Radiology meets Biology: New imaging technology for the mouse

Simon Cherry
University of California at Davis
Bioengineering and Molecular Imaging

Animal models are widely used to study human diseases and potential new
therapies. Many of these models are based around the mouse, because of its
genetic similarity to man (roughly 95% of the genes are common), the many
sophisticated techniques that now exist for genetic manipulation in mice,
their fast breeding cycle, and their low housing costs. The rat is a
commonly used model in neuroscience, because of its larger brain size
compared to the mouse. Most existing techniques designed to assess
biological effects in these animals require that the animal be sacrificed.
This precludes repeat studies within a single animal and leads to the need
to study groups of experimental animals at each time point of interest. One
of the major breakthroughs over the past five years has been the adaptation
of imaging technologies that are familiar in the clinical setting, for use
in imaging of the mouse and rat. CT, MRI, PET, SPECT and ultrasound systems
have been developed specifically for small animal imaging, and the results
have been spectacular. These in vivo imaging technologies now allow
anatomy, physiology, metabolism and specific molecular targets to be studied
repeatedly and non-invasively in living animals.

This presentation will provide an overview of small animal imaging, and the
imaging technologies that are currently available. To demonstrate some of
the opportunities and challenges in the world of mathematics and statistics,
the specific case of imaging with positron emission tomography (PET) will be
explored in depth. The many factors inherent in the design of PET systems
(both for animals and humans) that determine the quality of the
reconstructed tomographic image will be explained. This will lay the
groundwork for Dr. Richard Leahy's presentation which will review how all
these factors, along with an appropriate statistical model of the data, can
be placed in a unifying framework for reconstruction of high resolution,
high signal-to-noise volumetric images. Challenges related to the analysis
of data from PET studies of small animals will also be identified,
particularly in the areas of brain imaging in rodents, multi-modality
imaging and high throughput imaging.


Back to Long Programs