DIABETES COMPLICATIONS

Development of a New Tool for Obliterating Arteriopathy of the Lower Limbs

Detection

Stéphane Roeslin

, Nadège Marthouret

, Louis Benazet

, Christophe Roncato

, Gabriel Camelot

Anca Loppinet

, Xavier Racadot

, Sylvie Grandperret

, Hayet Bourezane

, Christian Pieralli

Lionel Pazart

and Bruno Wacogne

1,2

FEMTO-ST Institute, UMR 6174, 16 Route de Gray, 25030 Besançon cedex, France

INSERM CIT 808, Besançon University Hospital, Place Saint Jacques, 25030 Besançon cedex, France

Chirurgie Vasculaire et Médecine Vasculaire, CHU Minjoz, Bd Fleming, 25000 Besançon, France

Diabétologie et Endocrinologie, CHU Minjoz, Bd Fleming, 25000 Besançon, France

Keywords: Diabetes complications, Medical device, Obliterating arteriopathy of the lower limbs, Point of care,

Photoplethysmography, Signal processing.

Abstract: Obliterating arteriopathy of the lower limbs (OALL) is a common complication in diabetes. This

vasculopathy, which is associated with mild injury and with the diabetic neuropathy, is the source of

diabetic foot ulcers which precede approximately 85% of amputations. Simple measures may avoid this

dreaded complication if it is identified in time. OALL detection is currently undertaken by measuring ankle

systolic pressure. The latter could be evaluated with microcirculatory technique but these techniques have a

number of limitations: time consumption and cost. OALL detection is therefore limited to a small number of

specialized units. In order to allow detection of OALL in ambulatory medicine, we propose a simple system

based on photoplethysmography. The idea is to apply a pre-set "warning" pressure to the patient's toe and to

optically check if arterial pulsation still exists. If not, the patient is directed to the adequate hospital unit for

full diagnosis. This "warning" system which can easily be used at the general practitioner’s office is meant

to help detecting the OALL at an early stage, hence reducing the number of amputations. In this position

paper, we present the system, some early results and we propose a discussion concerning the screening of

OALL.

1 INTRODUCTION

Obliterating arteriopathy of the lower limbs (OALL)

is a common complication in diabetes (between 17

and 21% of the diabetic population between 50 and

75 years old) (HAS, 2006). This vasculopathy,

which is associated with mild injury and with the

diabetic neuropathy, is the source of diabetic foot

ulcers which precede approximately 85% of

amputations (VALMI, 2008). The prevalence of

amputations would be 1.3% of diabetic patients.

Simple measures may avoid this dreaded

complication if it is identified in time (Boccalon,

2004 – Girach, 2006), and one of the five-year

objectives fixed by the Saint Vincent European

Declaration is to reduce the rate of foot amputations

among diabetic patients by 40%.

Given the extent of this issue and the objectives

fixed, Alfediam, followed by HAS (French Health

High Authority), recommend OALL screening

among patients over 40 years old or having suffered

from diabetes for over 20 years.

This detection is currently made by measuring

ankle Systolic Pressure Index (SPI). This

examination consists in measuring the humeral

systolic pressure and the ankle systolic pressure. The

SPI is simply given by the ratio of these two

measurements.

These ankle examinations, however, are easily

disturbed due to the common existence of

mediacalcosis among diabetic patients (prevalence

Roeslin S., Marthouret N., Benazet L., Roncato C., Camelot G., Loppinet A., Racadot X., Grandperret S., Bourezane H., Pieralli C., Pazart L. and

Wacogne B..

DIABETES COMPLICATIONS - Development of a New Tool for Obliterating Arteriopathy of the Lower Limbs Detection.

DOI: 10.5220/0003703700930098

In Proceedings of the International Conference on Biomedical Electronics and Devices (BIODEVICES-2012), pages 93-98

ISBN: 978-989-8425-91-1

 2012 SCITEPRESS (Science and Technology Publications, Lda.)

between 17 and 24% of diabetics). SPI could be

evaluated with microcirculatory techniques (TcPO2

and Laser Doppler) which usually offer a reliable

insight into the quality of distal vascularisation

(Becker, 1989 – De Graff, 2003). But, these

techniques have a number of limitations: time limits

(45 mins) undervalue (15.01 euros) requiring a

considerable initial investment (42000 € for a

PERIMED brand microcirculation unit), precise

interpretation requiring specialist expertise, and

numerous limitations in the process itself (infection,

oedema, general haemodynamic, etc). This situation

has confined its use to rare specialist centres which

are almost entirely based in hospitals, while these

examinations have become essential for classifying

and managing arteriopathy in diabetic patients.

Ankle SPI can be extremely perturbed by the

presence of a mediacalcosis which leads to an over

estimation of the SPI. In this case toe SPI can be

used together with a clinical and Doppler

examination (Moe, 2002 – Kröger, 2003 – Carter,

1996). The PERIMED system can be used for SPI

measurement. However, due to its cost, its use is

limited to a small number of specialized centres.

