Portable brain scanner allows PET in motion

For original article site, click here.

Clinical brain imagers currently require a trade-off between motion and resolution. Large, bolted-to-the-floor devices such as MRI, PET and magnetoencephalography systems provide high-resolution images but require the participant to remain completely still. Meanwhile, approaches such as electroencephalography and near-infrared spectroscopy can be used with a moving subject, but have low spatial resolution and cannot visualize important structures such as the hippocampus and amygdala deep within the brain.

To address this dilemma, a US research team is developing a wearable brain-scanning device that will provide high-resolution images of the whole brain, including deep structures, while the subject is moving about and performing everyday activities. The new approach – called ambulatory microdose positron emission tomography (AMPET) – miniaturizes the PET technology such that it fits onto a helmet that's worn during scanning.

"Standard PET scanners in hospitals utilize large photomultiplier tubes; but recent advances in material science have led to the development of detectors using silicon photomultiplier (SiPM) technology, which is orders of magnitude more compact," explained Julie Brefczynski-Lewis, from the West Virginia University (WVU) Centers for Neuroscience. "We made a ring of SiPM detectors around the head and are developing this into a wearable PET scanner."

Stan Majewski, an original co-principal investigator on the project, explained that the team's motivation for developing the portable PET brain scanner lay both in the challenge and the assumption that improved instruments will offer new opportunities. "As a developer of nuclear medicine organ-specific imagers, and coming originally from the high-energy physics community, I have seen what "my brothers and sisters" from the Brookhaven National Laboratory developed for rats – wearable Rat-Cap PET imagers," said Majewski, currently at the University of Virginia. "Then came the development in Japan of the so-called PET-Hat. The patient was in the upright position, but had to sit. My vision was to free the patient/subject and make him/her stand or even walk within a limited space."

Initial simulations of the device showed that the helmet scanner offers a more than 400% increase in sensitivity over a conventional whole-body PET scanner. AMPET also requires a much lower radiation dose than conventional scanners. "The detectors are close to the head and thus capture more photons," said Brefczynski-Lewis, who presented the team's work at the recent Neuroscience 2015 annual meeting.

"We have already demonstrated that we can reconstruct images from this close geometry using 10-25% of the dose used for standard PET," she added. "Low dose has advantages both in reducing radioactivity exposure – making longitudinal scans, purely research scans and scans using young populations more feasible – as well as allowing detection of lower concentration targets like rarer neuron types and ligands with a low binding efficiency."

The researchers predict that further improvements in time-of-flight and depth-of-interaction information could reduce the required radioligand dose to less than one tenth of the standard dose. OAS AD 'Middle' begin

Applications abound

While still in the prototype stage, AMPET has a wide variety of potential research and clinical applications including studies of balance, physical therapies, natural social interactions and virtual reality, as well as disorders like stroke, Alzheimer's and Parkinson's disease, multiple sclerosis and traumatic brain injury.

"Imagine imaging a savant while painting or a chess master in action: We might be able to tap into the mechanisms behind these super abilities," said Brefczynski-Lewis. "Clinically, we may better understand why people with autism react differently to social situations and this may inform diagnosis and therapies. Moving forward, we'll be talking with researchers and clinicians working in stroke rehabilitation, Parkinson's disease, and balance disorders, as well as neuroscientists who study social, cognitive, and emotional processes, to determine how such a scanner could be used and what features are most necessary."

"I was excited to discover that the system can potentially assist with early diagnosis of Alzheimer's by combination with Virtual Reality," added Majewski. "Apparently, the earliest signs of Alzheimer's are not seen in the memory loss but in the loss of navigational skills. Well-controlled navigational tests in the Virtual Reality environment, correlated with functional/molecular images of the brain obtained at low radiation dose with our upright high-resolution PET imager, can be really revolutionary."

Last year, the WVU team and collaborators – from the University of Virginia, GE Global Research, UC Davis and the University of Washington – were awarded $1.5 million from the USA's National Institutes of Health (NIH). The funding was through the newly established BRAIN Initiative, established by President Barack Obama to accelerate the development and application of innovative imaging technologies.

"Our proposed imager will be like no technology currently existing, in that one can have high resolution, high sensitivity whole brain images while the participant is actively moving," Brefczynski-Lewis told medicalphysicsweb. "The untapped niche is quite remarkable and seemed to be just what the NIH was looking for."

The team is currently testing detector modules to determine the ideal coupling of detectors and electronics – information that will allow them to better understand the weight/resolution trade-offs and optimize prototype design.

"In tandem, we are also building an improved ring prototype," said Brefczynski-Lewis. "This won't have all the latest advances, but will help us understand how motion will affect the images, and better see real world issues that must be addressed in the mechanical support of the device. Results from both of these sets of tests will allow us to improve simulations and reconstruction algorithms."

To help potential users find out more about AMPET, the researchers have set up www.pethelmet.org, where visitors can compare the advantages and trade-offs of three distinct AMPET models and fill out a short survey informing which features would be important to their research or clinical practice. "We need input from neuroscientists and clinicians as to what kind of imager we should build," said Brefczynski-Lewis.