Moving from CCD to CMOS in Medical imaging: RUBY, the next generation Camera Lens Assembly for C-arms (part 2 towards RSNA 2017)

Next generation camera lens assemblies with CMOS image sensors for C-arm X-ray systems can improve image quality and reduce patient does. Adimec’s RUBY delivers full digital real-time 1k2 30fps imaging for Image Intensifier based medical X-Ray imaging systems. This is the second of a three-part series on Adimec’s new Ruby camera lens assembly (CLA) for X-Ray imaging. RUBY is specifically designed to reduce costs in fluoroscopy C-arm systems compared to existing CCD based solutions (Part 1), while improving image quality and reducing patient dose (Part 2), and doing that in a system that is simple to install and service (Part 3).

There are five key parameters that define the quality of an X-Ray image in mobile C-arm systems: Resolution, Contrast, Dynamics, Noise, and Artifacts. This blog will focus on these parameters relative to the RUBY camera lens assembly (CLA).

Resolution or the number of pixels per unit area provides the overall sharpness of the image sensor given focus and dose are optimized for the scene. The RUBY provides industry standard 1k2 or 1 Megapixel resolution. Full automated precision alignment of the sensor to the optics on optimal image MTF at the factory assures the best possible sharpness of the entire 1k2 image, repeatable for each RUBY produced.

Contrast or the ability to see details in the bright and dark portions of the image is a strong point of CMOS imagers. RUBY’s CMOS technology out-performs CCD in low-light contrast sensitivity which could both add contrast to the fluoroscopy images and/or reduce patient dose however this is left to the system manufacturer and in part the radiologist who sets the dose parameters. The widely used CCD sensor in mobile C-arms (KAI1020CCD) has a dynamic range (DNR) of 60 dB which is 1000 gray levels while the CMOS sensor used in RUBY has a DNR of over 74 dB or 4000 gray levels. Grey levels translate to the camera’s ability to see details. This contrast benefit is not at the cost of forced sensor cooling and offers dynamic snapshot imaging at better sensitivity, lower read noise and higher dynamic range (14 bit). X-Ray medical imaging is a low light imaging application. The clinical information is typically in the dark portions of the image (5 to 20% saturation level). RUBY, with its CMOS sensor, captures more signal details in dark which results in better details after clinical processing. Actual patient dose depends on the choices of the radiologist in setup and usage of the system. RUBY’s low light imaging capabilities in combination with effective image processing may help to reduce the actual X-Ray exposure levels at which the radiologist acquires the image which therefore would reduce the dose that the patient receives.

Dynamics or blurring due to motion can be handled easily with the CMOS global shutter sensor. RUBY is operated in the same way as existing global shutter CCD cameras, supporting both pulsed and continuous fluoroscopy acquisition.

Noise. For a detector in an X-ray imaging system it is essential that its contribution to the total noise perception by the end user is not noticeable. Often the X-ray photon noise is the dominant noise source in the image. How does noise contribution of CMOS related to the known reference of CCD? Let’s make a technical comparison between the KAI1020CCD and the CMOS technology in RUBY. The KAI1020 CCD has a read noise of 40 electrons while the RUBY CMOS sensor has a read noise of 3.6 electrons and thus provides much greater contrast sensitivity from an application point of view. A visual example analog of noisy X-Ray images is shown below of the same scene, in low light, from the perspective of a CMOS camera (RUBY) and a recent CCD camera (newer and higher performance than the KAI1020 CCD). It is clear that the image on the left is easier to read. At low light levels, the CMOS technology outperforms the traditional CCD. As X-ray medical imaging is in essence a low light imaging application, RUBY with state of the art CMOS technology contributes to seeing more details in dark, thus at low dose.

Artifacts are defect pixels that provide signal that is either higher or lower than those around it on average. They are typically not dead pixels. Adimec uses proprietary algorithms to average out defect pixels and thus get useful data from them. Only single isolated pixels are replaced when ‘defective’. Clusters of defect pixels are not allowed. This ensures clinical details in the image will not be obscured for the end user. For sure, defect corrections can be turned off when needed. In addition, RUBY drives its CMOS sensor in such a way that large gain stages can be applied to increase image brightness without introducing blemishes such as known with uncooled CCD technology.

In summary, the Adimec RUBY CLA has the five parameters of X-Ray image quality covered and is at or exceeds those parameters for comparable CCD based systems.

Adimec is visiting RSNA2017 in Chicago. Feel free to contact us directly at strategy@adimec.com to meet us at RSNA and learn more about RUBY, PEARL, CMOS vs CCD in medical imaging and our other cameras for Clinical Diagnostics applications.

Stay tuned for part 3 about the RUBY by subscribing to our blog mailing list.