Preoperative assessment of parietal pleural invasion/adhesion of subpleural lung cancer: advantage of software-assisted analysis of 4-dimensional dynamic-ventilation computed tomography

Eur Radiol. 2019 Oct;29(10):5247-5252. doi: 10.1007/s00330-019-06131-w. Epub 2019 Mar 26.

Abstract

Objective: To evaluate the accuracy of four-dimensional (4D) dynamic-ventilation computed tomography (CT) scanning coupled with our novel image analysis software to diagnose parietal pleural invasion/adhesion of peripheral (subpleural) lung cancer.

Methods: Eighteen patients with subpleural lung cancer underwent both 4D dynamic-ventilation CT during free breathing and conventional (static) chest CT during preoperative assessment. The absence of parietal pleural invasion/adhesion was surgically confirmed in 13 patients, while the presence of parietal pleural invasion/adhesion was confirmed in 5 patients. Two chest radiologists, who were blinded to patient status, cooperatively evaluated the presence of pleural invasion/adhesion using two different imaging modalities: (i) conventional high-resolution CT images, reconstructed in the axial, coronal, and sagittal directions, and (ii) 4D dynamic-ventilation CT images combined with a color map created by image analysis software to visualize movement differences between the lung surface and chest wall. Parameters of diagnostic accuracy were assessed, including a receiver operating characteristic analysis.

Results: Software-assisted 4D dynamic-ventilation CT images achieved perfect diagnostic accuracy for pleural invasion/adhesion (sensitivity, 100%; specificity, 100%; area under the curve [AUC], 1.000) compared to conventional chest CT (sensitivity, 60%; specificity, 77%; AUC, 0.846).

Conclusion: Software-assisted 4D dynamic-ventilation CT can be considered as a novel imaging approach for accurate preoperative analysis of pleural invasion/adhesion of peripheral lung cancer.

Key points: • 4D dynamic-ventilation CT can correctly assess parietal pleural invasion/adhesion of peripheral lung cancer. • A unique color map clearly demonstrates parietal pleural invasion/adhesion. • Our technique can be expanded to diagnose "benign" pleural adhesions for safer thoracoscopic surgery.

Keywords: Computer-assisted image analysis; Four-dimensional computed tomography; Lung cancer.

Publication types

  • Observational Study

MeSH terms

  • Aged
  • Aged, 80 and over
  • Female
  • Four-Dimensional Computed Tomography / methods
  • Humans
  • Lung / pathology
  • Lung Neoplasms / diagnostic imaging
  • Lung Neoplasms / pathology*
  • Lung Neoplasms / surgery
  • Male
  • Middle Aged
  • Neoplasm Invasiveness
  • Pleural Neoplasms / diagnostic imaging
  • Pleural Neoplasms / pathology*
  • Preoperative Care / methods
  • ROC Curve
  • Respiration
  • Retrospective Studies
  • Sensitivity and Specificity
  • Software
  • Tissue Adhesions / diagnostic imaging
  • Tissue Adhesions / pathology