Yukihiro Nagatani1, Masayuki Hashimoto2, Yasuhiko Oshio3, Shigetaka Sato4, Jun Hanaoka3, Kentaro Fukunaga5, Ryo Uemura4, Makoto Yoshigoe4, Norihisa Nitta4, Noritoshi Usio4, Shinsuke Tsukagoshi6, Tatsuya Kimoto7, Tsuneo Yamashiro8, Hiroshi Moriya9, Kiyoshi Murata4, Yoshiyuki Watanabe4. 1. Department of Radiology, Shiga University of Medical Science, Otsu, Shiga, 520-2192, Japan. Electronic address: yatsushi@belle.shiga-med.ac.jp. 2. Department of Thoracic Surgery, Kyoto Medical Center, Kyoto, Kyoto, 612-8555, Japan; Division of General Thoracic Surgery, Department of Surgery, Shiga University of Medical Science, Seta-tsukinowa-cho, Otsu, Shiga, 520-2192, Japan. 3. Division of General Thoracic Surgery, Department of Surgery, Shiga University of Medical Science, Seta-tsukinowa-cho, Otsu, Shiga, 520-2192, Japan. 4. Department of Radiology, Shiga University of Medical Science, Otsu, Shiga, 520-2192, Japan. 5. Division of Respiratory Medicine, Department of Internal Medicine, Shiga University of Medical Science, Seta-tsukinowa-cho, Otsu, Shiga, 520-2192, Japan. 6. CT System Division, Canon Medical Systems, Otawara, Tochigi, 324-8550, Japan. 7. Department of Radio Center for Medical Research and Development, Canon Medical Systems, Otawara, Tochigi, 324-8550, Japan. 8. Department of Radiology, Graduate School of Medical Science, University of the Ryukyus, Nishihara, Okinawa, 903-0215, Japan. 9. Department of Radiology, Ohara General Hospital, Fukushima, Fukushima, 960-8611, Japan.
Abstract
PURPOSE: To assess the usefulness of software analysis using dynamic-ventilation CT for localized pleural adhesion (LPA). MATERIALS AND METHODS: Fifty-one patients scheduled to undergo surgery underwent both dynamic-ventilation CT and static chest CT as preoperative assessments. Five observers independently evaluated the presence and severity of LPA on a three-point scale (non, mild, and severe LPA) for 9 pleural regions (upper, middle, and lower pleural aspects on ventral, lateral, and dorsal areas) on the chest CT by three different methods by observing images from: static high-resolution CT (static image); dynamic-ventilation CT (movie image), and dynamic-ventilation CT while referring to the adhesion map (movie image with color map), which was created using research software to visualize movement differences between the lung surface and chest wall. The presence and severity of LPA was confirmed by intraoperative thoracoscopic findings. Parameters of diagnostic accuracy for LPA presence and severity were assessed among the three methods using Wilcoxon signed rank test in total and for each of the three pleural aspects. RESULTS: Mild and severe LPA were confirmed in 14 and 8 patients. Movie image with color map had higher sensitivity (56.9 ± 10.7 %) and negative predictive value (NPV) (91.4 ± 1.7 %) in LPA detection than both movie image and static image. Additionally, for severe LPA, detection sensitivity was the highest with movie image with color map (82.5 ± 6.1 %), followed by movie image (58.8 ± 17.0 %) and static image (38.8 ± 13.9 %). For LPA severity, movie image with color map was similar to movie image and superior to static image in accuracy as well as underestimation and overestimation, with a mean value of 80.2 %. CONCLUSION: Software-assisted dynamic-ventilation CT may be a useful novel imaging approach to improve the detection performance of LPA.
PURPOSE: To assess the usefulness of software analysis using dynamic-ventilation CT for localized pleural adhesion (LPA). MATERIALS AND METHODS: Fifty-one patients scheduled to undergo surgery underwent both dynamic-ventilation CT and static chest CT as preoperative assessments. Five observers independently evaluated the presence and severity of LPA on a three-point scale (non, mild, and severe LPA) for 9 pleural regions (upper, middle, and lower pleural aspects on ventral, lateral, and dorsal areas) on the chest CT by three different methods by observing images from: static high-resolution CT (static image); dynamic-ventilation CT (movie image), and dynamic-ventilation CT while referring to the adhesion map (movie image with color map), which was created using research software to visualize movement differences between the lung surface and chest wall. The presence and severity of LPA was confirmed by intraoperative thoracoscopic findings. Parameters of diagnostic accuracy for LPA presence and severity were assessed among the three methods using Wilcoxon signed rank test in total and for each of the three pleural aspects. RESULTS: Mild and severe LPA were confirmed in 14 and 8 patients. Movie image with color map had higher sensitivity (56.9 ± 10.7 %) and negative predictive value (NPV) (91.4 ± 1.7 %) in LPA detection than both movie image and static image. Additionally, for severe LPA, detection sensitivity was the highest with movie image with color map (82.5 ± 6.1 %), followed by movie image (58.8 ± 17.0 %) and static image (38.8 ± 13.9 %). For LPA severity, movie image with color map was similar to movie image and superior to static image in accuracy as well as underestimation and overestimation, with a mean value of 80.2 %. CONCLUSION: Software-assisted dynamic-ventilation CT may be a useful novel imaging approach to improve the detection performance of LPA.