PT - JOURNAL ARTICLE AU - Keisuke Kiso AU - Emi Tateishi AU - Yoshihiro Nishimura AU - Yoshitake Takahashi AU - Norifumi Nakanishi AU - Hatsue Ishibashi-Ueda TI - Clinical utility of fractal analysis with lung perfusion scintigraphy for the differential diagnosis of pulmonary veno-occlusive disease DP - 2014 May 01 TA - Journal of Nuclear Medicine PG - 532--532 VI - 55 IP - supplement 1 4099 - http://jnm.snmjournals.org/content/55/supplement_1/532.short 4100 - http://jnm.snmjournals.org/content/55/supplement_1/532.full SO - J Nucl Med2014 May 01; 55 AB - 532 Objectives Pulmonary veno-occlusive disease (PVOD) is rare disease which causes severe pulmonary artery hypertension (PAH). Moreover, PVOD is refractory to recent advanced medical therapy and its prognosis is very poor. It is extremely difficult to make early diagnosis of this disease, because of poor pathognomonic findings, except histopathologic diagnosis. Some reports revealed that fractal analysis of lung perfusion scintigraphy (LPS) with 99mTc-MAA (MAA) might help differentiation of PAH diseases. In this study, we investigated the clinical utility of fractal analysis with LPS for differential diagnosis of PVOD, compared with normal subjects and other PAH disease. Methods We evaluated antemortem LPS with MAA of 5 patients with PVOD who were diagnosed by autopsy. And we also evaluated LPS of six normal subjects and 7 patients with chronic thromboembolic pulmonary hypertension (CTEPH). Then we applied “pixel” and “box” counting method of fractal analysis to these LPS data, and we compared fractal dimensions in these study groups. Results In box counting method, PVOD showed significantly higher fractal dimension than other two groups (CTEPH: Normal: PVOD = 1.34 ± 0.090: 1.35 ± 0.063:1 .42 ± 0.067, P < 0.05). In pixel counting method, PVOD also showed higher fractal dimension than other two groups (CTEPH: Normal: PVOD = 0.18 ± 0.041: 0.21 ± 0.037: 0.24 ± 0.055, P < 0.01). Moreover, combined analysis with box and pixel counting methods could differentiate PVOD from other two groups with sensitivity=60% and specificity=100%. Conclusions Fractal analysis combined with pixel and box counting methods might be helpful for the differential diagnosis of PVOD from normal and CTEPH subjects.