Highly accelerated quantitative MRI of skeletal and cardiac muscle in muscular dystrophy and type 2 diabetes

Magnetic resonance imaging (MRI) has revolutionised clinical medical imaging and research by providing repeatable, non-invasive measurements of tissue structure and function. Since its invention, scanning efficiency has improved due to scientific developments in hardware and software. Even so, MRI examinations tend to concentrate on one part of the body and usually last no longer than 1 hour. This is a problem if we want to study a disease which affects more than one body part (ie muscle and heart), since the exam will take several hours. This is (1) difficult for some patients to bear (children, adults with breathing difficulties) and (2) very expensive, which means that less research can be performed.

This aim of this project is to reduce the time required to get the images by a factor of three. The basic principle is familiar: in digital photography, photos are stored using compressed techniques such as JPEG which use the fact that large areas of the image are the same. This research takes this idea further. By knowing the properties of an image we want to acquire we can compress the acquisition of the data, not just the storage, so that the scan time is reduced by at least factor three . New mathematical techniques will be used to optimise the compression as much as possible to see if reductions greater than two-thirds are possible. The accelerated method will be validated for two important clinical conditions: for measuring skeletal and cardiac muscle damage in muscular dystrophy and measuring excess fat accumulation in the liver and pancreas in type 2 diabetes.

There are presently no established cures for the muscular dystrophies, where healthy muscle tissue is replaced by fibrosis and fat. It affects both skeletal muscle and cardiac muscle. MRI has the ability to measure the amount of fat going into the muscles, which is useful for measuring how the disease has progressed between two points in time. Cardiac MRI is known to be the most sensitive method of detecting early damage. Now that trials of therapies are moving from animal models to the first human trials, there is an urgent need to measure the response of skeletal and cardiac muscle to determine the benefits/drawbacks of the intervention. This research develops a method of significantly speeding up the acquisition of this information, to take 40 minutes rather than 2 hours.

Research Objectives: To accelerate magnetic resonance imaging (MRI) evaluation of cardiac and skeletal muscle using a novel acquisition and reconstruction technique. This will enhance the use of MRI in clinical trials and scientific studies with higher patient compliance and lower cost. This will be achieved by applying a form of mathematical reconstruction called compressed sensing which exploits the redundancy in transformed images to reduce the amount of data that needs to be acquired. We will evaluate its validity in muscular dystrophy and type 2 diabetes patients. The overall aim is to reduce what is presently a two hour examination for muscular dystrophy to take no longer than 40 minutes, a significant advantage for all patients, particularly young children. The project has two aims:

Aim 1 : To accelerate the acquisition of quantitative fat mapping in skeletal muscle of patients with muscular dystrophy and in the liver and pancreas of patients with type 2 diabetes.

Aim 2 : To accelerate the acquisition of cardiac cine imaging and tagging, a technique which sensitively measures cardiac strain and torsion and has been found to be altered in the muscular dystrophies. This will substantially reduce the acquisition time and the number of breath holds required. Respiratory muscles are often compromised and breath-holding is time consuming for this patient group.

Aim 1 will be achieved by: (1) optimising the compressed acquisition by building a computer model and testing it on our library of previously acquired data; (2) programming, testing and validating the optimum pulse sequence to acquire the compressed data on the MRI scanner, and (3) performing pilot studies in 10 healthy subjects and 10 adults with muscular dystrophy, analyzing 9 muscle groups and performing a Bland-Altman analysis of the accelerated technique to determine repeatability and bias. It will also be validated for pancreas and liver fat measurements in 10 subjects with type 2 diabetes and 10 healthy controls.

Aim 2 will be achieved by: (1) using previously acquired cardiac tagging data to optimise the compressed acquisition, (2) programming, testing and validating the optimum pulse sequence, (3) performing in vivo validation in 10 healthy volunteers and 10 adult muscular dystrophy patients, using Bland-Altman analysis.

The research will be exploited in peer-reviewed journals, international and national conferences and results presented directly to patient groups. Subsequently the acquisition and processing algorithms developed by the research will be made publicly available.