Abstract
PurposeIn the present work, we describe the correction of diffusion‐weighted MRI for site and scanner biases using a novel method based on invariant representation.Theory and MethodsPooled imaging data from multiple sources are subject to variation between the sources. Correcting for these biases has become very important as imaging studies increase in size and multi‐site cases become more common. We propose learning an intermediate representation invariant to site/protocol variables, a technique adapted from information theory‐based algorithmic fairness; by leveraging the data processing inequality, such a representation can then be used to create an image reconstruction that is uninformative of its original source, yet still faithful to underlying structures. To implement this, we use a deep learning method based on variational auto‐encoders (VAE) to construct scanner invariant encodings of the imaging data.ResultsTo evaluate our method, we use training data from the 2018 MICCAI Computational Diffusion MRI (CDMRI) Challenge Harmonization dataset. Our proposed method shows improvements on independent test data relative to a recently published baseline method on each subtask, mapping data from three different scanning contexts to and from one separate target scanning context.ConclusionsAs imaging studies continue to grow, the use of pooled multi‐site imaging will similarly increase. Invariant representation presents a strong candidate for the harmonization of these data.
Highlights
Observational conditions may vary strongly within a medical imaging study
In the present work we focus on harmonization for diffusion magnetic resonance imaging (MRI), a modality known to have scanner/site biases[8,9,10,11,12,13,14,15,16] as well as several extra possible degrees of freedom with respect to protocol
We propose that a subset of harmonization solutions may be found by learning scanner invariant representations, that is, representations of the images that are uninformative of which scanner the images were collected on
Summary
Observational conditions may vary strongly within a medical imaging study. Researchers are often aware of these conditions (eg, scanner, site, technician, facility) but are unable to modify the experimental design to compensate, due to cost or geographic necessity. In magnetic resonance imaging (MRI), variations in scanner characteristics such as the magnetic field strength, scanner vendor, receiver coil hardware, applied gradient fields, or primary image reconstruction methods may have strong effects on collected data[1,2,3]; multi-site studies in particular are subject to these effects.[4,5,6,7] Data harmonization is the process of removing or compensating for this unwanted variation through post hoc corrections. While all methods require paired scans to correctly validate their results (subjects or phantoms scanned on both target and reference scanners), supervised methods require paired training data. The collection of such data is expensive and difficult to collect at a large scale
Sign-in/Register to view highlights & summary Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
