Search

Patient safety during Magnetic Resonance Imaging (MRI) examinations is currently guaranteed through the compliance with the limitations on the Specific Absorption Rate (SAR) exposure provided by the current regulations. SAR limits are implemented by scanner vendors in a strictly conservative way to nullify the risk of patient hazard, even at the cost of limiting sometimes the diagnostic power of the exam. Nonetheless, both in the case of ultra-high field applications (i.e. with static magnetic field of 7 T and above) and at clinical field strengths for example in the presence of metallic prosthetic implants, the inhomogeneities in the RF excitation field distribution may cause a local and subject-specific SAR increase. In those cases, local tissue temperatures could reach damaging levels. To directly quantify the local temperature increase caused by the more commonly used RF acquisition sequences in MRI of the human extremities, we have implemented Magnetic Resonance Thermometry (MRT) techniques on a 7 T MRI scanner. We demonstrated that temperature increase maps are easily and fast obtainable with MRT techniques, which are sensitive enough to detect potential hazardous tissue heating.

Sodium Magnetic Resonance Imaging (MRI) is an attractive imaging technique for early detection of osteoarthritis, as sodium concentrations correlate with proteoglycan content. However sodium yields much less signal in human tissues compared to hydrogen: it has a very rapid transverse signal decay (T2) and rapid longitudinal relaxation (T1), low gyromagnetic ratio and much smaller concentration than hydrogen. Ultra High Field (UHF) 7T MRI systems play an essential role to enhance the SNR, to achieve sufficient spatial resolution and to obtain images with diagnostically relevant contrast. This work shows how to design, develop and test a dual-tuned proton/sodium surface radio-frequency coil for 7T UHF MRI systems.

A high performance implementation of an algorithm based on the learning process of the human vision as an edge-enhancing filter for medical images is proposed. It is based on a cognitive image processing algorithm that emulates the contour identification process that is believed to be used by the human brain. The algorithm is implemented on an innovative and high performance embedded system for real-time pattern matching. The system design uses Field Programmable Gate Arrays (FPGAs) and the powerful Associative Memory chip (an ASIC) to achieve real-time performance. We present the first results of the algorithm as a 3-D edge detection filter using Magnetic Resonance Images (MRI).

This work reports a study of Neuromuscular Diseases (NMD) by Magnetic Resonance Imaging as reliable and non invasive instrument for NMD diagnosis and follow up. The evaluation of the images is now only visual, while standardization procedures and quantitative methods could be very useful instruments to optimize the diagnostic performances. We propose a new method to evaluate the fat infiltration in tissues developed and retrospectively applied to images of the human leg. Through a muscle segmentation algorithm on structural T1-weighted magnetic resonance images (MRIs), the estimated non-muscle percentage (eNMP) in the segmented muscle area has been evaluated in healthy subjects as a reference value. A semi-automated procedure allows extending the algorithm to MRIs of NMD patients. A strong correlation has been demonstrated between this index and the disease severity. The final aim is to obtain a quantitative evaluation of fat infiltration percentage and to relate it to the grade of muscle impairment in subjects affected by Neuromuscular Diseases.

Ultra-High Field (UHF) (4–9.4T) Magnetic Resonance Imaging (MRI) and Spectroscopy (MRS) are valuable tools in the diagnosis and monitoring of many diseases thanks to the enhanced Signal-to-Noise Ratio (SNR) and spectral/spatial resolution. However, such UHF MR applications require the development and optimization of specially designed Radio Frequency (RF) coils. In this study we report the design, construction, workbench and MR testing of a novel 7T double-tuned (1H/31P) RF surface coil made with microstrip technology. The RF coil is suitable for MRI and MRS human studies of the lower limbs and also of the human head. The double-tuned RF coil was tested at the IMAGO7 Foundation in Pisa (Italy). Ex vivo MR 1H images and 31P spectra obtained with a large size fresh calf veal sample showed an excellent SNR within about 50 mm from the RF surface coil. It is believed that the combined 7T MR 1H/31P will provide benefits in performing clinical follow-up of muscular disorders in young and aging patients. The proposed design can be easily extended to the simultaneous detection of the 1H and other low-gamma nuclei, such as 7Li, 11B, 13C, 17O, 23Na.

The hippocampus segmentation in Magnetic Resonance (MRI) scans is a relevant issue for the diagnosis of many pathologies.

The evaluation of the local Specific Absorption Rate (local SAR) is a major concern in high field Magnetic Resonance (MR) systems. The investigation here presented is focused on the estimation of the Specific Absorption Rate in human head excited by a 7T surface coil. The aim of our work is to evaluate the maximum local SAR of the sample inside the scanner under realistic condition of use. The proposed methodology can be adopted to avoid the over-exposure of the body during the MR exam.

Several clinical magnetic resonance applications at 1.5 T and higher field strengths require a careful selection of the RadioFrequency (RF) coil design to optimize the RF spatial distribution and sensitivity. Specifically, two basic requirements must be fulfilled for obtaining high quality images: in the transmission mode, RF coils must be able to produce a uniform magnetic field in the volume of interest so that the nuclei can be properly excited; in the receiving mode, a high signal to noise ratio is needed, and the coil must be able to collect the signal emitted by the nuclei with better sensitivity throughout the volume of interest. A number of analytical and numerical methods are reported in the literature to simulate the RF field distributions of surface and volume coils. In this work, we compared the performances of two computational methods (Method of Moments and Finite Elements Method), considering the modelling of both surface (single loop, figure of eight coil - FO8, dual loop) and volume (TEM) coils. Low (1.5 T) and high (7 T) field operating regimes have been analysed. Measurements obtained on the workbench and with a 1.5 T scanner are used for validation.

A computer-aided detection (CADe) system for microcalcification cluster identification in mammograms has been developed in the framework of the EU-founded MammoGrid project. The CADe software is mainly based on wavelet transforms and artificial neural networks. It is able to identify microcalcifications in different datasets of mammograms (i.e. acquired with different machines and settings, digitized with different pitch and bit depth or direct digital ones). The CADe can be remotely run from GRID-connected acquisition and annotation stations, supporting clinicians from geographically distant locations in the interpretation of mammographic data. We report and discuss the system performances on different datasets of mammograms and the status of the GRID-enabled CADe analysis., 6 pages, 4 figures; Proceedings of the IEEE NNS and MIC Conference, October 23-29, 2005, Puerto Rico