Periodic Reporting for period 2 - PASMAR (Portable, accessible and sustainable magnetic resonance)
Berichtszeitraum: 2023-03-01 bis 2024-08-31
Magnetic resonance imaging (MRI) is a key diagnostic imaging modality, used for diagnosis and treatment monitoring, with over 150 million scans performed annually. However, MRI is a highly expensive modality, with its use compared to ultrasound and X-ray, for example, limited by several factors including:
(i) purchase costs in the millions of euros,
(ii) annual maintenance costs in the hundreds of thousands of euros,
(iii) highly stringent siting requirements, typically needing a large electromagnetically-shielded room to house the system, access to high power and high fidelity electricity lines, chilled water for cooling different components, and a humidity/temperature controlled environment,
(iv) extensive training for experienced personnel
All of these considerations mean that MRI facilities are confined to centrally-located medical centres in large towns and cities. Globally over 70% of the world’s population has absolutely no access to MRI, and clinical conditions such as hydrocephalus, stroke, pneumonia, head trauma and many manifestations of viral diseases, which could benefit from even very simple scans, cannot be fully treated.
A potential solution to increasing the availability and accessibility of MRI in both the developed and developing worlds is to design new low-field systems based on permanent magnet arrays. In the developed world these could be used as screening devices, or via new geometrical designs to open up areas, such as surgical interventions, in which MRI is not used. In the developing world, these types of systems could significant increase the accessibility of MRI, particularly if they can be made portable, as well as sustainable in terms of maintenance and repair.
Advantages:
(i) The most obvious practical advantage of low field MRI is the significantly lower financial burden (system costs, maintenance costs, running costs and siting costs),
(ii) The physical size of the magnet can be small and the geometry open, which means that patient comfort is much higher In addition the acoustic noise (the most complained-about aspect of MRI) from the gradient coils is extremely low,
(iii) The power required to run the gradient and RF amplifiers can be provided by conventional mains supplies, and can even be supplied by rechargeable batteries, greatly increasing the number of potential sites for such systems, as well as portability between sites, and
(iv) Contra-indications due to medical implants are dramatically reduced: forces on any metal implant, heating around implants, and any image artifacts close to the implants are orders of magnitude lower than for conventional clinical field strengths.
Major challenges (addressed in this proposal):
(i) The major technical challenge is that lower magnetic fields result in severely reduced MR signal intensity, with for example a 50 mT system having ~400 lower signal-to-noise ratio (SNR) than a conventional 1.5 T scanner, and
(ii) to be truly operational in the field (for example rural settings or in an ambulance) issues of changes in temperature causing small changes in the system’s magnetic field, and cancellation of environmental/man-made electromagnetic noise need be addressed.
(i) purchase costs in the millions of euros,
(ii) annual maintenance costs in the hundreds of thousands of euros,
(iii) highly stringent siting requirements, typically needing a large electromagnetically-shielded room to house the system, access to high power and high fidelity electricity lines, chilled water for cooling different components, and a humidity/temperature controlled environment,
(iv) extensive training for experienced personnel
All of these considerations mean that MRI facilities are confined to centrally-located medical centres in large towns and cities. Globally over 70% of the world’s population has absolutely no access to MRI, and clinical conditions such as hydrocephalus, stroke, pneumonia, head trauma and many manifestations of viral diseases, which could benefit from even very simple scans, cannot be fully treated.
A potential solution to increasing the availability and accessibility of MRI in both the developed and developing worlds is to design new low-field systems based on permanent magnet arrays. In the developed world these could be used as screening devices, or via new geometrical designs to open up areas, such as surgical interventions, in which MRI is not used. In the developing world, these types of systems could significant increase the accessibility of MRI, particularly if they can be made portable, as well as sustainable in terms of maintenance and repair.
Advantages:
(i) The most obvious practical advantage of low field MRI is the significantly lower financial burden (system costs, maintenance costs, running costs and siting costs),
(ii) The physical size of the magnet can be small and the geometry open, which means that patient comfort is much higher In addition the acoustic noise (the most complained-about aspect of MRI) from the gradient coils is extremely low,
(iii) The power required to run the gradient and RF amplifiers can be provided by conventional mains supplies, and can even be supplied by rechargeable batteries, greatly increasing the number of potential sites for such systems, as well as portability between sites, and
(iv) Contra-indications due to medical implants are dramatically reduced: forces on any metal implant, heating around implants, and any image artifacts close to the implants are orders of magnitude lower than for conventional clinical field strengths.
Major challenges (addressed in this proposal):
(i) The major technical challenge is that lower magnetic fields result in severely reduced MR signal intensity, with for example a 50 mT system having ~400 lower signal-to-noise ratio (SNR) than a conventional 1.5 T scanner, and
(ii) to be truly operational in the field (for example rural settings or in an ambulance) issues of changes in temperature causing small changes in the system’s magnetic field, and cancellation of environmental/man-made electromagnetic noise need be addressed.
We have performed work in three major areas, namely system design, safety analysis and obtaining medical ethics approval for local scanning, and developing open source software which can be used to emulate different low field MRI systems.
