Design of processes supporting the development of medical devices
|Title||Design of processes supporting the development of medical devices|
|Department||Applied Informatics and Applied Mathematics Doctoral School|
Hemodialysis machines are responsible for removing metabolic waste products from the human blood when the body cannot excrete them on its own. With their help it is possible not only to improve the life quality of patients with kidney malfunction, but in certain cases this is the only way to keep someone alive.
The software component in these machines has significantly improved with the accessibility of powerful computing devices. Not only the expectations of experts has changed (more accurate device, improved patient safety, cost efficient operation), but also the international regulations have been modified. Altogether, these facts raise a challenge for development teams. In this thesis I am analyzing two aspects of the aforementioned problems.
During treatment, the blood of the patient is extracted and filtered extracorporeally. The purification is done by using special dialysis fluids where at the end, excess fluid is removed. This means that the fluid transfers have to be supervised to keep the patient fluid balance. Soft computing methods were developed and compared with conventional control methods in the first thesis group. This approach is novel for the industry and analyzes have proved it to be effective. The developed controllers are comparable with the existing solution with the advantage of possible incorporation of additional expert knowledge. The controllers were developed to be not only suitable for testing and analyzes, but they can be easily integrated into existing machines.
The quality of a newly developed machine is highly depending on the applied development process. Different supporting tools are used to decrease documentation related burden, eliminate redundancies and to reduce human workload. The tools in the development toolchain are called application lifecycle management system and they can be homogeneous or heterogeneous depending on the number and connectivity of software providers. The second thesis group discusses how enhanced traceability and consistency can be achieved when using these systems. The idea of augmented lifecycle space was created as a software independent solution to find missing artifacts in software development. Furthermore, it makes possible for stakeholders to select and prioritize the deficiencies.