Industrial Case Study: Low-Dose Dental 3D X-ray Imaging

In years 2001-2007, Finnish researchers and engineers in industry and academia joined forces and developed a novel low-dose tomographic X-ray imaging device. At the time I was working at Instrumentarium Imaging, having completed my an applied math PhD degree in 1999. My role was not only that of an R&D scientist, but also that of an interpreter between the company and the universities involved. The invention is called the VT product (for Volumetric Tomography) and it has been used at dental clinics around the world from year 2008. This is the story behind the VT device.

References:

Hyvönen N, Kalke M, Lassas M, Setälä H and Siltanen S 2010,
Three-dimensional X-ray imaging using hybrid data collected with a digital panoramic device.
Inverse Problems and Imaging 4(2), pp. 257-271. PDF (724 KB)

These days it is possible to replace a missing tooth with a dental implant. It consists of a titanium screw and an artificial tooth attached on top of the screw. The dentist needs to drill a hole in the bone for the attachment of the screw. The hole needs to be deep enough for firm connection, but not too deep as crucial nerves may be damaged. How to measure the appropriate depth of drilling? Well, using X-ray imaging, obviously, but how, exactly?

Since the 1970's, panoramic dental imaging has been a basic workhorse of dentists. Panoramic imaging was invented by a Finnish university professor called Yrjö Paatero in the 1950's. He grew tired of taking piles of small intraoral radiographs of each patient and decided that there has to be a way of imaging all the teeth on one big sheet of film in one fluid movement. Shown below are technical drawings from Paatero's patent and an early model of his Orthopantomograph device.

This is the classical imaging situation of dental panoramic imaging. Thanks to Maaria Rantala for her help and to Palodex Group for the opportunity to record this video at their showroom. The X-rays were disabled in this test sequence.

A panoramic dental image shows pretty much all the teeth and surrounding tissues such as sinuses. It is an overall view of dento-maxillofacial structures. There is a curved sharp layer located roughly along the dental arc. Any details away from the sharp layer are blurred by the mechanical movement of the device. However, the panoramic image involves some geometric distortion (for example, fan-beam geometry vertically and parallel-beam geometry horizontally), so dentists cannot reliably measure the correct depth of drilling when planning implantation.

We reprogrammed a digital Orthopantomograph device so that it collects two-dimensional X-ray projection images with a scanning movement. Note that the detector is vertically long but horizontally narrow, so scanning is necessary. Also, the degrees of freedom of the machine as well as the potiioning of the patient's head forced us to restrict the imaging to a limited angle of view.

Here you can see the geometry of imaging as well as some example X-ray projection radiographs taken from a dry skull specimen.

This slide shows the basic philosophy behind VT imaging: radiate the patient as little as possible, only just enough to recover the crucial information. The 3D reconstruction below is far from perfect since we can see ringing artefacts and blurring of details. However, the dentist can reliably measure the important thing: appropriate depth of drilling for attaching the implant.

In 2006, this was the radiation dose of the novel VT imaging modality compared to competition. It reduces the dose with one or two orders of magnitude.