Fully Automated Scan Technique: Optimisation of Scan Timing in chest CT [FAST START trial]

Computed Tomography Angiography (CTA) is a non-invasive imaging tool widely
used for various indications. Contrast media (CM) is used to enhance the
intravascular lumen and organ parenchyma, depending on the indication. Recent
technical advances in CT scan techniques allowed for a very fast scan
acquisition with substantially increased image quality in terms of temporal and
spatial resolution. These faster scan times account for a significant reduction
in radiation dose, which is desirable in light of the *As Low As Reasonably
Achievable* (ALARA) principle. Another advantage of the newer *high-end*
scanners is the use of lower tube voltages, since many studies have shown that
CM volumes can be reduced with usage of lower tube voltages.
However, with faster scan acquisition, challenges arise with regard to CM bolus
timing. The risk of outrunning the CM bolus in these fast acquisitions is
higher, subsequently leading to a decreased or even non-diagnostic enhancement
(in Hounsfield Units (HU)). In addition, decreased CM volumes due to usage of
lower tube voltages also add to the risk of outrunning the bolus. When reducing
the CM bolus, the injection time decreases and the window of peak enhancement
is shorter and more narrow. Also, when injecting these smaller CM volumes at
higher flow rates, although the peak enhancement is increased, the window of
peak enhancement decreases more rapidly. Thus, when administered with the same
flow rate, the peak of the enhancement curve will be lower, narrower and faster
compared to larger CM volumes. This, in combination with the faster scan
acquisition makes the timing of the start of the scan (scan start delay) highly
important, since scanning at the peak enhancement is necessary to achieve a
diagnostic image quality.
To determine scan delay, two techniques that are used frequently in daily
clinical routine are the *bolus tracking* and *test bolus* technique. With the
latter, a smaller CM bolus is administered before the actual scan, and the time
to peak of the intravascular enhancement is determined with help of dedicated
software (DynEva, Siemens Healthineers, Forchheim, Germany). When using the
*bolus tracking* technique, no additional CM volume is administered. A region
of interest (ROI) is placed in a large artery of interest (e.g. ascending or
descending aorta), and a threshold enhancement is set prior to the scan (e.g.
100 HU). Repetitive low dose helical scans are acquired at the same level and
the arrival of the CM bolus is followed. Once the threshold is reached, the
scanner automatically starts the scan. Between reaching the threshold and the
actual start of the scan, a manual post-tracking delay is set prior to
scanning. This delay is necessary for both the table movement of the scanner to
the starting point of the scan and the breath hold command. The problem is that
this manual post-tracking delay is set prior to the scan, without information
of the patient*s cardiovascular dynamics (e.g. cardiac output). Since cardiac
output can vary greatly inter- and intra-patient, this fixed post-tracking
delay may not be appropriate for all patients. Scanning with a sub-optimal post
tracking delay could potentially result in suboptimal arterial enhancement and
insufficient diagnostic quality.

With new bolus tracking auto-delay software (Fully Automated Scan Technique,
FAST, Siemens Healthineers) the incidence of scans made at a suboptimal
attenuation could be reduced. This software works the same as the *bolus
tracking* technique, the difference is that the manual post-tracking delay is
calculated automatically by the software7. During the low-dose helical scans at
the level of the ROI, the attenuation in the ROI is used to predict the optimal
enhancement curve. The software takes the injection protocol, tube voltage and
patient parameters into account. A previously acquired database of numerous
enhancement curves is consulted to predict this enhancement curve of the
individual patient. The software then calculates the optimal post-tracking scan
delay to scan at the peak enhancement. Thus, the optimal individual scan delay
and enhancement, based on the patients physiology can be achieved, and the risk
of non-diagnostic scans should decrease. Therefore, this study aims to evaluate
the performance of the FAST software in patients receiving standard chest CT
with regard to the number of non-diagnostic scans (< 300 HU) and compare this
with standard care (manual set pre-scan delay).

Primary objective:
- To evaluate the performance of the FAST software in patients receiving a
chest CT with regard to the number of non-diagnostic scans (> 300 HU) in
comparison with standard care ('default' delay).

Secondary objectives:
- To assess the enhancement curses calculated by the FAST software with regards
to scan timing and delay and compare it with the scan timing and delay of the
control group ('default' delay)
- To assess the objective (intravascular attenuation, image noise,
signal-to-noise ratio, and contrast-to-noise ratio) image quality parameters in
patients receiving standard chest CT with the FAST software and compare it with
the control group ('default' delay).
- To assess the subjective (Likert scale) image quality parameters in patients
receiving standard chest CT with the FAST software and compare it with the
control group ('default' delay).
- Calculation of CO with help of the testbolus attenuation curves.

This study is an observer blinded randomized controlled trial conducted
according to Guidelines of GCP. This prospective study will assess the
performance of the FAST software in patients receiving standard chest CT.
We are aiming to prospectively enrol 316 consecutive patients, based on a
sample size calculation (see 4.4; page 13), who will be referred for a chest
CT. The inclusion period will be two years. Standard chest CT consists of an
contrast-enhanced CT scan of the chest, performed with the bolus tracking
technique. An additional test bolus will be added for quality assurance, no
other changes will be made in the standard scan protocol. All patients who are
referred for a standard chest CT will be eligible for inclusion. Patients will
be enrolled in one of two groups, according to the software used. Group 1: FAST
group, scanned with the new FAST software; Group 2: Control group, scanned
with the a default delay set prior to scanning. No adjustments will be made in
the standard scan settings used in daily practice, only scan delay will differ.

NVT

The participants will receive a scan on referral from their clinician and the
scan will be performed according to normal clinical routine. Only patients
already scheduled for clinically-mandated chest CT will be recruited. Scan
protocol will not be altered for this research. Differences in intravascular
attenuation may potentially impact the diagnostic accuracy of a chest CTA,
especially when the attenuation is below a diagnostic level. However, since the
scan delay calculated with the FAST software is more specific to patient
physiology, it is not expected that the attenuation will be lower compared to
the control group. Also, the FAST software has the possibility to manually
start the scan in case the technician feels the delay is too long. This can be
seen as an extra security measurement. An additional test bolus will be added
to asure the quality of the software. Lastly, the CM protocol in this study is
the standard injection protocol in our medical centre and is known to be
sufficient for clinical purposes.

Participation in this study will not cause any delay in the standard CTA
procedure. We therefore do not expect participation in this study to give any
disadvantages for the subjects relative to the standard CTA protocol, which
they would have undergone as part of their clinical care. The only
trial-related burdens will be the randomization and additional low-dose
low-volume test bolus.