RSNA 2014 

Abstract Archives of the RSNA, 2014


VSIR21-11

Predictive Quantification of Infarction Volume before Partial Splenic Embolization for Hypersplenism

Scientific Papers

Presented on December 1, 2014
Presented as part of VSIR21: Interventional Series: Embolotherapy

Participants

Toshihiro Tanaka MD, Presenter: Nothing to Disclose
Tetsuya Masada, Abstract Co-Author: Nothing to Disclose
Hideyuki Nishiofuku, Abstract Co-Author: Nothing to Disclose
Takeshi Sato, Abstract Co-Author: Nothing to Disclose
Shinsaku Maeda, Abstract Co-Author: Nothing to Disclose
Kimihiko Kichikawa MD, Abstract Co-Author: Nothing to Disclose
Hiroshi Anai MD, PhD, Abstract Co-Author: Nothing to Disclose
Masayoshi Inoue MD, Abstract Co-Author: Nothing to Disclose

PURPOSE

To obtain the optimal splenic infarction volume is the key to achieve high efficacy and to reduce the risk of complications after partial splenic embolization (PSE). We have developed a new system to predict the infarction splenic volume before PSE using computed volumetric analysis software. The aim of this study is to evaluate the accuracy of this prediction system.

METHOD AND MATERIALS

The data, from 12 patients with hypersplenism who had received PSE, was retrospectively analyzed. 3-dimensional (3-D) arteriography image was reconstructed from the contrast enhanced CT obtained before PSE. Using a 3-D image analysis system (SYNAPSE VINCENTTM), the volume of the area supplied from each splenic branch was calculated based on the Voronoi Diagram. The estimated infarction volume was defined by the total sum of the volume supplied from each embolized branch. The actual infarction volume was calculated on the contrast enhanced CT obtained 1 week after PSE. Pearson Correlation Coefficients was used to assess the correlation between the estimated infarction volume and the actual infarction volume.

RESULTS

The mean estimated and actual infarction volumes were 65.4±14.6% and 60.9±10.2%, respectively. The mean difference between them was 7.29±6.93%. The actual infarction volume was strongly correlated with the estimated infarction volume (ρ= 0.791, P=0.002). There were no complications in any of the patients. The mean platelet count significantly increased from 7.96×104/µL before PSE to 15.6×104/µL two weeks after PSE (the increased ratio: 259±93.9%).

CONCLUSION

Our results demonstrated that infarction splenic volume can be precisely predicted before PSE using computed volumetric analysis software. This new system could be helpful for tailoring planning of PSE to achieve optimal splenic infarction volume in patients with hypersplenism.

CLINICAL RELEVANCE/APPLICATION

Predictive quantification of splenic infarction volume using the Voronoi Diagram method is accurate, which could be useful for planning before PSE.

Cite This Abstract

Tanaka, T, Masada, T, Nishiofuku, H, Sato, T, Maeda, S, Kichikawa, K, Anai, H, Inoue, M, Predictive Quantification of Infarction Volume before Partial Splenic Embolization for Hypersplenism.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14012541.html