Medical imaging: when disruption comes from merging technologies

Photo Serge Ripart & Ludovic Le Meunier / Head of MRI Marketing, Siemens France & Business Unit Manager, Molecular Imaging, Siemens Healthcare / October 23rd, 2013

With 3-D medical imaging rapidly coming on line, a silent revolution is under way in our hospitals and research establishments. Practitioner’s techniques are constantly improving and gaining ground, but there are limits. New strides forward will come from combinatory techniques, for example, by marrying PET-Scan and MRI: two very different technologies, one of which is familiar to patients, the other of a more confidential nature. Magnetic Resonance Imaging (MRI), the older technique, has been in use for several decades now. The second, developed later is the Positron Emission Tomography (PET), usually referred to as the PET-Scan.

ParisTech Review – PET and MRI are two very different technologies. Where did the idea to make the combination originate?

Serge Ripart – To better understand this move forward and the progress thus far, we can sum up for you the respective features of MRI and PET-Scans and mention their limits.

MRI technology became available about 50 years ago. It allowed us to diagnose anatomic or morphological pathologies, for example, cancer tumours. The underlying physical principle is nuclear magnetic resonance, based on the quantum properties of atomic nuclei to use them in chemical spectroscopic analysis. The patient is first positioned in a strong magnetic field, which magnetises biological tissues; he or she is is then subjected to oscillating magnetic fields of lower intensity which trigger electromagnetic signals. The MRI technique consists of localising the sources of the signals accurately and, on this basis, can map out the chemical composition of the tissues being explored.

It proves extraordinarily useful to detect numerous neurological or vascular illnesses (except our coronaries) and osteo-articular disorders such as those affecting our ligaments. Magnetic resonance proves very efficient to discover and analyse tumours in the abdomen, the liver, the pancreas or the prostate gland. In the latter case, MRI technologies are absolutely necessary since the prostate is a difficult organ to visualise using other diagnosis tools. MRI techniques, it should, be noted, are totally inoffensive inasmuch as they do not produce/use ionising radiation (as do X-rays). There are, however, certain constraints: the patient is enclosed for about 20 minutes in a tunnel, not exactly a comfortable situation, with various ambient noises and it can be an anguishing experience, especially for children. More important is the fact that certain tissues do not provide a very sharp image, because they have a low hydrogen content, e.g., our bones and lungs.

Is this the reason why MRI diagnoses are often completed by a Pet-Scan?

Ludovic Le Meunier – Both technologies are in fact complementary. MRI lies in the category of structural imaging tools: it allows practitioners to access images of the patient’s anatomy and to visualise his/her organs. Pet-Scans are used to visualise cellular activity: in this case we refer to functional imaging. It is a very useful tool to detect pathologies with modifications of normal physiological functions.

Another difference is that PET-Scans are a form of scintigraphy, i.e., they use radio-isotopes, injected in the zone we want to analyse, for example, 18F a fluoride isotope. This radioactive material is produced specifically for this utilisation [in the form of a radiolabelled sugar (glucose) molecule] where the 18F has a half-life of only 110 minutes. When it disintegrates, the 18F atom emits a positron which is annihilated when it encounters a local electron, after travelling 1mm approx. The annihilation event produces two photons that travel in opposite directions on the same vector. Sensors positioned round the patient detect the photon pairs and thereby localise the emission vector. A specific software package then rebuilds a 2-D or 3-D image, with an optical resolution of a few millimetres.

In a Pet-Scan image we do not see the body as such but rather teaches us how it functions. For example, we can distinguish how a tumour reacts. But PET-Scans also have their draw-backs. Firstly there is the need to inject a radioactive tracer, which will attach itself to certain zones depending on a tumour’s activity. That is why we often prefer to use a MRI for pregnant women or for young patients. Another difficulty with PET-Scans is that they can assess physiological situations but do not offer very sharp images. This is why the MRI and the Pet-Scan are complementary – the images from one technology can compensate for the shortcomings of the other.

Left: MRI image; centre: PET; right: combination (source: Siemens)

How did the idea to combine both technologies arise?

Serge Ripart – The need was identified by radiologists about a decade ago. The obvious complementarity of both tools (MRI and PET) but more than that, our strong desire to make the diagnostics more accurate, encouraged us to carry out research to combine both functionalities in one machine. When we took two images at different times (on separate machines), there was not an optimal correlation of the information. Between a MRI examination and a PET-Scan the condition of the patient’s pathology can evolve. Moreover, phenomena such as the patient simply breathing, or moving in the tunnel, make the results difficult to compare and seriously compromise their interpretation. Bringing both technologies together brings with it the advantage of accessing ‘real-time’ physiological and anatomic information, in a single examination session. The objective was to replace a sequential examination by a simultaneous examination.

The idea, per se, seems attractive, but concretely, how did you ‘fuse’ the two tools together?

Serge Ripart – Well, rather than ‘fuse’, I would say we integrated the tools. To be more precise here, we integrated a PET-Scan camera in a MRI machine. This, of course, required a long period of R&D before the first prototype was readied and presented in 2006. We notably had to replace one of the central components of the PET since the magnetic fields produced in the MRI seriously impaired the electronics of the PET equipment.

Associating these two technologies therefore forced us to evolve. We had to develop a specific PET technology that was not influenced by the very strong MRI magnetic fields. And for the latter, we had to develop new arrays of antennae (positioned round the patient’s body or on the body itself to pick up the MR signals), such that the information coming from the PET-Scan was correctly received.

