Assessment of Gait Symmetry and Gait Normality Using Inertial Sensors: In-Lab and In-Situ Evaluation - Biomedical Engineering Systems and Technologies - page 227

for speed. However, in Part I of this study symmetry and normality measures are shown
to correlate to joint-angle curves normalized to stride time, not containing any velocity
information. The normality index is also normalized to stride time and as such is inde-
pendent of walking cadence. It is expected that the normality index would differentiate
between normal and abnormal patterns at the same speed. Further investigations are
needed to support this assumption.
Another factor supporting the usefulness of the normality index is its correlation
with the type of walking aid used during the test, Figure 8. The test-retest reliability
and discriminatory power of the index were also satisfactory. Overall, the proposed
index can possibly be developed into a reliable and clinically relevant measure of gait
normality.
Another interesting result was the correlation between improvement of normality and
number of days spent at the ward, Figure 13. Whereas there was no correlation between
improvement in normality and number of days between baseline and follow-up. This
possibly suggests that the rate of recovery at the ward is indicative of the total rate of
recovery, which is little affected by the recovery time at home. This assumption should
be further investigated.
Normality results and the answers to the EQ-5D TM questionnaire showed some pos-
itive trends. Greater discomfort and difficulties in performing usual activities seem to
be accompanied with worse normality, Figure 11. Besides the self-assessment question-
naire, the use of walking aids was also considered an indication of how well the patient's
health status was, i.e. patients who did not need any walking aid were, on average, in
better condition than those who used one crutch. Another indicator of recovery was the
number of days the patient spent at ward, assuming that patients who recovered better
or more quickly were discharged sooner. The normality index seems to be in agreement
with all the above mentioned qualitative health status assessments.
Symmetry results are difficult to judge due to the variety of walking aids used. The
large variety of symmetry at follow-up, Figure 5, was mostly influenced by the patients
using one crutch only. This could be explained by the fact that some patients were
more dependent on the crutch and consequently leaned more to one side. Whereas some
patients barely used the crutch for support.
Due to their recent surgery, patients were very uncomfortable during the baseline
measurements. It was important to keep the data collection as simple and quick as pos-
sible. No more than five minutes had to be spared by the patient to complete the entire
procedure, and they were all willing to participate in the study. Briefness is also impor-
tant for the staff responsible for the procedure so that the addition of GA is not an extra
burden. The placement of the sensors was also quick and easy. However, in the future,
the waist sensor should be placed on the lower back so as not to be affected by subjects'
different shapes and sizes.
Another issue with the present study is that the number of participants was very
small. Any statistical inference on the results is greatly affected by the sample size.
However, results are promising and suggest that a larger study will likely produce pos-
itive results.
At the ward where the data was collected, gait analysis is not normally used, and
most records are based on rough qualitative descriptions. This lack of quantitative