Williams, Mark B.
Professor, Radiology and Medical Imaging
- BS, Physics, Grinnell College
- MS, Physics, Wake Forest University
- PhD, Physics, University of Virginia
Radiology and Medical Imaging
Charlottesville, Virginia 22908
Biomedical Engineering, Biophysics, Translational Science
Design, Development and Optimization of Medical Imaging Technologies
The primary areas of research are the development of novel systems for medical imaging, quantification of basic performance characteristics and efficacy of imaging systems. Of particular interest are imaging systems utilizing x-rays (i.e. radiography, x-ray tomosynthesis, x-ray computed tomography (CT)) and/or nuclear medicine (i.e. scintigraphy, gamma ray emission tomosynthesis, single photon emission computed tomography (SPECT), and positron emission computed tomography (PET)). In recent years we have focused on the development of multimodal hybrid systems that integrate spatially correlated anatomic and functional image sets. Current focus application areas include breast cancer detection and characterization, intraoperative image guidance, and pre-clinical in vivo imaging.
Current Major Projects
1) Dual modality tomosynthesis breast imaging: We have developed an integrated imaging system that combines the sensitivity of digital breast tomosynthesis with the specificity of molecular breast imaging (breast-specific gamma imaging) in a single upright unit. The system is designed to obtain diagnostic information regarding suspicious or radiographically occult mammographic findings, and functions by obtaining sequential series of x-ray transmission and gamma emission images over a limited range of viewing angles with the breast in a single configuration. Co-registration then correlates the 3-dimensional x-ray and gamma ray tomosynthesis images to within a fraction of a voxel. This system is currently being evaluated in a human study among women schedules for breast biopsy.
2) Tomographic molecular breast imaging: We are adapting a two-head gamma camera designed for conjugate imaging of the breast to permit dedicated breast SPECT. The acquisition geometry is upright and seated. Iterative SPECT image reconstruction incorporates resolution recovery and results in nearly isotropic spatial resolution. The utilization of two independently positioned cameras permits sampling artifact-free performance with clinically acceptable acquisition time. Breast immobilization (rather than compression) is accomplished using a unique structure that permits small collimator-to-skin separation throughout the scan while maintaining a natural, comfortable breast shape.
3) Dedicated breast ring PET: Clinical whole body PET (WBPET) scanners have inadequate spatial resolution to reliably detect primary breast malignancies smaller than 1 cm. Our group is developing a breast ring PET (BRPET) scanner that uses a full ring of 12 detectors that surround the breast for high spatial resolution and high photon detection sensitivity. Each detector in the ring contains a 5 cm x 5 cm array of high density lutetium-yttrium oxyorthosilicate (LYSO) scintillator crystals whose light emission is channeled to a position sensitive photomultiplier tube via a slanted fiber optic light guide. The slant design permits the scintillator crystals to be placed up inside the table hole, mitigating the deficiency of virtually all other prone table breast imaging systems: limited visualization of structures or tracer near the chest wall. The BRPET scanner will be evaluated in a pilot study comparing its sensitivity, specificity, positive and negative predictive values against those of contrast-enhanced breast MRI, using the results of breast biopsy as ground truth.
4) Intraoperative SPECT imaging using a handheld gamma camera: In collaboration with Dilon Technologies, SurgicEye, and the Jefferson Lab, we are developing and evaluating a system for radio-guided surgical procedures such as sentinel lymph node biopsy. The system uses infrared tracking of a handheld gamma camera to create a 3-D map of the distribution of gamma-emitting compounds (e.g. lymphatic or tumor-targeting radiotracers) during surgery. The camera is based on silicon photomultiplier (SiPM) technology, resulting in a compact device that is light enough to be hand-operated. The system will be evaluated among melanoma patients undergoing sentinel lymph node biopsy. It is anticipated that, compared to current practice utilizing pre-operative lymphoscintigraphy using large area, general purpose gamma cameras followed by intraoperative node excision using non-imaging gamma probes, the handheld SPECT system will result in more accurate node localization, less ambiguity for the surgeon, and greater comfort for the patient.