Author of

• 10946

  +

  P1.19 - Poster Session 1 - Imaging (ID 179)

  • Event: WCLC 2013
  • Type: Poster Session
  • Track: Imaging, Staging & Screening
  • Presentations: 1
  • Moderators:
  • Coordinates: 10/28/2013, 09:30 - 16:30, Exhibit Hall, Ground Level

  • 10951

    +

    P1.19-005 - Computer-aided lung nodule analysis focused on ground glass opacity and consolidation on thin-section computed tomography. (ID 1447)

    09:30 - 09:30  |  Author(s): E. Yamada
    • Abstract

    Background
    Ground-glass opacity (GGO) component in a nodule on thin-section computed tomography (TSCT) often corresponds to a lepidic growth pattern of adenocarcinoma. In contrast, solid attenuation or consolidation on TSCT corresponds to invasive components. Many researchers reported consolidation tumor ratio (CTR; defined as the ratio of the size of solid attenuation to the maximum tumor dimension) was a reliable parameter in predicting tumor invasiveness. However, it has been pointed out that inter-/intra-observer variability in CTR measurement is a major problem in precise and reproducible evaluation of tumor characteristics. The aim of this study was to determine the optimal CT settings to reproducibly diagnose GGO and consolidation areas on TSCT by using an imaging software.

    Methods
    We reviewed preoperative TSCT images of the patients undergoing surgical resection for T1 lung adenocarcinoma in our institution between 2005 and 2009. The TSCT images were obtained without contrast enhancement and reconstructed in 1.0 or 2 mm thickness, using several CT systems. The imaging software colored GGO areas with cut-off CT levels of -800, -700 and -600 HU. Consolidation areas were colored with cut-off CT levels of -300, -200 and -100 HU. These GGO/consolidations identified by the software were compared with those visually determined by the consensus of the 4 authors (EY, KA, YM, HO). The 4 authors scored the correspondence between visual evaluation and software identification according to the cut-off levels. The scores were summarized to determine the optimal cut-off CT levels.

    Results
    We have reviewed 20 patients so far. Figure 1 shows a TSCT image and software-yielded image showing good correspondence with each other of GGO and consolidation areas. The best score was obtained when the cut-off level was -700 HU for GGO and -200 HU for consolidation. Figure1. Figure 1

    Conclusion
    Although based on a small cohort, we found optimal cut-off CT levels to identify GGO and consolidation areas using an imaging software. We need to analyze more cases, but this image analysis method is promising in determining CTR reproducibly.