Kana Sakamoto, Presenter: Nothing to Disclose

Hiroshi Mogami MD, Abstract Co-Author: Nothing to Disclose

Tomoko Hyodo Kajiwara, Abstract Co-Author: Nothing to Disclose

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

Yoshifumi Yasuhara MD, Abstract Co-Author: Nothing to Disclose

Masao Miyagawa MD,PhD, Abstract Co-Author: Nothing to Disclose

Teruhito Mochizuki MD, Abstract Co-Author: Nothing to Disclose

et al, Abstract Co-Author: Nothing to Disclose

The aim of this study was to evaluate detectability of lung cancer by a chest radiograph with a one-shot dual energy subtraction method (ES method) that produces, in addition to conventional chest radiographs, both bone and soft-tissue images.

We used a computed radiography system (FCR XU-D1) with a dedicated ES software, which was originated by Fuji Computed Radiography, Tokyo, Japan. Six radiologists read two sets of chest radiographs from 60 pts (66.7±9.3 y.o.) with histologically proven lung cancer (49 adenoca., 11 squamous cell ca.) and those from 60 normal subjects (57.6±10.7y.o.). They read standard computed radiography images (CR) only: A method, or the combination of CR and ES images: B method, and we evaluated the effect of ES method in detection of lung cancer with chest radiographs via an observer performance study with use of receiver operating characteristic (ROC) analysis. All patients were classified into three groups according to the appearance on high-resolution CT; 1) non-solid ground glass opacity: pure GGO group, 2) part-solid GGO: mixed group, 3) solid part only group: solid group. And mixed group was subclassified into the GGO-dominant type, that contains a solid part but did not exceed 50% of whole area, and the solid-dominant type, that contains a solid part more than 50% of whole area.

In the pure GGO group, detection of lung cancer was difficult both with A and B methods (0/7). In the solid group, we could detect 30/34 (88%) with A method and 34/34 (100%) with B, however, there had no difference between the detectability of two methods (A: Az=0.994, B: Az=0.994, n.s.). In the mixed group, B method had significantly superior detectability to that of A method in the GGO-dominant type, (A: Az=0.708, B: Az=0.861, p<0.01), while there had no difference between the detectability of two methods in the solid-dominant type (A: Az=0.821, B: Az=0.864, n.s.). Consequently, overall detection of lung cancer was significantly better with B method than with A method from ROC analysis (A: Az=0.841, B: Az=0.892, p<0.05).

Thus, a one-shot ES method is superior to the standard CR images in the detection of lung cancer, especially in mixed group.

The aim of this study was to evaluate detectability of lung cancer by a chest radiograph with a one-shot dual energy subtraction method (ES method) that produces, in addition to conventional chest radiographs, both bone and soft-tissue images.

We used a computed radiography system (FCR XU-D1) with a dedicated ES software, which was originated by Fuji Computed Radiography, Tokyo, Japan. Six radiologists read two sets of chest radiographs from 60 pts (66.7±9.3 y.o.) with histologically proven lung cancer (49 adenoca., 11 squamous cell ca.) and those from 60 normal subjects (57.6±10.7y.o.). They read standard computed radiography images (CR) only: A method, or the combination of CR and ES images: B method, and we evaluated the effect of ES method in detection of lung cancer with chest radiographs via an observer performance study with use of receiver operating characteristic (ROC) analysis. All patients were classified into three groups according to the appearance on high-resolution CT; 1) non-solid ground glass opacity: pure GGO group, 2) part-solid GGO: mixed group, 3) solid part only group: solid group. And mixed group was subclassified into the GGO-dominant type, that contains a solid part but did not exceed 50% of whole area, and the solid-dominant type, that contains a solid part more than 50% of whole area.

In the pure GGO group, detection of lung cancer was difficult both with A and B methods (0/7). In the solid group, we could detect 30/34 (88%) with A method and 34/34 (100%) with B, however, there had no difference between the detectability of two methods (A: Az=0.994, B: Az=0.994, n.s.). In the mixed group, B method had significantly superior detectability to that of A method in the GGO-dominant type, (A: Az=0.708, B: Az=0.861, p<0.01), while there had no difference between the detectability of two methods in the solid-dominant type (A: Az=0.821, B: Az=0.864, n.s.). Consequently, overall detection of lung cancer was significantly better with B method than with A method from ROC analysis (A: Az=0.841, B: Az=0.892, p<0.05).

Thus, a one-shot ES method is superior to the standard CR images in the detection of lung cancer, especially in mixed group.

Sakamoto, K, Mogami, H, Kajiwara, T, Inoue, T, Yasuhara, Y, Miyagawa, M, Mochizuki, T, et al, , The Evaluation of the Detectability in Lung Cancer by a Chest Radiograph Using a One-shot Dual Energy Subtraction Method.  Radiological Society of North America 2005 Scientific Assembly and Annual Meeting, November 27 - December 2, 2005 ,Chicago IL. http://archive.rsna.org/2005/4406433.html