Detecting mammography of breast microcalcification with SOL-based self-organization neural network
Chien-Pen Chuang, Shiunn-Shin Lee, Jia-Shiunn Tsai, Tai-Jung Kuo · 2010 Sixth International Conference on Natural Computation · 2010
The progress of medical imaging technologies, from X-ray radiography, ultrasonic graph to modern age's Magnetic Resonance Imaging (MRI) and Computed Tomography (CT) scan has helped the advance of the medical technology as well as the improvement of medical care quality all over the world. It is essential to promote our own medical imaging technologies so as to reduce the future overall medical expense. In this study, we applied the numerical method of SOL computer-aided system. The SOL includes self-organizing mapping (SOM) method and Learning Vector Quantization (LVQ) on artificial neural network for feature extraction, clustering and filtering, to improve the detection resolution and to upgrade the target recognition efficiency in clinical decision support systems. We also developed an image processing toolbox as our auxiliary instrument. All newly developed numerical methods and functions of SOL (including SOM and LVQ) can be easily retrieved for data analysis. We also analyze the complexity of our algorithm and calculate the processing speed for our simulation results. Based on our analysis, we proposed how to improve the efficiency in the computer aided clinical decision system.