Search for: All records

Abstract MRI acquisition and reconstruction research has transformed into a computation-driven field. As methods become more sophisticated, compute-heavy, and data-hungry, efforts to reproduce them become more difficult. While the computational MRI research community has made great leaps toward reproducible computational science, there are few tailored guidelines or standards for users to follow. In this review article, we develop a cookbook to facilitate reproducible research for MRI acquisition and reconstruction. Like any good cookbook, we list several recipes, each providing a basic standard on how to make computational MRI research reproducible. And like cooking, we show example flavours where reproducibility may fail due to under-specification. We structure the article, so that the cookbook itself serves as an example of reproducible research by providing sequence and reconstruction definitions as well as data to reproduce the experimental results in the figures. We also propose a community-driven effort to compile an evolving list of best practices for making computational MRI research reproducible. 

Although open databases are an important resource in the current deep learning (DL) era, they are sometimes used “off label”: Data published for one task are used to train algorithms for a different one. This work aims to highlight that this common practice may lead to biased, overly optimistic results. We demonstrate this phenomenon for inverse problem solvers and show how their biased performance stems from hidden data-processing pipelines. We describe two processing pipelines typical of open-access databases and study their effects on three well-established algorithms developed for MRI reconstruction: compressed sensing, dictionary learning, and DL. Our results demonstrate that all these algorithms yield systematically biased results when they are naively trained on seemingly appropriate data: The normalized rms error improves consistently with the extent of data processing, showing an artificial improvement of 25 to 48% in some cases. Because this phenomenon is not widely known, biased results sometimes are published as state of the art; we refer to that as implicit “data crimes.” This work hence aims to raise awareness regarding naive off-label usage of big data and reveal the vulnerability of modern inverse problem solvers to the resulting bias.