At least three technology-based trends are shaping medical imaging. One is the greater detail from CT and MRI equipment that captures and processes more data in one scan. That's good for patients because shorter CTs expose them to less radiation. Secondly, software and computers more quickly crunch through enormous files and present results in clearer images that let doctors check organs and even cells for signs of disease. And lastly, nearly all images are digital, making them easier to transmit and store.

The trends are letting doctors alter surgical strategies, especially for risky operations such as separating conjoined twins. Take the recent case of twins joined at their skull. “After examining the images and viewing holograms of the shared vasculature and bone, surgeons decided on several separate operations rather than a single long one,” says Daniel Burman, president of Holorad Inc., Salt Lake City, (holorad.com). His company transforms CT and MRI data into 3D images that are more telling than either format alone.

Of course, a variety of supporting technologies surround the trends. Higher density flat panels, for example, let doctors see details more clearly, and sensors are now available that register lower dose X-rays without losing information.

More detail, lower dose

The goal of gleaning more visual information from shorter scans has companies almost reinventing their CT equipment. “We're working with new CT technology that will let clinicians see more,” says GE Healthcare's research manager for molecular imaging Sholom Ackelsberg.

He says resolving the paradox of improving image clarity while simultaneously reducing patient dose requires examining the fundamentals of spatial resolution, low-contrast discrimination, and dose efficiency. “Adding more slices or X-ray sources does nothing to improve image clarity” says Ackelsberg. Consequently, GE is revamping its CT systems, from X-ray tubes through detectors and data acquisition, even reinventing how images are reconstructed.

Part of the revamping is in finding ways to acquire data from two energy levels with improved temporal registration, using a single source and detector. Dual-energy acquisition involves using X-rays generated at two different peak energy levels to better discriminate between different tissue types. This new approach involves switching X-ray generation between 80 kV and 140 kV and back in less than 1 msec during a scan. Conventional scanning works with a spectrum of X-ray energy, from 0 to a peak of 140 kV. Dual-energy scans make it possible to synthesize any single energy scan from the two spectra. This has potential to significantly improve the quality of CT images.

To make a good CT image you need about 1,000 views per rotation from a spinning gantry. With the new data acquisition technology running at 7.3 kHz, a 0.35-sec rotation provides about 2,500 views. Interlacing the views from each energy level provides sufficient data to reconstruct two complete sets from a single rotation. Data from the two power levels are then combined into one image.

Even though the X-ray generator and detectors are rotating and the patient is moving, Ackelsberg says HDCT (high definition CT) images have temporal registrations about 150 times better than those taken by current technologies. “Dual-energy imaging has the potential to more accurately quantify body-tissue densities. This may let doctors see inside stents in cardiac arteries, and without artifacts common to imaging dense objects in the body,” says Ackelsberg.

Competing systems, he says, take sequential patient scans from two tubes separated by 90°. This setup takes about 83 msec to switch from one voltage to another, so a scan can take 100 times longer and that introduces a loss of clarity called temporal misregistration.

In addition, dual-energy data processing lets GE developers reconstruct CT images that show reduced beam-hardening artifacts, and hence, more accurate images. Artifacts come from some X-rays penetrating more than others (become harder) while passing through matter. Artifacts show up as streaks or dark shadows that reduce image quality.

A new, faster scintillator or X-ray detector based on the optical properties of the garnet gemstone is in development. In a nutshell, a scintillator converts X-rays into light through a photodiode which is then converted to digital data. Ackelsberg says this new and faster detector and data-acquisition system are key to making possible fast-switching, dual-energy imaging.

“Most manufacturers of CT scanners have similar spatial resolutions, but recent developments should increase spatial resolution by 30% throughout the body and about 50% in the heart.

A key contributor to detecting subtle differences is the amount of noise in the image. Less noise makes it easier to differentiate objects. New image reconstruction techniques can reduce noise. These methods can be applied to either reduce noise from the same X-ray exposure (thus improving low-contrast detectability) while maintaining spacial resolution, or clinicians can decrease the dose by 50% and maintain the same noise and image quality available today.

Other companies are also working to minimize patient exposure to X-rays yet capture enough information for useful images. Dalsa Corp., Waterloo, Ontario, Canada (dalsa.com) says its recent series of 12-µm CCD sensors can further minimize patient exposure to X-rays. Their full-frame sensors have improved quantum efficiency (it turns more X-rays into electrons) and a sensitivity that generates a useful image with lower X-ray doses than previous sensors. Recent CCD designs also produce high signal-to-noise ratios.