiographs.

Using an FPD radiography system (Revolution XR/d, GE) we obtained 41 sets of chest radiographs (Cx) from 26 patients (aged 62.9 ± 14.0 years) with solitary pulmonary nodules measuring ≤ 2.0 cm and 15 normal subjects (aged 55.5 ± 7.0 years). Each data set also contained a standard- and a corresponding energy-subtracted Cx [ES-Cx] (soft tissue image). There were 6 non-solid ground glass opacity (GGO)-, 10 part-solid GGO-, and 10 solid nodules. They were histologically confirmed as bronchioloalveolar carcinoma (n=13), adenocarcinoma (n=9), squamous cell carcinoma (n=1), metastatic lung tumor (n=2), and hamartoma (n=1). The mean size of the 26 nodules was 15.0 ± 4.8 mm (standard deviation). We performed receiver operating characteristic (ROC) analysis using a continuous rating scale to compare the observers’ performance in detecting the nodules, first without and then with ES-Cx. Seven board-certified radiologists participated in this observer study.

For the 7 observers, the mean value of the area under the best-fit ROC curve (Az) without and with ES-Cx was 0.62 ± 0.05 and 0.68 ± 0.06, respectively; the difference was statistically significant (P<0.05). For mixed GGO nodules, the difference of the mean Az values obtained without and with ES-Cx was statistically significant (Az=0.61 ± 0.07 vs 0.69 ± 0.05, respectively; P<0.01); for pure GGO nodules it was not (Az=0.62 ± 0.10 vs 0.61 ± 0.09, respectively; P=0.73) and for solid nodules it was not (Az=0.75 ± 0.096 vs 0.78 ± 0.08, respectively; P=0.26). For nodules with overlapping by bone shadows the difference of the mean Az values was statistically significant (Az=0.66 ± 0.05 vs 0.72 ± 0.05, respectively; P<0.05), for nodules without overlapping it was not (Az=0.58 ± 0.06 vs 0.62 ± 0.06, respectively; P=0.26)

Use of ES-Cxs improved the diagnostic performance of radiologists detecting small pulmonary nodules, especially for mixed GGO nodules and nodules with overlapping bone shadows.

The dual-energy subtraction technique may improve the detectability of small pulmonary nodules on FPD chest radiographs, especially in mixed GGO nodules and nodules with overlapping bone shadows.

Oda, S, Awai, K, Yoshinori, F, Nakaura, T, Yanaga, Y, Yamashita, Y, Nakamura, S, Kawanaka, K, et al, 0, Detection of Small Pulmonary Nodules by Using Dual-Energy Subtraction Chest Radiographs: Effect of Nodule Density and Location on Radiologists' Diagnostic Performance.  Radiological Society of North America 2009 Scientific Assembly and Annual Meeting, November 29 - December 4, 2009 ,Chicago IL. http://archive.rsna.org/2009/8005019.html Accessed October 23, 2025