Neural Network Design - Report Example

The neural networks aid in the relationship of parameters of the tumor with the profile of temperature over the breast model. There is a comparison of the obtained related to the one obtained by an element software (Bradbury 2007). There is a possibility of determining the parameters of cancer with a five and ten percent random noise. With ten percent noise, the estimation accuracy deteriorates for tumors that are deep-seated. INTRODUCTION Breast cancer in women is commonly making it an international concern.

Every year about two hundred thousand new cases are reported. There is an estimation that more than one million women are not aware that they have cancer (Bradbury 2007). Infrared imaging is one of the conventional ways that were used in the detection of cancer. Abnormal skin temperatures may be used as an indication of benign tumors and cancer. The objective of the neural networks is to develop a methodology to be used in the estimation of different parameters of the cancerous tumor (Kings 1996).

The report relates to breast cancer and the different neural networks used in the methodology estimation. The demonstration of the methodology is done by the use of a simple model, in the illustration of the details and the involved procedures. Various conditions of the cancerous tumor were used to illustrate the neural network generality (Millburn 2000). The diagnosis of breast cancer is a significant medical problem in the real world. An important class of medical science problems involves disease diagnosis.

This is based on various performed tests upon the patient (Rockville 2008). When there is an involvement of several tests, the final diagnosis may be obtained in a difficult manner, even for experts in the medical field. This has given an increase, over a few most recent decades, to diagnostic computerized tools (Hill 2012). This is intended to help the physician in reaching a sensible conclusion out of data confusion. Much research has been done on the medical diagnosis of cancer. A learning algorithm combines annealing logarithmic simulated with a perceptron algorithm and accuracy of almost one hundred percent is reported (Hagan 1996).

Breast cancer is the commonest cancer type in women. Men can also get breast cancer, but this accounts for a much lesser percentage of all diagnosed cases of breast cancer (Hill 2012).