Diagnosis of breast cancer based on modern mammography using hybrid transfer learning

Autor: Deepak Gupta, Dang N. H. Thanh, Aditya Khamparia, Ashish Khanna, Thai Kim Phung, Subrato Bharati, Prajoy Podder
Rok vydání: 2020
Předmět:
FOS: Computer and information sciences
Computer Science - Machine Learning
3D-Mammography
Computer science
Computer Vision and Pattern Recognition (cs.CV)
3D mammography
Computer Science - Computer Vision and Pattern Recognition
Early detection
02 engineering and technology
Hybrid transfer learning
Machine learning
computer.software_genre
Quantitative Biology - Quantitative Methods
Convolutional neural network
Article
Machine Learning (cs.LG)
Breast cancer
Artificial Intelligence
FOS: Electrical engineering
electronic engineering
information engineering
0202 electrical engineering
electronic engineering
information engineering
medicine
Mammography
Survival rate
Quantitative Methods (q-bio.QM)
medicine.diagnostic_test
business.industry
Applied Mathematics
Image and Video Processing (eess.IV)
Cancer
020206 networking & telecommunications
Electrical Engineering and Systems Science - Image and Video Processing
Medical image segmentation
medicine.disease
Computer Science Applications
Hardware and Architecture
FOS: Biological sciences
Signal Processing
020201 artificial intelligence & image processing
Convolutional neural networks
Artificial intelligence
business
Transfer of learning
computer
Software
Information Systems
Zdroj: Multidimensional Systems and Signal Processing
ISSN: 0923-6082
Popis: Breast cancer is a common cancer for women. Early detection of breast cancer can considerably increase the survival rate of women. This paper mainly focuses on transfer learning process to detect breast cancer. Modified VGG (MVGG), residual network, mobile network is proposed and implemented in this paper. DDSM dataset is used in this paper. Experimental results show that our proposed hybrid transfers learning model (Fusion of MVGG16 and ImageNet) provides an accuracy of 88.3% where the number of epoch is 15. On the other hand, only modified VGG 16 architecture (MVGG 16) provides an accuracy 80.8% and MobileNet provides an accuracy of 77.2%. So, it is clearly stated that the proposed hybrid pre-trained network outperforms well compared to single architecture. This architecture can be considered as an effective tool for the radiologists in order to reduce the false negative and false positive rate. Therefore, the efficiency of mammography analysis will be improved.
24 pages, 11 figures
Databáze: OpenAIRE
Externí odkaz: