Electrical impedance tomography (EIT) is a medical imaging technology that uses boundary electrodes to inject stimulus currents and measure the resulting potential distributions. These potentials are measured using electrodes which are in turn used to reconstruct conductivity changes within the body. EIT has been studied for its ability to image both the flow of blood and the delivery of blood to a tissue (perfusion). However, cardiac-related signals are challenging to image accurately due to their small amplitude and the limited sensitivity of EIT systems to impedance changes deep in the body. This thesis presents techniques to improve perfusion imaging with EIT by generating more accurate meshes and using internal electrodes to obtain higher sensitivity in the centre of the body. This work develops a tool to generate accurate, customized meshes from diagnostic computed tomography (CT) images, and a technique to reconstruct images with internal electrodes. Custom models reconstructed the location of impedance changes due to ventilation with higher accuracy, and internal electrodes yielded an increase in internal sensitivity over traditional external configurations. Shifting the position of electrodes on an internal probe by as little as 1% of the tank radius in simulation created artefacts in images reconstructed using existing approaches without motion correction. A novel technique to correct for probe motion is presented that improved reconstruction accuracy and reduced background noise compared to existing techniques. The presented work contributes to increasing internal sensitivity of EIT measurements and demonstrates that refined meshes and internal electrodes may improve measures of perfusion and help to make EIT a viable tool for continuous perfusion monitoring at the bedside.