Physics of bio-imaging "needs better understanding"

11 Feb 2013

Images contain a wealth of information but much of what it means remains a mystery.

by Tim Hayes
Biomedical imaging techniques continue to develop rapidly, but making a solid connection between the images obtained and eventual clinical outcomes remains challenging – according to a special “hot topics” presentation at SPIE’s Photonics West event last week.

"This is the really beautiful, but also somewhat horrifying, thing about this field," said Bruce Tromberg of the Beckman Laser Institute and Medical Clinic, addressing the special BiOS session on biomedical optics.

"We are able to obtain so much information and content from various medical optics techniques, but can sometimes have little idea exactly where it comes from or what it means."

Some of the disconnection between the two aspects stems from the basic physics involved. "We need to understand the functional origins of the contrast we see in the images, which means better measurement of the path lengths taken by the photons in the tissues," said Tromberg.

Techniques to achieve that, such as frequency-domain photon migration analysis with pulsed near-infrared illumination, are currently moving to market in oximetry systems to monitor blood oxygen content, although not all have received FDA approval as yet.

Blood oxygenation
The connection between blood oxygenation in the brain, as measured by the “StO2” blood saturation parameter, and the long-term survivability of certain cardiothoracic surgeries is now becoming better understood, making this particular bioimaging technique attractive for anesthesiology and for monitoring hemodynamics in the operating room.

Oxygenation is also a key metric in breast cancer surgery, an area where the need for repeat procedures is a known concern. A two-year trial is underway to assess whether a combination of optical imaging and oxygenation measurements made even before anti-cancer treatments begin can be correlated with the ultimate success of tumor removal.

"Discovering the exact outlines of tumors optically is hard enough; assessing their response to chemotherapy can be harder," commented Tromberg, who believes that using steady-state spectroscopy alongside frequency-domain photon migration analysis can monitor a broad enough range of spectral features to yield functional images of anti-cancer treatments.

"Separating absorption effects from photon scattering or determining blood perfusion and chemical concentration are questions of physics and physiology," he said "But from the medical perspective, we need outcome studies. Those are the important things."