In system design we have developed a fully automated pathway in which the individual components (magnet, gradient and RF coils, gradient and RF amplifiers) can be designed or chosen to perform a specific imaging task on a specific
organ. So, for example, the system might be designed to perform relatively coarse resolution scans of the human brain. This would be sufficient to study hydrocephalus and would comprise a relatively inexpensive option. Alternatively, if
much higher spatial resolution was needed, the system would have to be more sophisticated and therefore ultimately more expensive to produce. Smaller systems for extremities such as the wrist can also be designed using this approach.
We have also developed new approaches for reducing the effect of any electromagnetic interference, which would allow the system to be operated without an expensive shielded enclosure.
For scanning either healthy volunteers or patients, extensive safety analysis has to be performed since we are producing these MRI systems ourselves. We have performed and published a number of different studies, which will also be
very useful to the scientific community as a whole, which has shown that the risks associated with potential heating of the patient are negligible. In addition we showed that image artifacts from medical implants are much lower than
at clinical field strengths, and that one can obtain information right next to such implants which is currently not possible.
Since there is still very limited access to low field MRI systems, we have developed an open-source "simulator" which is python based, can be easily edited by anyone with basic programming experience, and runs on a normal laptop.
This simulator allows operators to gain knowledge about the types of image contrast which are obtainable at low fields: these are sometimes quite similar to those at conventional clinical field strengths (for example T2-weighted or
diffusion weighted images) but often are quite different. As with flight simulators, the idea is that the operator can try out many different combinations of parameters to get a good feel of what works and what doesnt. We have also
included a module where the latest machine learning based algorithms for image enhancement can be integrated into the imaging process.
In system design we have developed a fully automated pathway in which the individual components (magnet, gradient and RF coils, gradient and RF amplifiers) can be designed or chosen to perform a specific imaging task on a specific
organ. So, for example, the system might be designed to perform relatively coarse resolution scans of the human brain. This would be sufficient to study hydrocephalus and would comprise a relatively inexpensive option. Alternatively, if
much higher spatial resolution was needed, the system would have to be more sophisticated and therefore ultimately more expensive to produce. Smaller systems for extremities such as the wrist can also be designed using this approach.
We have also developed new approaches for reducing the effect of any electromagnetic interference, which would allow the system to be operated without an expensive shielded enclosure.
For scanning either healthy volunteers or patients, extensive safety analysis has to be performed since we are producing these MRI systems ourselves. We have performed and published a number of different studies, which will also be
very useful to the scientific community as a whole, which has shown that the risks associated with potential heating of the patient are negligible. In addition we showed that image artifacts from medical implants are much lower than
at clinical field strengths, and that one can obtain information right next to such implants which is currently not possible.
Since there is still very limited access to low field MRI systems, we have developed an open-source "simulator" which is python based, can be easily edited by anyone with basic programming experience, and runs on a normal laptop.
This simulator allows operators to gain knowledge about the types of image contrast which are obtainable at low fields: these are sometimes quite similar to those at conventional clinical field strengths (for example T2-weighted or
diffusion weighted images) but often are quite different. As with flight simulators, the idea is that the operator can try out many different combinations of parameters to get a good feel of what works and what doesnt. We have also
included a module where the latest machine learning based algorithms for image enhancement can be integrated into the imaging process.
Our work currently leads the way in terms of system design for neuroimaging applications of very low field MRI. We are also working with commercial low field companies to ensure that this technology is
certified, and therefore can be more widely used in patient studies.
In the remaining part of the project, we hope to make significant progress in exploring the design of new types of magnet for new types of clinical applications. For example, we will look at rotationally optimized designs, which
means that the position of each magnet is optimized to produce overall the most uniform magnetic field, in order to make the neuroimaging systems more compact. We will also look at more unconventional designs, ones which
are open on the upper side to increase patient comfort and acceptance, and nested magnets which can be placed inside one another so that the field strength can be tailored to the particular application without having
to replicate the system for each application.
A second area which we will actively explore is the use of artificial intelligence to maximize the amount of information that we can obtain for a given overall scan time. This will involve a combination of
optimizing data acquisition protocols and image processing algorithms. This will be particularly important in applications such as pediatric imaging, where the patient is not able to remain still for long periods of time.
certified, and therefore can be more widely used in patient studies.
In the remaining part of the project, we hope to make significant progress in exploring the design of new types of magnet for new types of clinical applications. For example, we will look at rotationally optimized designs, which
means that the position of each magnet is optimized to produce overall the most uniform magnetic field, in order to make the neuroimaging systems more compact. We will also look at more unconventional designs, ones which
are open on the upper side to increase patient comfort and acceptance, and nested magnets which can be placed inside one another so that the field strength can be tailored to the particular application without having
to replicate the system for each application.
A second area which we will actively explore is the use of artificial intelligence to maximize the amount of information that we can obtain for a given overall scan time. This will involve a combination of
optimizing data acquisition protocols and image processing algorithms. This will be particularly important in applications such as pediatric imaging, where the patient is not able to remain still for long periods of time.