Accurate multiple sclerosis atrophy measurement system

With the help of Magnetic resonance imaging (MRI) shrinkage of the brain can be
monitored. This shrinkage is usually referred to as atrophy and has been
observed to be 10 times faster in multiple sclerosis (MS) compared to normal
aging of the brain. The responsible processes for this decline have not fully
been identified. However, several treatments for MS have recently shown to
reduce brain atrophy rates. Although this seems promising, atrophy rates are
rarely is used for the decision of treatment for patients. One of the reasons
for this is that systematic differences between scanners arise, possible
leading up to a 10% different measured volume on different scanners. Therefore,
current volume measurement software cannot account for these differences and
the measurements of the brain volume could be unreliable. We will develop a
measurement system to remove differences between scanners, thus creating a
reliable method to assess brain atrophy with the help of conventional MRI.

The main objective of this project is to validate a standardized volume
measurement system (using eight in-house 3D printed phantoms) to allow for
reliable atrophy measurements in MS, by quantifying between-scanner and
within-scanner reproducibility in volume measurements. Secondly, we will
quantify between-scanner and within-scanner reproducibility in a longitudinal
setup.

The method we are going to test uses so-called *phantoms*. Using 3D printing
and precision mechanical techniques, we simulate brains with different severity
of brain shrinkage. We use this as a standard (reference) to be able to align
(standardize) the measurements in patients between different scanners and
institutions. Such objects whose properties are known and which are used for
reference are called *phantoms* in the MRI field. Our phantoms consist of
compartments that are very precise throughout and mimic the cortex, the
ventricles (chambers of cerebrospinal fluid), and white matter, ensuring that
we know the volumes of all compartments very precisely. In a small pilot we
made a first prototype of such a phantom. It turns out that we can produce the
phantoms with this technology and that they are imagable on the MR scanner.

In this project, we will create a set of eight phantoms and use them to
standardize volume measurements in 30 MS patients and 10 healthy subjects.
These participants are all scanned on the same day on three different MRI
scanners as much as possible. If that does not fit into the schedule, these
three scans can also be spread over several days. We ask these participants
back after 2 years ± 5 months for the same scans, to investigate the effect of
the standardization on the measurement of volume change. No invasive procedures
are performed or substances are administered. The volume measurements are
standardized retrospectively, based on additional MRI measurements performed
without the volunteer being present.

With the following scan procedure:
1. Scan and re-scan on scanner number 1.
2. Scan and re-scan on scanner number 2.
3. Scan and re-scan on scanner number 3.
Each pair of scan or re-scan takes approximately 50 minutes.

The scheduling of the various scans is done in consultation with the
participant and based on the availability of the MRI scanners for this
examination. The scan may, in consultation with the participant, be made on the
same day or spread over a few days. Analysis: Intra- and inter-scanner
reproducibility will be quantified using the median absolute difference as well
as limits of agreement based on linear mixed model analyzes. These analyzes
will be performed for both the corrected and uncorrected volumes measured with
state-of-the-art methods: SIENA (X), FIRST and FreeSurfer.

Althoughthere are no immediate risks or side effects, and the patient undergoes
the same clinical scan that he or she would otherwise have received, there is a
significant burden. This is because all patients and healthy subjects undergo
MRI scans for a total duration of 3 hours on one day. In the meantime, the
movement from one scanner to another still has to take place (not all scanners
are located in the same building section). All in all, this is a considerable
burden, in addition to the usual inconveniences of undergoing an MRI scan, such
as the noise and having to stay still in a narrow tunnel. We will explicitly
inform potential participants about this, in order to prevent someone from
starting the research insufficiently prepared and encountering unpleasant
surprises. Because no contrast agent is administered, there is no additional
burden on the MRI scans. The phantoms are scanned without the patient being
present, so this also does not play a role. We will clearly advise participants
about all possibly burdensome aspects, to avoid exposing anyone to this burden
unknowingly. In a previous project, we have used a similar set-up to compare
three scanners in 21 MS patients. The patients in that study indicated that
although long, the burden was tolerable. No immediate benefits are expected for
the patients.