Modeling and Analysis of the interaction between magnetic fields and biological tissues in MRI

Magnetic Resonance (MR) is a well known diagnostic technique today widely used especially for cardiac and neurological applications. MR has an excellent spatial resolution and a good temporal resolution and it allows obtaining high quality clinical images often essential to diagnosis or monitoring of several diseases. Moreover, MR is considered a safe technology since it has just the ability to change the position of atoms, but not to alter their structure, composition, and properties, as the ionizing radiations do attempts. In fact, the electromagnetic radiations involved in a MR procedure have no enough energy to detach electrons from atoms or molecules, such as other higher energetic radiations can do (X-rays, radiations used in nuclear medicine etc). However, as in any sanitary interventions, even in a MR diagnostic procedure there are intrinsic hazards that must be understood, acknowledged and taken into consideration: for this reason, the analysis of the interaction between the magnetic fields an the biological tissues underwent to a MR procedure, is essential. These interactions are caused from different physical phenomena which, on the hand, are responsible to the signal generation that contribute to the final image but, on the other hand, can cause dangerous biological effects for the patient or signal artefacts. The study of these interactions has been become more important in the last years due to the growing interest for high static magnetic field MR scanners which assure a higher signal to noise ratio (SNR), and hence a better quality for the final images, but imply heavier risks for patients and occupational workers. Although the radiation used is not ionizing, there are several effects to consider for safety assurance and engineering aspects relative to MR signal and image generation. The knowledge of these phenomena is important not only for the design of transmission/reception coils and acquisition sequences but also for the choice of acquisition parameters for each diagnostic exam. This thesis deals to the study of the interactions between the MR magnetic field and the biological tissues and, in particular, to the modeling and analysis of the adsorbed power estimation in a subject underwent a MRI exam. The goal of this work is the detailed understanding of the interaction mechanisms and the design of numerical models and experimental methods for the adsorbed power estimation using the evaluation of Specific Absorption Rate (SAR) (radiofrequency (RF) dosimetry). Regarding the numerical calculation of SAR, the Finite Difference Time Domain (FDTD) method was used; a birdcage coil model was designed for RF transmission and a human thorax model was designed as biological sample. A relationship between local peak SAR to average SAR ratio and sample body metric was found. A new experimental method for RF induced heating measurement on biomedical implants was designed based on a laboratory set up to reproduce the RF transmission of a MR exam. The method was tested on a home made phantom with a commercial cardiovascular stent. Then two methods to obtain the local SAR distribution using MR images were tested and compared. Finally, a study for possible genotoxic effects of MR procedure was conducted both in vitro and in vivo sample using the micronuclei (MN) test of DNA damage. For this study, a new method to measure electric conductivity of liquid sample at radiofrequency used in MR applications was also realized. This study has been the first, and up to now the only one, demonstrating any genotoxic effects induced by MRI. Although, the results, showing a modest increased MN frequency and a return to normal frequency within a few days, cannot confirm a conclusion of risk, a prudent attention should be adopted in order to avoid unnecessary MR examinations, according to the precautionary principle