| Autor: |
R. Widita, R. Kurniadi, F. Haryanto, Y. Darma, Y. S. Perkasa, S. S. Zasneda, Zaki Su’ud, A. Waris |
| Rok vydání: |
2010 |
| Předmět: |
|
| Zdroj: |
AIP Conference Proceedings. |
| ISSN: |
0094-243X |
| DOI: |
10.1063/1.3462757 |
| Popis: |
We have been successfully developed a new software, Automated Image Registration and Segmentation (AIRS), to fuse the CT and SPECT images. It is designed to solve different registration and segmentation problems that arises in tomographic data sets. AIRS is addressed to obtain anatomic information to be applied to NanoSpect system which is imaging for nano‐tissues or small animals. It will be demonstrated that the information obtained by SPECT/CT is more accurate in evaluating patients/objects than that obtained from either SPECT or CT alone. The registration methods developed here are for both two‐dimensional and three‐dimensional registration. We used normalized mutual information (NMI) which is amenable for images produced by different modalities and having unclear boundaries between tissues. The segmentation components used in this software is region growing algorithms which have proven to be an effective approach for image segmentation. The implementations of region growing developed here are connected threshold and neighborhood connected. Our method is designed to perform with clinically acceptable speed, using accelerated techniques (multiresolution). |
| Databáze: |
OpenAIRE |
| Externí odkaz: |
|
