Samenvatting
About 30 years ago the first commercial Magnetic Resonance Imaging (MRI) scanner was installed at the Hammersmith Hospital in London. This revolutionary technique made it possible to image tissues surrounded by bone. This was a big advantage in comparison to X-ray based imaging methods. However, resolution of the first magnetic resonance images was low and the scanning time was long, due to problems of weak signal and high sensitivity to the patient motion. Since then a lot of research has been done to improve the overall performance of the machines. In mid 90s fast imaging techniques were developed that had a tremendous impact on the popularity of MRI among other medical imaging methods. Nowadays there are a lot of clinical imaging applications where MRI overtakes the X-ray successors. Moreover, MRI is believed to be harmless to the patient, because no ionizing radiation is utilized. However, the main disadvantages of MRI are strong magnetic field, extreme expense, and relatively long examination time when compared to X-ray. The first factor imposes high safety standards that must be respected in an MRI scanner room, whereas the last two factors prevent hospitals from fast investments return. Moreover, due to high demand on MRI examinations, the patient waiting lists in hospitals are often several weeks long. This backlog decreases patient satisfaction. In this dissertation, a new approach to reduce the examination time of MRI systems is described. The time reduction is accomplished by dividing parts of the MRI examination into segments that are then intermixed. The intermixing algorithms are based on scheduling technique from the field of Operations Research. There are a number of physical parameters that restrict performance of MRI systems, such as temperature of MRI hardware during the examination. Also, due to electromagnetic effects inside the bore of MRI scanner, the temperature of patient’s body can get close to an uncomfortable level. In current practice, all these duty cycle limitations are modeled and verified before the MRI examination starts. Then, if necessary, the MRI examination time is prolongated, in order not to exceed the temperature limits. Typical MRI examination consists of several discrete parts, i.e., scans. Different types of scans impose different duty cycle limitations. The approach proposes that the examination can be divided into small segments that are rescheduled in such a way that the adverse effects of duty cycle limited scans are reduced by non-limited scans. In this thesis, several scheduling algorithms are described that were designed to deal with different kinds of duty cycle limitations and to improve performance of MRI systems. The algorithms were verified on a large number of MRI examinations. According to collected statistics, time of MRI examinations can be reduced by up to 22%. As a result, the capacity of one MRI system can be increased by up to 4 patients per day. Moreover, special MRI experiments were carried out to validate the algorithms. Finally, the thesis presents an approach to patient flow modeling in MRI departments in hospitals. The patient flow is modeled by means of queuing theory in
order to uncover bottlenecks. Then, discrete-event computer simulations are performed
to overcome limitations of the classical queuing theory assumptions. The current hospitals practice demonstrates that the MRI scanners are not always the bottleneck in the overall examinations workflow. The resulting models can be utilized to predict patient flow for various layouts of MRI departments and appointment scheduling strategies. Based on these detailed models, recommendations on improving MRI departments’ workflow can be derived. The results of this study can be used to optimize performance of MRI departments in hospitals or free-standing imaging centers. First, the MRI scanning time can be reduced. Second, the patient flow can be optimized that yields the overall MRI examination time reduction. This will result in better patient comfort and
faster return on investments in MRI equipment. The research described in this thesis was carried out as a part of the DARWIN project at Philips Healthcare under the responsibilities of the Embedded Systems Institute (ESI). This project is partially supported by the Dutch Ministry of Economic Affairs under the BSIK program.
Originele taal-2 | Engels |
---|---|
Kwalificatie | Doctor in de Filosofie |
Toekennende instantie |
|
Begeleider(s)/adviseur |
|
Datum van toekenning | 27 mrt. 2012 |
Plaats van publicatie | Eindhoven |
Uitgever | |
Gedrukte ISBN's | 978-90-386-3112-7 |
DOI's | |
Status | Gepubliceerd - 2012 |