But, without minimising the technological efforts we had to deploy, we noted that the real break-through lay in the practitioners’ practice. A real revolution was underway. The doctors now had two simultaneous images to read. One of the obvious advantages is the time factor to handle the patient. But there is also a gain in quality and in information content: what we see is an association of the extraordinary analytical power of an MRI in terms of morphology, with the functional analyses from the PET-scan. In terms of the diagnosis of the patient, the leap is enormous, like shifting from 2-D to 3-D, viz., the image became far more legible, more informative. The diagnosis was better by far. We all know how important this phase is in the case of cancer diagnosis. Moreover, the fact that the machine delivered two sets of results simultaneously means that it was possible to draw conclusions rapidly and if need be to do a second MRI run immediately, with an adjusted target area to scan. The machine consequently speeds up the diagnosis procedure.

Ludovic Le Meunier – Let me just mention another potential benefit here: using MRI to enhance the images from the PET-Scan – where, as I mentioned, the resolution or sharpness is not yet optimal – for very simple reasons. All the 18F atoms do not disintegrate at the same time, and when we see that the patient’s body moves as he/she breathes, the image becomes somewhat blurred. It is precisely here, with the conjunction of the two technologies, that this blurred effect can be corrected and thereby offer the practitioners a much improved PET image. But, I hasten to add, this is still a case of “work in progress” with, needless to say, enormous potential.

Serge Ripart – One final, but important, point to note – in keeping with the general evolution of diagnosis protocols today – is that the techniques used are less and less invasive to the patient’s body and less damaging. In the case, for instance, of certain tumour pathologies where frequent examinations of the patient are needed, the progress is noteworthy. There is also a welcome progress in paediatrics. In New York, there is now a clinic for children that uses only this combined approach for its analyses…

Can we stop for a moment on the question of disorders. What are the illnesses most concerned by this technique?

Serge Ripart – In the beginning, our objective was to improve the diagnosis of neurodegenerative diseases such as multiple sclerosis or Alzheimer. The first piece of equipment we assembled was limited to doing head scans. But fairly quickly we decided to widen the scope of our machine’s possibilities and assembled a machine capable of analysing a complete body to be used in diagnosis of thorax, breast or prostate cancers. A second prototype was presented in December 2010.

Obviously, it is in the field of cancer diagnosis that these machines and this combined technology will prove most useful. We can observe a dual trend: continued improvement of diagnosis protocols on one hand and an increase in better, targeted, analyses, on the other. For certain cancers, such as prostate cancer, the earlier the diagnosis is made, the easier it is to prescribe targeted and less invasive remedial therapies. We have a similar case for lung cancer or breast cancer. Today we can have better targeted chemotherapies, notably for soft tissues. But again, the detection and diagnosis must intervene fairly soon and be accurate.

Biograph exemple
A better diagnosis (source: Siemens)

We suppose that this technology is quite expensive. Can we envisage today that it will spread as rapidly as MRI equipment did, for example?

Ludovic Le Meunier – We are at the start of a life cycle; the first machine was installed in July 2011, in Germany. There are now about 50, two thirds of which are in service: in the USA, in Korea … Early 2014 another unit will be installed in Lyons, at the CERMEP research establishment [the acronym being for the Centre for exploration and medical research using positron emissions.] This unit is already seen as very attractive for various anti-cancer centres and the universities.
The investment is high and we can only surmise that such machines will be reserved for rich countries. However, we can envision and even express the wish that they become accessible to emerging countries. As for any other imaging process, there could be a progressive development of units adapted to these countries’ financial investment possibilities. Technically speaking, there is no major obstacle to seeing this innovative technology become wide-spread.

Serge Ripart – I could add that for the moment Siemens is the only player in the field to have developed the dual MRI-PET technology, but obviously the arrival on the market place of another constructor would make the market more dynamic. The potential for development we have identified through our market studies indicates that about 50 machines could be sold in the world each year.

Nevertheless, you have to realize that we talking about a market that differs from others. The regulations that apply to equipment of this category are more or less stringent. In a country such as France, so-called ‘heavy’ imaging protocols have to be approved and the market is extremely controlled: only the Regional Health Service Agencies are entitled to deliver the authorisations. Second point, even in developed countries, there are considerable differences in terms of the equipment, from Region to Region, and likewise from country to country. France now has one MRI for 90 000 inhabitants; Germany has one for 30 000. And the question of the cost of the equipment cannot be taken for granted; they are expensive, whatever the services they render – around 200 000 euros for a classic MRI scanner and up to several million euros for more sophisticated models.

And you will note that we are referring to a technology that today is fairly ‘old’ and you have noted that it has spread in a remarkable manner over the past two decades. When dealing with innovating technologies and in the early stages of a life cycle, we must be patient. At this point in time, MRI-PET machine locations will mainly be in research establishments. Naturally – and very rapidly we hope – the hospitals and clinics will follow suit.

Note from the editors: Siemens is one of ParisTech Review’s patrons. This article is part of a series started with an interview with Dr. Wafa Skalli (Arts et Métiers ParisTech) and to be continued in the next months.


  • Technological Breakthrough: New Scanner Could Revolutionize Cancer Diagnosis (Spiegel Online, Jan. 2011)
  • Positron emission tomography (Wikipedia)

More on paristech review

By the author

  • Medical imaging: when disruption comes from merging technologieson October 23rd, 2013