193 / 236
navigation of endoscopic and ablative instruments. Accu-
racy was universally judged to be equal or superior to
the current OR standards and sufficient for the applica-
tions provided. The main benefit over a conventional
system was the speed at which navigation assistance
could be provided and interpreted during ablative tasks.
Task Workload
The NASA-TLX scores for mental demand, effort,
and frustration were significantly reduced when using
the LIVE-IGS system in comparison to conventional nav-
igation (
P
<
.05). There was no significant difference in
physical demand or perceived performance. Despite open
feedback suggesting a potential operative time saving
during real cases, no significant change in temporal
demand was found during this trial (
P
5
0.19; Table I).
Questionnaire
The seven-point Likert scale questionnaire state-
ments and median (IQR) responses are shown in Table
II. No subject disagreed (score 1–3) with any of the
statements. One gave a neutral response (score of 4) for
question 1. There was universal agreement (score 5–7)
for all other questions, with fairly uniform responses
across the subjects.
Below is a summary of the feedback gathered for
each theme of investigation. Specific responses from
each subject are summarized in Supplementary Appen-
dix 1.
Ergonomics
All subjects agreed that the laboratory layout and
equipment were consistent with the OR. Minor changes,
including mounting the reflective markers to face
obliquely opposite the surgeon, were made to minimize
optical tracking interference (Fig. 2d). Occasionally, the
drill dropped out of vision when it was rotated; however, all
of the surgeons were familiar with optical IGS systems and
seemed to intuitively recognize when this was a problem.
Visual Display
Image guidance was provided in two ways: a 3D vir-
tual view and cross-sectional, triplanar CT images.
Three participants preferentially referenced the virtual
view, stating that it was intuitive, allowed faster assess-
ment of proximity to critical structures, and was easier
to synthesize. Two preferred the triplanar views, as they
were more often used to this display, and thought the
virtual view was cluttered or lacked depth information
and precision. The other two used both displays fairly
equally; the virtual for a quick assessment and the tri-
planar for fine detail.
The contours were thought to be accurate, and
there were mixed opinions as to whether the pixilation
TABLE I.
Task Workload Assessment.
NASA-TLX Subscale
NASA-TLX Scores
Conventional Mean (SD)
LIVE-IGS Mean (SD)
P
, Wilcoxon SR Test
Mental Demand
10.6 (4.9)
6.9 (3.3)
.006*
Physical Demand
9.3 (5.6)
7.4 (3.1)
.094
Temporal Demand
6.6 (4.5)
5.1 (2.4)
.19
Performance
5.4 (3.8)
4.6 (2.8)
.496
Effort
9.8 (5.0)
6.1 (3.2)
.011*
Frustration
7.7 (5.5)
4.6 (2.7)
.032*
Lower performance score indicates higher perceived performance.
*Significant improvement with LIVE-IGS (
P
<
.05).
LIVE-IGS
5
localized intraoperative virtual endoscopy image-guided surgery; NASA-TLX
5
National Aeronautics and Space Administration Task Load
Index; SD
5
standard deviation; SR
5
signed rank.
TABLE II.
Questionnaire Responses.
Statements for Questionnaire
Median (IQR)*
I felt it was faster to perform surgery when aided by the virtual view.
6 (5–6)
The system appeared to be sufficiently accurate for its intended use.
6 (6–6)
The dynamic tool tracking allowed me to quickly assess my proximity to critical
structures without significantly interrupting dissection.
6 (5.5–7)
Proximity alerts increased my confidence during ablation close to critical structures.
6 (5.5–6)
The current technology is ready for clinical trial without significant changes.
5 (5–6)
*Based on a seven-point Likert scale (7
5
strongly agree, 1
5
strongly disagree).
IQR
5
interquartile range.
Laryngoscope 124: April 2014
Dixon et al.: Real-Time Navigation for Endoscopic Surgery
171