PET and MRI have long been used to obtain molecular, functional, and morphologic information to study the human body in health and disease. In addition to allowing the simultaneous acquisition of these complementary datasets, fully integrated PET/MRI systems have the potential to combine their strengths and alleviate many of their limitations. Although the initial efforts to combine these 2 imaging modalities were made in the preclinical arena in the 1990s (1), it took almost a decade for the major medical equipment manufacturers to recognize the potential of this emerging field. The first prototype device developed for human use allowed the simultaneous examination of the brain (2). This prototype was followed by the introduction in 2010 of the first fully integrated whole-body PET/MRI system, called the Biograph mMR (Siemens Healthineers). The results of the performance characterization measurements for this scanner were reported by Delso et al. in The Journal of Nuclear Medicine the following year (3).
At first sight, the article by Delso et al. (3) is similar to other papers reporting the initial results obtained with a novel imaging system. However, a careful analysis reveals, with the benefit of hindsight, that this paper was distinctive in many ways and managed to highlight the requirements for progress in the field and forecast some of the challenges it has faced over the last decade. Starting with the list of authors, the need for close collaboration between academia and industry, nuclear medicine and radiology, and physicians and physicists was emphasized. The methods section still serves as a blueprint for the types of studies that need to be performed for characterizing the performance of each of the components of a hybrid device within the constraints imposed by the other modality. As one example, assessing the PET image quality using standard phantoms required a calculated attenuation map as the MR-based methods specifically developed for human imaging were not adequate for this purpose. Although substantial progress has been made and correct human attenuation maps can now be generated for most body parts, phantom imaging is still challenging a decade later. From a technical point of view, the results presented confirmed that the hardware of PET and MRI components can be successfully integrated and that their performance is on a par with that of stand-alone devices. This confirmation was particularly relevant on the PET side, as the Biograph mMR was the first commercial system to use semiconductor-based photon detectors (i.e., avalanche photodiodes) as a replacement for photomultiplier tubes, which, until then, were used in virtually all commercially available PET/CT scanners. This advance arguably opened the road to the subsequent adoption of an even more advanced semiconductor-based photon detector technology (i.e., silicon photomultipliers) in PET/CT (4) and in latest-generation PET/MRI scanners (5). Finally, although only 2 proof-of-principle human studies were presented, this first technical report enabled clinicians to subsequently focus on assessing the clinical potential of this novel technology. In fact, the results of a comparison between PET/CT and PET/MRI in oncologic patients were reported by the same group in another highly cited paper published the following year (6).
As of July 2020, the paper by Delso et al. (3) had been cited more than 500 times by authors from more than 25 countries, proving its far-reaching influence. Furthermore, attesting to its multidisciplinary impact, the citing papers belong to a wide range of scientific areas (e.g., physics, chemistry, mathematics, engineering, and computer science) and clinical areas (e.g., radiology and nuclear medicine, oncology, cardiology, neuroscience, pediatrics, hematology, endocrinology, and gastroenterology). Remarkably, the Hirsch index (a measurement of the impact of a particular scientist rather than a journal) of this paper, which continues to be cited, is currently 50, with many of the citing papers having reached in turn the “highly cited” status in their fields.
Although 3 imaging equipment manufacturers are currently commercializing fully integrated PET/MRI systems for human use, the clinical adoption of PET/MRI has been much slower than that of PET/CT, with only approximately 250 systems being operational around the world a decade after its introduction. The exact role of PET/MRI within our health-care system is being explored for routine and advanced applications in oncology, neurology, and cardiology. PET/MRI could also enable several recent developments from the research arena (e.g., improved quantification enabled by MR-assisted PET data optimization, machine learning applied to multimodal datasets, and precision medicine informed by molecular imaging) to soon become clinical reality.
Just as the Biograph mMR marked the evolution of PET/MRI hardware from the prototype phase to the product phase, the report by Delso et al. (3) paved the way toward fully integrated PET/MRI investigations and served as the technical validation that facilitated the transition of this field from the research arena to the clinical arena. In addition to its substantial impact on nuclear medicine and radiology, this paper contributed to the wider embracing of these imaging modalities by other areas of medicine.
DISCLOSURE
No potential conflict of interest relevant to this article was reported.
- © 2020 by the Society of Nuclear Medicine and Molecular Imaging.
REFERENCES
- Received for publication June 23, 2020.
- Accepted for publication August 11, 2020.