17-03-2014, 08:53 PM
Abstract-The foot complications constitute a tremendous
challenge for diabetic patients, caregivers, and the healthcare
system. With current technology, in-shoe monitoring systems can
be implemented to continuously monitor foot's at-risk ulceration
sites and send feedback to patients and physicians. The few
available high resolution in-shoe pressure measuring systems are
extremely expensive and targeting clinical use only. The more
affordable price ranges can be reached by limiting the number
of sensors in the shoe. Precise subject-specific sensor placement
is still a challenge in such platforms. Moreover, there is no good
way to estimate pressure on other points of the foot. In this
paper, we address these technical challenges by proposing SCPM
algorithm that reconstructs a continuous foot plantar pressure
image from a sparse set of sensor readings. Using our technique,
sensor placement can be the same in every electronic insole.
However, the SCPM's trained parameters are unique for every
subject and foot.
Index Terms-Continuous Image Reconstruction, Diabetic
Foot Ulcers, In-shoe Pressure Monitoring, Plantar Pressure
Modeling.
I. INTRODUCTION
A. Motivation
According to the American Diabetes Association (ADA),
about 25.8 million people in the United States suffer from
diabetes [1]. Up to 25% of diabetic individuals will develop
a foot ulcer during their lifetime [2]. Of these people, 12%
to 24% eventually must undergo amputation as a result of
infection due to the foot ulcer [3]. Any reduction in the rate of
diabetic foot complications would be significant to healthcare
providers, and more importantly, would improve the quality
of life for many individuals.
Diabetes over time can damage the nervous system and
cause neuropathy. A patient with neuropathy is not able to
feel his or her feet properly to allow the traumatized foot
to recover [4]. With current technology, electronic pressure
monitoring systems can be placed as insoles into normal shoes
to continuously monitor at-risk ulceration sites of foot and give
early warning/feedback to patients and physicians.
B. Related Works
Several systems for measuring high resolution plantar pressure
of foot are commercially available. Among these are
the Pedar [5], F-Scan [6] and Parotec [7]. These systems are
extremely expensive (order of $10k-$20k in 2012 models) and
mostly aim at athletic foot/running or clinical analysis.
For such a system to be affordable to regular patients, it
would need to be at a much lower price. This price point
can be reached primarily by limiting the number of sensors in
the shoes. An early foot-to-ground force measurement device
was reported by Spolek and Lippert in 1976 [8]. The system
was restricted to measure heel and toe forces during several
steps. Wertsch et. al. presented a portable plantar pressure
measurement system with seven sensors located under high
risk area of foot [9]. The more recent studies aimed at more
in-situ processing ability by using a computing device/gateway
attached to the waist [10-12].
In all of these system, a limited number of sensors are
mounted on a few regions of foot plantar area. These regions
are selected and fixed based on either the medically-predefined
high risk ulceration sites of foot [12], or foot pressure image
obtained from a pressure indicating film while subject standing
on it [10]. The main draw back of these systems is except
for a few exact points under the sensors, pressure distribution
information on other points of plantar area is completely lost.
Moreover, plantar pressure distribution based on foot shapes
and existing deformities is very subject-specific. So, sensor
placements should be carefully adjusted for every patients, and
even for a given subject, exact sensor placement is critical to
capture accurate data. Therefore, finding sensor placements
remained as a research problem and did not become practical
for widely-used reasonable-priced electronic insoles.
Plantar pressure modeling can be used as a method to
estimate the foot pressure distribution all over plantar area
for each individual by using limited number of parameters.
