Practice SafeCT

Axial scan, 0.6 mm slice thickness, acquired with a LightSpeed VCTTM by GE Healthcare and displayed in a Soft Tissue window of W=60; C=40. SafeCT processed image on right.

Image courtesy medicvision.com 

This entry, by the way, does NOT qualify as a Naughty Bit. In fact, it's about something rather impressive.

Last year, we were quite involved in selecting an advanced visualization system. We were at one time under the impression that one of the products under consideration was going to deliver a added bonus: Dose Reduction! But sadly, this turned out not to be the case. Please read this Old Post for details. As you know, we went with TeraRecon, although it has yet to be installed. Good things come to those who wait.

In the meantime, we have been fortunate enough to afford to replace several of our CT's with the latest and greatest from Philips, which have quite advanced dose reduction built into their hardware or software. But what of the old scanners we can't yet replace? If a patient is so unlucky as to be scanned when the low slice slicers are the only thing available, do they deserve a higher dose of radiation? Now, I think I need to mention that one scan on the higher dose machines might theoretically carry a greater risk of something, but really, the excess dose is probably not all that dangerous. Still, we treat radiation with respect, and try to reduce the cumulative dose by limiting the dose of each individual event. Alas, we cannot go back and retrofit the old but serviceable scanners with the new dose reduction technology.

Or can we?

I don't know off hand if Atlantis Worldwide contacted us, or if our physicist found them. Atlantis deals mainly in used equipment, an honorable and difficult task. But they are the US representatives of an Israeli company called MedicVision, the creators of the product in question. I've had the joy of a WebX demo, and today, the chance to speak to some of the scientists behind a new approach, SafeCT. To be brutally honest, when I first heard of it, I thought it was nothing but bullsh*t. After much more exposure, I think I was full of bullsh*t, and SafeCT might well be the answer for many in our position.

So what is it? What it is NOT is a simple filter. What it is NOT is a hardware solution. It IS a software solution, utilizing a "novel" non-linear three-dimensional post-processing iterative image reconstruction algorithm that increases SNR and allows CT radiation dose reduction. From the MedicVision website:

Medic Vision’s SafeCT is based on proprietary patented iterative volumetric algorithms technology for Signal-to-Noise Ratio (SNR) enhancement of CT studies acquired over a wide range of exposure parameters on conventional CT scanners. SafeCT is compatible with all major CT platforms and PACS systems. It can serve multiple CT scanners simultaneously via their DICOM network. SafeCT has been in clinical operation at leading medical centers and private radiology practices in the Unites States, serving thousands of patients. SafeCT is FDA-cleared for distribution in the USA.
Excellent Image Quality

• Reduced image noise
• Improved visualization of anatomical detail
• Preserves the look and feel of images the site is accustomed to. 

High End-User Value

• Add-on product compatible with any CT scanner
• Eliminates the need for costly scanner upgrades or replacement 
• A single SafeCT simultaneously serves multiple CT scanners from different vendors and models, providing a solution for the entire department.
• Rapid processing—no impact on image reconstruction time at the CT console
• Unchanged clinical workflow—seamless to both technologist and radiologist
• Scalable product capable of accommodating future scanner additions.  

The website has links to dozens of papers which validate the stuff. However, I looked at a bunch of pictures which were truly worth a thousand words. If everything is as it seems, the technique enhances noisy images and reveals pathology, EVEN ON IMAGES SCANNED WITH REDUCED PARAMETERS. In other words, pump in less radiation for the scan, and use SafeCT to rescue the image, and maybe even get a better image than you would have otherwise. (They compare favorably to GE's ASIR for example.) MedicVision suggests that we could routinely cut the doses by 50%, and with more experience and experimentation, possibly to an even greater degree.

I was worried that dropping the dose and impairing the image would lose data and detail. As near I can tell, SafeCT does just the opposite. The processed images seem quite diagnostic, with lesions clearly detailed that were essentially invisible or at least not well-demonstrated prior to their digital massage. In fact, the algorithm works in part, I was told, by edge-detection, incorporated within iterative reconstruction, thus detecting lesions in a way. I suggested to the inventors that they parlay this into a CAD display, lighting up the detected anomalies in red, for example. I'll take a 5% royalty on that one.

SafeCT works by placing a processing box between the modality and the PACS, rather like any other advanced imaging processor. It's a little pricey, but in the right setting, it may be well worth the cost.

This is a product worthy of your attention.  Have a look!