Gaussian-based Spatial Hybrid Distances for Detection and Segmentation of Lymphoid Lesions in 3D PET Images
| Autor: | Pierre Vera, Haigen Hu, Pierre Decazes, Jerome Lapuyade-Lahorgue, Su Ruan |
|---|---|
| Přispěvatelé: | Equipe Quantification en Imagerie Fonctionnelle (QuantIF-LITIS), Laboratoire d'Informatique, de Traitement de l'Information et des Systèmes (LITIS), Institut national des sciences appliquées Rouen Normandie (INSA Rouen Normandie), Institut National des Sciences Appliquées (INSA)-Normandie Université (NU)-Institut National des Sciences Appliquées (INSA)-Normandie Université (NU)-Université de Rouen Normandie (UNIROUEN), Normandie Université (NU)-Université Le Havre Normandie (ULH), Normandie Université (NU)-Institut national des sciences appliquées Rouen Normandie (INSA Rouen Normandie), Normandie Université (NU), Service de médecine nucléaire [Rouen], CRLCC Haute Normandie-Centre de Lutte Contre le Cancer Henri Becquerel Normandie Rouen (CLCC Henri Becquerel) |
| Rok vydání: | 2019 |
| Předmět: |
DBSCAN
Similarity (geometry) business.industry Computer science Gaussian Pattern recognition 3. Good health 030218 nuclear medicine & medical imaging 03 medical and health sciences symbols.namesake 0302 clinical medicine Feature (computer vision) [INFO.INFO-TI]Computer Science [cs]/Image Processing [eess.IV] symbols Gaussian function Segmentation Artificial intelligence Cluster analysis business ComputingMilieux_MISCELLANEOUS 030217 neurology & neurosurgery |
| Zdroj: | CISP-BMEI 12th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI) 12th International Congress on Image and Signal Processing, BioMedical Engineering and Informatics (CISP-BMEI), Nov 2019, Suzhou, China |
| DOI: | 10.1109/cisp-bmei48845.2019.8965932 |
| Popis: | Detection and segmentation of lymphoid lesions from PET/CT images is a crucial task for cancer staging and treatment monitoring. However, it is still a challenging task owing to the fact that there are not fixed locations, obvious feature information, shapes and sizes for lymphomas in PET images. In this work, a spatial hybrid distance (SHD) methodology is proposed for detection and segmentation of 3D PET images. Based on Gaussian kernel function, the SHDs are defined by combining different feature attributes including some priori knowledge about organ distribution, and spatial positions. A clustering method using SHD is proposed to detect and segment the lymphoma. Finally, a series of comparison experiments are performed based on DBSCAN algorithm, and the results show that the propoded SHD can achieve better performance than traditional similarity measures, and it can provide an unsupervised detection and segmentation of Iymphoid lesions in 3D PET images. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|