Abstract | ||
---|---|---|
Pulmonary nodule classification has aroused much interest for a long time, which can be an effective method for early diagnosis of lung cancer. Computer-aided diagnosis (CAD) systems are helpful to realize totally automatic classification for pulmonary nodule, based on computed tomography (CT) scans. It is important to emphasize that even we are easy to access sufficient CT scans, there are always very few annotations related to pulmonary nodules and the annotation process is a huge burden for radiologists. We propose a novel unsupervised representation learning method integrated with convolutional architecture to face the challenge and obtain promising results on the the National Cancer Institute (NCI) Lung Image Database Consortium (LIDC) dataset. |
Year | DOI | Venue |
---|---|---|
2017 | 10.1109/ISCID.2017.34 | 2017 10th International Symposium on Computational Intelligence and Design (ISCID) |
Keywords | Field | DocType |
CAD system,unsupervised representation learning,ELM | Lung cancer,CAD,Annotation,Pattern recognition,Computer science,Feature extraction,Artificial intelligence,Computed tomography,Image database,Cancer,Machine learning,Feature learning | Conference |
Volume | ISSN | ISBN |
1 | 2165-1701 | 978-1-5386-3676-3 |
Citations | PageRank | References |
0 | 0.34 | 4 |
Authors | ||
4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Xinyu Jin | 1 | 93 | 16.67 |
Fenghao Zhu | 2 | 0 | 0.34 |
Lanjuan Li | 3 | 10 | 1.74 |
Qi Xia | 4 | 132 | 21.76 |