Foot modeling in diabetes research are mainly offered in two
categories. First, Finite Element (FE) approaches to model
foot mechanical structure and tissue characteristics [13]. These
models can be used to predict the effect of accommodative
in-shoe orthoses or insoles. The complexity of this approach
prevents its real-time applications. Second, various mathematical
modeling methods to extract relevant features for different
classification purposes. Authors in [14] applied Principal
Component Analysis (PCA) on dynamic foot pressure image
to extract eigenvalues as features and used a Fuzzy classifier
to distinguish normal and diabetic subjects. In this line of
research, models are not used to predict more detailed data,
but only used for classification.
challenge for diabetic patients, caregivers, and the healthcare
system. With current technology, in-shoe monitoring systems can
be implemented to continuously monitor foot's at-risk ulceration
sites and send feedback to patients and physicians. The few
available high resolution in-shoe pressure measuring systems are
extremely expensive and targeting clinical use only. The more
affordable price ranges can be reached by limiting the number
of sensors in the shoe. Precise subject-specific sensor placement
is still a challenge in such platforms. Moreover, there is no good
way to estimate pressure on other points of the foot. In this
paper, we address these technical challenges by proposing SCPM
algorithm that reconstructs a continuous foot plantar pressure
image from a sparse set of sensor readings. Using our technique,
sensor placement can be the same in every electronic insole.
However, the SCPM's trained parameters are unique for every
subject and foot.
Index Terms-Continuous Image Reconstruction, Diabetic
Foot Ulcers, In-shoe Pressure Monitoring, Plantar Pressure
Modeling.
I. INTRODUCTION
A. Motivation
According to the American Diabetes Association (ADA),
about 25.8 million people in the United States suffer from
diabetes [1]. Up to 25% of diabetic individuals will develop
a foot ulcer during their lifetime [2]. Of these people, 12%
to 24% eventually must undergo amputation as a result of
infection due to the foot ulcer [3]. Any reduction in the rate of
diabetic foot complications would be significant to healthcare
providers, and more importantly, would improve the quality
of life for many individuals.
Diabetes over time can damage the nervous system and
cause neuropathy. A patient with neuropathy is not able to
feel his or her feet properly to allow the traumatized foot
to recover [4]. With current technology, electronic pressure
monitoring systems can be placed as insoles into normal shoes
to continuously monitor at-risk ulceration sites of foot and give
early warning/feedback to patients and physicians.
B. Related Works
Several systems for measuring high resolution plantar pressure
of foot are commercially available. Among these are
the Pedar [5], F-Scan [6] and Parotec [7]. These systems are
extremely expensive (order of $10k-$20k in 2012 models) and
mostly aim at athletic foot/running or clinical analysis.
For such a system to be affordable to regular patients, it
would need to be at a much lower price. This price point
can be reached primarily by limiting the number of sensors in
the shoes. An early foot-to-ground force measurement device
was reported by Spolek and Lippert in 1976 [8]. The system
was restricted to measure heel and toe forces during several
steps. Wertsch et. al. presented a portable plantar pressure
measurement system with seven sensors located under high
risk area of foot [9]. The more recent studies aimed at more
in-situ processing ability by using a computing device/gateway
attached to the waist [10-12].
In all of these system, a limited number of sensors are
mounted on a few regions of foot plantar area. These regions
are selected and fixed based on either the medically-predefined
high risk ulceration sites of foot [12], or foot pressure image
obtained from a pressure indicating film while subject standing
on it [10]. The main draw back of these systems is except
for a few exact points under the sensors, pressure distribution
information on other points of plantar area is completely lost.
Moreover, plantar pressure distribution based on foot shapes
and existing deformities is very subject-specific. So, sensor
placements should be carefully adjusted for every patients, and
even for a given subject, exact sensor placement is critical to
capture accurate data. Therefore, finding sensor placements
remained as a research problem and did not become practical
for widely-used reasonable-priced electronic insoles.
Plantar pressure modeling can be used as a method to
estimate the foot pressure distribution all over plantar area
for each individual by using limited number of parameters.
Foot modeling in diabetes research are mainly offered in two
categories. First, Finite Element (FE) approaches to model
foot mechanical structure and tissue characteristics [13]. These
models can be used to predict the effect of accommodative
in-shoe orthoses or insoles. The complexity of this approach
prevents its real-time applications. Second, various mathematical
modeling methods to extract relevant features for different
classification purposes. Authors in [14] applied Principal
Component Analysis (PCA) on dynamic foot pressure image
to extract eigenvalues as features and used a Fuzzy classifier
to distinguish normal and diabetic subjects. In this line of
research, models are not used to predict more detailed data,
but only used for classification.