Numerous MR image acquisition and reconstruction methods are devised each year, notably by researchers in academia. Only a few of them can be commercialized by industry, a prerequisite for them to become available on many MR scanners and to have significant impact on clinical care. In this presentation, some of the main criteria applied in this selection process and some of the key steps involved in the subsequent development process are outlined. In addition, some subjective suggestions are made to researchers interested in improving the chances of success of their methods in this respect.
Initial interest in a method proposed by researchers in academia may arise from publications, presentations, or discussions, for instance at this conference. It may be triggered by a particular problem or question, or it may be stimulated accidentally. It may also be driven by customer feedback or competitive information. The plausibility and the potential of the underlying idea are crucial at an early stage, while compelling results become decisive later on.
The method is then primarily assessed as to its promise to improve effectiveness or increase efficiency of an accepted MR examination, or to establish a new MR examination. This involves identifying which clinical needs it can meet and which clinical applications it can serve. Commonly, a broader scope is preferred. The method is also rated as to its disruptiveness. Generally, potential for a new product is favored over potential for a new feature or a mere improvement of an existing product. However, this is weighed against the required effort to commercialize the method, amongst other things. Furthermore, the method has to fit in with the overall business strategy.
Finally, a decision is taken whether the method is to be investigated internally, whether a collaborator in academia is supported in exploring it further, or whether progress on it is to be monitored only. In the first case, the maturity of the method determines which steps are still required in preparation of a development project and who within the organization is entrusted with them.
Starting from a publication or presentation, corporate research typically designs and implements a first prototype on one platform. Alternatively, a collaborator in academia shares his or her prototype. Initial experience is then gained with this prototype in-house on phantoms and volunteers in selected applications. If the method lives up to the expectations, the first prototype will gradually be refined. Simultaneously, the application range will successively be extended and the needs of potential users will be considered.
The code is then transferred to advanced development, where it is merged to the central software repository. The implementation is optimized to be sufficiently fast, and remaining tuning parameters are eliminated or fixed. Moreover, the user interface is streamlined. The scope of the evaluations is extended to all relevant platforms and applications. For this purpose, application specialists are involved. Additionally, patches are generated and released for initial external evaluations, which are overseen by clinical scientists. Discussions with marketing are initiated to develop a business case.
In a feasibility report, the status of the method is then summarized. The current implementation, the considered clinical applications, and the expected clinical benefits are described in detail, and the results of the evaluations, as well as lists of remaining work and risks, are compiled.
If the method seems sufficiently attractive and mature at this point, a development project may take it up. In it, the quality of the code is improved and extensive documentation is created to meet product standards. Moreover, a formal review of the design and the implementation is conducted. Test cases are collected for all target platforms and applications, and expected results are specified. Additionally, suitable metrics are defined to assess the performance of the method. Finally, verification and validation tests are performed.
For regulatory approval of the method, which is an introduction of new functionality as part of a medical device in this case, the FDA in the US demands evidence for safety and efficacy, for instance. This requires documented proof of appropriate image quality, lack of potentially dangerous artifacts, and accuracy of derived metrics. Clinical testing for this purpose can be facilitated by well-planned research studies prior to product release that demonstrate the usefulness of the technology, if a rationale for equivalence of the technology employed in the studies and in the product can be provided.
The development process may be postponed or stopped at any stage for various reasons. Among these are technical reasons, such as disappointing performance, insufficient robustness, and excessive demands on the hardware, as well as commercial reasons, such as low user acceptance or changes in priorities. The development process may also continue, but with another method. These risks commonly decrease over time.
Researchers in academia interested in seeing their methods adopted by industry may find considering the following subjective suggestions helpful: