Title | ||
---|---|---|
Improve threshold segmentation using features extraction to automatic lung delimitation. |
Abstract | ||
---|---|---|
With the consolidation of PACS and RIS systems, the development of algorithms for tissue segmentation and diseases detection have intensely evolved in recent years. These algorithms have advanced to improve its accuracy and specificity, however, there is still some way until these algorithms achieved satisfactory error rates and reduced processing time to be used in daily diagnosis. The objective of this study is to propose a algorithm for lung segmentation in x-ray computed tomography images using features extraction, as Centroid and orientation measures, to improve the basic threshold segmentation. As result we found a accuracy of 85.5%. |
Year | DOI | Venue |
---|---|---|
2013 | 10.3233/978-1-61499-289-9-1159 | Studies in Health Technology and Informatics |
Keywords | Field | DocType |
Algorithm,Lung,Segmentation | Computer vision,Scale-space segmentation,Segmentation,Artificial intelligence,Medicine | Conference |
Volume | ISSN | Citations |
192 | 0926-9630 | 0 |
PageRank | References | Authors |
0.34 | 0 | 6 |
Name | Order | Citations | PageRank |
---|---|---|---|
Cleunio França | 1 | 0 | 0.68 |
Germano C. Vasconcelos | 2 | 104 | 12.25 |
Paula Diniz | 3 | 1 | 2.37 |
Pedro Melo | 4 | 29 | 4.61 |
Jéssica Diniz | 5 | 0 | 0.68 |
Magdala Novaes | 6 | 5 | 4.23 |