Microwave hyperthermia, a process that involves heating tumour cells, has been proposed as an alternative method of treating breast cancer. By sending microwave energy of a suitable frequency range to the breast, localised heating is induced by the applied electromagnetic radiation. It is important that the tumour is heated to a temperature greater than 42℃, (the required temperature for hyperthermia treatment) while keeping healthy tissue at normal body temperature. Several useful approaches have been proposed to direct microwave hyperthermia to cancerous tissue, including novel focusing methods and innovative microwave hardware designs. However, a number of challenges restricting the use of microwave systems in clinical medicine remain. Arguably, the most persisting engineering problem to overcome is achieving locally focused microwave power in the heterogeneous environment of the breast. In order to understand the challenges of microwave hyperthermia, considerably large computational efforts are required to solve 3D Electromagnetic (EM) problems in a complex model of breast tissue.The success of using microwave hyperthermia to treat breast cancer is highly dependent on the accuracy of the excitation signals utilised for each antenna element. Several research papers have reported various approaches on this topic. Although recent progress has been made in the investigation of microwave hyperthermia, the studies were mainly reliant upon oversimplified Radio Frequency (RF) models which assume point source applicators, and human breast phantoms which are mostly 2D models lacking tissue thermal-electric properties. As a result, it is infeasible to implement the proposed systems for clinical applications. This thesis aims to address the limitations of these earlier studies by modelling and constructing realistic human breast models and 3D antenna arrays. This will require the determination of correct phases and amplitudes for the excitation signals in a realistic environment by using a global optimisation method.The approaches proposed use actual antenna arrays to transmit microwave power while the excitation signals are optimised based on power distributions and thermal profiles induced in the patient-specific breast models. The first study based on 2D antenna arrays and patient-specific breast models is demonstrated in Chapter 3. In this study, microwave hyperthermia is performed in a realistic environment rather than over-simplified scenarios. In the recommended technique, electromagnetic focusing on patient-specific breast models concentrates the power at the tumour position while successfully keeping the power levels at other positions (healthy tissue) at minimum values. Several patient-specific breast models ranging from fatty tissue to highly dense tissue are used to investigate the effect of breast anatomy on the proposed focusing technique.ii Chapter 4 introduces an improved hyperthermia approach for breast cancer treatment, overcoming the limitations of previously analysed techniques. In this appro...