Development of breast cancer diagnosis tool using hybrid magnetoacoustic method and artificial neural network

• Main Authors: Mohamad Salim, Maheza Irna, Ahmad, Abd. Hamid, Ariffin, Ismail, Rosidi, Bustanur, Supriyanto, Eko
• Format: Article
• Published: North Atlantic University Union 2012
• Subjects: RA0421 Public health. Hygiene. Preventive Medicine; RC0254 Neoplasms. Tumors. Oncology (including Cancer)
• Online Access: http://eprints.utm.my/id/eprint/31697/; http://www.naun.org/multimedia/NAUN/bio/17-664.pdf

Breast cancer is a metabolic disease that causes the breast cells to acquire genetic alteration and allows them to grow beyond the normal tissue limit. With the yearly increasing trend in new cases and mortality rate, new approach in diagnosis and treatment of breast cancer is crucial to improve the existing management of breast cancer cases. This paper presents a new approach in breast cancer diagnosis by using Hybrid Magnetoacoustic Method (HMM) and artificial neural network. HMM is a newly developed one dimensional imaging system that combines the theory of acoustic and magnetism for breast imaging. It is capable to produce 2 outputs, the attenuation scale of ultrasound and the magnetoacoustic voltage. In this study, an artificial neural network is developed to automate the output of HMM for breast cancer classification. The ANN employs the steepest gradient descent with momentum back propagation algorithm with logsig and purelin transfer function. The best ANN architecture of 3-2-1 (3 network inputs, 2 neurons in the hidden layer, one network output) with learning rate of 0.3, iteration rate of 20000 and momentum constant of 0.3 was successfully developed with accuracy of 90.94% to testing data and 90% to validation data. The result shows the advantages of HMM outputs in providing a combination of bioelectric and acoustic information of tissue for a better breast cancer diagnosis consideration. The system’s high percentage of accuracy shows that the output of HMM is very useful in assisting diagnosis. This additional capability is hoped to improve the existing breast oncology diagnosis.