Two novel and innovative segmentation algorithms are presented, for liver metastases and necrotic tissues in pre- and post-ablation, respectively. Both are based on Gaussian Mixture Model (GMM), adapted to include healthy liver knowledge and proximity information. Preliminary validation results on 20 patients and 42 lesions are discussed showing high accuracy versus manual segmentation gold standard outperforming simpler GMM and thresholding methods.
Pre and Post Liver Lesion Thermal ablation FDG-PET: background driven GMM segmentation
MOCCIA, SARA;SOFFIENTINI, CHIARA DOLORES;BASELLI, GIUSEPPE;
2015-01-01
Abstract
Two novel and innovative segmentation algorithms are presented, for liver metastases and necrotic tissues in pre- and post-ablation, respectively. Both are based on Gaussian Mixture Model (GMM), adapted to include healthy liver knowledge and proximity information. Preliminary validation results on 20 patients and 42 lesions are discussed showing high accuracy versus manual segmentation gold standard outperforming simpler GMM and thresholding methods.File in questo prodotto:
| File | Dimensione | Formato | |
|---|---|---|---|
|
embc.pdf
accesso aperto
:
Publisher’s version
Dimensione
92.47 kB
Formato
Adobe PDF
|
92.47 kB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
