Courtesy: Avnet
When a doctor tells you to get a CT scan, they’re calling on a powerful medical imaging technology for insights only otherwise possible through invasive procedures.
Computed tomography (CT) exploits the penetrating nature of X-rays. A standard X-ray shines a 2D beam of high-energy photons through the subject. How these photons are detected has changed over the years. It was once just photographic film, but today it is more likely to be a digital detector.
Since bone, muscle and fat each absorb X-rays differently, the image captured is effectively the shadow cast by the mix of tissues in the body. Rather than create an image directly from the photons detected, computed tomography processes that captured data to synthesize an image.
In CT scanning, a source illuminates the subject using a fan-shaped beam of X-rays that are picked up by an arc-shaped array of digital detectors. The source and the detector are mounted on a circular gantry, which rotates around the patient, taking scans from multiple angles.
The resulting scans are not directly interpretable as an image. The scans are combined in a computer, creating a more detailed 2D “slice” through the body. Many CT scanners also coordinate the movement of the patient with the gantry’s rotation, creating a sequence of slices through the body that can be processed into a 3D image.
Seeing the advantages of CT in medical imaging
patients.
CT scans can render more detail about internal structures than ordinary X-rays and can present that data in augmented 2D or 3D, making it easier to interpret. They are also relatively fast, which makes them useful for providing insights about injuries to the head, spine, chest, abdomen and pelvis.
The ability to post-process CT images means that they can provide a useful basis for detecting tumors and cancers, their size, location, and how they have spread. They can also reveal internal bleeding and the spread of infection as well as enable doctors to visualize blood vessels, aneurysms and blockages throughout the body.
CT imaging can be used to reduce the invasiveness of some procedures. The ability to differentiate diseased tissue helps surgeons avoid removing healthy tissue unnecessarily. Similarly, the 3D detail possible with CT imaging can provide a useful basis for planning procedures such as biopsies, surgery, implants, and radiation treatment.
The technique can reveal the detailed health of bones and joints, making it easier to understand wear or disease and to diagnose fractures. CT scans can also help track the progress of disease and reveal the effectiveness of treatments such as chemotherapy.
There are challenges associated with CT scanning. They usually involve greater exposure to ionizing radiation than is common with standard X-rays. Patients may also react badly to the contrast agents used to improve the CT scan’s effectiveness.
Navigating the key trade-offs in CT scanner design
Developers of CT scanners work with two forms of constraints. The first is the paramount nature of combining innovation with patient safety when using ionizing radiation. The second is the tension that can bring for patients and care providers. The availability of new technologies and capabilities must always be met with the best judgment and a conservative attitude.
There are technical trade-offs in CT design too. Perhaps the most important of these is between image quality and radiation dose levels. Higher doses may improve image clarity at the cost of greater exposure. Technologies such as iterative image reconstruction and denoising algorithms based on machine-learning techniques can now replicate some of the image-quality gains of high-dose scans at lower doses.
Trade-offs also exist between the speed of scanning and image resolution. Faster scanning reduces artifacts introduced by patient movements but can result in lower spatial resolutions. Scanning more slowly can deliver higher resolution if the patient is still for long enough.
Designers also need to decide which market niche they want to address. For example, designing a scanner with a large field of view makes it easier to scan large body parts, but requires larger, more expensive detector arrays. Other trade-offs may have to be made between scanner flexibility and specialization, hardware quality and maintenance costs, and software capabilities and reliability.
There are also financial considerations over initial cost and long-term upgradability. Buying a CT scanner means a large upfront outlay and substantial operating costs. Buyers may be prepared to choose a scanner engineered to evolve, rather than a lower-cost machine with a more limited useful lifetime.
Making innovations in CT design
Despite the constraints and trade-offs outlined above, there are many avenues for innovation in CT design. Each detector usually has a scintillator, which emits visible light when it is hit with X-rays, mounted over a digitizing photodetector circuit. A basic CT scanner will have one arc of these detectors, but more sophisticated variants will have multiple arcs so that they can sample multiple “slices” simultaneously. There may be as many as 256 arcs.
To support the high number of detectors, semiconductor companies are engineering 128-channel analog-to-digital converters (ADCs). These ADCs can be mounted in modules to produce 256-channel capabilities. The chips have low-power, low-noise, low-input-current integrators. Simultaneous sample-and-hold circuits ensure that all samples are taken at once. Some ADCs targeting medical applications offer resolutions of up to 24 bits.
Achieving low-dose CT imaging
Different beam energies can reveal different things about the subject they are illuminating. Radiologists can adjust the beam strength used in the scan to pick out specific details. This is called the spectral CT technique.
Another approach is to use a dual-layer detector, with the top layer absorbing the lower-energy X-ray photons and a lower layer absorbing the higher-energy photons. This technique can reveal more about how the X-rays have been affected by their passage through the subject material.
A further innovation involves single-photon capture detection, in which a semiconductor device is used to directly count each X-ray photon. This gives scope for lower-dose CT imaging, since it does away with potential photon losses in the scintillation process of conventional detectors. It also makes it possible to measure the arrival energy of every photon, again giving greater insights into how it has been affected by passing through the patient.
In dual-source CTs, two source/detector array pairs are mounted on the rotating gantry ring at 90 degrees to each other. This arrangement gives good coverage of the patient while minimizing interference between the sources.
The two sources can run at different energies, which brings the advantages of spectral CT discussed above. They can also acquire a whole slice image more quickly than a single-source scanner, which gives them greater temporal resolution for imaging moving features such as a beating heart. This in turn reduces motion artefacts in the final scan. Faster scans may also be more acceptable to some patients.
Developing high-resolution CT scanners
High-resolution CT scanners produce very thin slices of less than 1 mm. They use more, smaller detectors, to achieve higher spatial resolutions than standard scanners. The extra resolution makes it easier to detect and characterize small features accurately.
Such scanners usually have sophisticated image-reconstruction algorithms to enhance image quality and detail, which is particularly important for visualizing fine structures and edges. They can also have features such as enhanced X-ray beam management. These techniques give higher contrast images than standard scanners.
CT scanners are enormously valuable for producing insights into patient health without the need for invasive procedures. Their developers can call on rapidly evolving technologies, such as detector electronics and machine-learning techniques, to provide enormous scope for innovation. Responsibility for patient safety means the adoption of new technologies can feel slow.
Fortunately, designers can make a real difference here by exploring the systemic trade-offs involved in the development of novel CT scanners to produce capabilities that are engineered to encourage rapid uptake. For example, designing a detector sampling and digitization circuit with a lower noise floor will enable higher-resolution scans at the same beam energy, or similar resolutions at lower doses.
An FPGA accelerator board may be used to speed up image-processing algorithms, increasing the scanner’s throughput and so cutting the cost of individual scans. Or perhaps there’s a better way to manage power use in the scanner, extending its reliability and so cutting its operating costs.
Avnet recognizes the holistic challenge of developing medical imaging products and has the resources to help OEMs address them.
