When
a rehabilitation physician meets with a postacute stroke patient for
counseling and rehabilitation program planning, decisions are usually
based on two types of information: the results of a clinical examination
of functional capacity (i.e., what a person can do in a standardized,
controlled environment) and the patient’s subjective report on
limitations and problems in daily life. With this information, the
rehabilitation professional and the patient set specific goals together(The professional should never be involved in goal setting, they will dumb down the goals to make themselves not look bad.),
with the objective of improving functional performance (i.e., what a
person does in his daily life)(Oh my God, you are immediately forcing your tyranny of low expectations on your patients.100% recovery is the only goal in stroke.) [1].
An objective measurement of functional performance was not available
for a long time, but with the development of wearable sensors, it is now
increasingly used in rehabilitation. Wearable sensor technology allows
the collection of data that had previously been missing: the ‘objective
measurement of clinically important naturalistic behaviors' [2].
Ideally, information about performance would be available for the
planning and monitoring of a rehabilitation program and would include
several aspects, such as overall physical activity, walking behavior and
upper-limb use.
Studies involving wearable sensors generally
report low physical activity levels, low walking performance and little
use of the affected arm in daily life in stroke patients at the
population level [3, 4]. However, the variability of daily life performance measures among patients was considerable in most studies [5, 6]. Demographic or stroke-related variables did not [7] or only partially [6, 8,9,10] explain the performance variability.
Potential applications of sensor-derived performance measures in rehabilitation programs have been described by many authors [11, 12],
but we are not aware of any studies that examined the value of such
performance information in individual patients receiving clinical
rehabilitation. Additionally, with few exceptions [13, 14],
most studies that employed wearable sensors to measure daily life
performance in stroke patients focused on either upper or lower limbs.
However, the clinical situation of a patient initiating a rehabilitation
program would, in most cases, require a comprehensive assessment of
upper-limb activity, walking behavior and physical activity.
We
hypothesize that comprehensive, sensor-derived performance information
is clinically valuable for the planning of rehabilitation programs for
individual stroke patients who live at home. Performance data are deemed
clinically valuable if they can be used as decision aids for
therapeutic management or for counseling in individual patients [15]. We explore the clinical value in a narrative style, with a focus on individual patient performance and capacity data.
Daily
life performance was recorded with a series of wearable sensors placed
on the upper and lower extremities and the trunk. The wearable sensors
were placed on the patient in the clinic by a clinical scientist as
suggested by others [16]
because the handling and placement of the wearable sensors was judged
too complicated to be done independently by the stroke patients.
Recordings were initiated during a routine medical consultation in the
morning and lasted until late afternoon of the same day. We intended to
measure performance under a scenario that is feasible in routine
clinical practice. Therefore, repeatedly visiting patients over several
days to help with sensor handling (e.g., for undoing/redoing or charging
of sensor modules) was not an option, considering the time and cost
constraints in most healthcare systems. On these grounds, a longer
recording period was not an option.
Masahiro Ishiwatari
Kaoru Honaga
Akira Tanuma1, 