Recently, the SysToe

system has been launched by

Atys Medical. This equipment, based on

photoplethysmography, is much more affordable

(2500 €) but still too expensive for being used on a

routine basis in the physicians office. Indeed, we

recall that the number of amputations could

drastically be reduced with a large scale screening of

OALL.

Figure 1: Description of the device.

In order to allow detection of OALL in

ambulatory medicine, we propose a simple system

based on photoplethysmography. The new concept

consists in applying a pre-set "warning" pressure to

the patient's toe and to optically check whether

arterial blood circulation is observed or not. If not,

the patient is directed to the adequate hospital unit in

order to undergo specialized diagnosis. This

"warning" system which can easily be used at the

general practitioner’s office is meant to help

detecting the OALL at an early stage, hence

reducing the number of amputations.

2 DESCRIPTION OF THE

DEVICE

We recall that our system is a "warning" system

used to screen the possible existence of an OALL.

The principle consists in applying a "warning

pressure" to the patient's toe and to check whether or

not an arterial circulation is detected. A schematic

description of the device is shown in figure 1. It

consists of a pneumatic and an electronic part that

are connected together via a pressure sensor. First of

all, the physician places the toe cuff and the

photoplethysmographic sensor onto the patient's toe.

The signal delivered by the optical sensor is pre-

processed in the pre-processing unit (see next

section).

At that moment, the cuff inflator has not yet

been used and the pressure in the pneumatic circuit

is zero. When the pressure is zero, only the driving

of the yellow LED is enabled. It pulsates at the

patient's heart rate and acts as a sensor position

monitor. Indeed, when the yellow LED is pulsating

the physician knows whether the sensor is correctly

installed or not.

Now that the optical sensor is correctly installed,

the physician uses the cuff inflator in order to reach

the "warning pressure" value. Let us call it Ps. A

calibrated valve is used to adjust the pressure to the

"warning" value. When the pressure Ps is reached,

the pressure electronics disables the yellow LED and

enables the pulsation detection unit. (Arterial

pulsation is detected by comparing the signal to a

threshold value that has been determined with

healthy volunteers.) Now, if an arterial pulsation is

detected, the green LED is switched on; everything

is alright with the patient. Conversely, if no

pulsation is detected, the red LED is switched on

and the patient is addressed to a specialized hospital

unit for further diagnosis.

Note that for the moment, and according to the

very little literature, the "warning" pressure is set to

BIODEVICES 2012 - International Conference on Biomedical Electronics and Devices

70 mmHg (Boccalon, 2004 – De Graff, 2003 –

Johanson, 2002). The actual value will have to be

confirmed with clinical trials we are setting up at the

moment.

3 EXPERIMENTAL RESULTS

3.1 Signal Pre-processing and

Pulsation Detection

3.1.1 Signal Preprocessing

While reading the above mentioned description one

could think that everything is trivial in this system.

However, photopletysmographic sensors are used to

measure absorption of light into the tissue. In order

to detect arterial pulsation, we must be able to

measure the small absorption variation due to

arterial blood only. Therefore, the signal to noise

ratio at the direct output of the optical sensor is quite

low. Furthermore, the optical sensor also detects

variations of light that can be due to patient's

movements or changes in the ambient light.

A signal pre-processing is therefore required. In

what follows, we briefly describe the signal

processing we used. The rest of the electronic

circuitry is somehow more conventional and the

description of it may lengthen this paper.

Figure 2 shows the signal pre-processing we

developed.

Figure 2: Signal pre-processing.

Right after the optical sensor head, we use a high

impedance follower in order to reduce the influence

of noise. The signal obtained at the output of this

stage is shown in figure 3(a).

At that stage, the signal exhibits a slowly varying

offset due to patient's movements and variations in

the ambient light. A second order high pass filter

(cut-off @ 1 Hz) is used to cancel this varying

offset. The result is shown in figure 3(b).

Here, the signal is somehow independent of

external perturbations but is still quite noisy. The

noise is now rejected by means of the second order

low pass filter (cut-off @ 10 Hz).

Figure 3: Detail of the signal pre-processing. (a) at the

optical sensor and follower output. (b) at the high pass

filter output.

Finally, a ×100 gain amplifier is used and an

offset is added in order to centre the signal around 6

Volts, a value compatible with the rest of the

electronic circuit. Figure 4 summarizes the signal

pre-processing. Figure 4(a) shows the signal after

the follower while figure 4(b) shows the signal after

pre-processing. It can be noted that the signal issued

from the optical sensor looks different between

figure 3(a) and 4(a). This difference arises because

the sensor is not placed exactly the same way

between the two measurements. This is not a

problem because signal filtering changes the signal

shape (frequency filtering). At the end, the signal

always looks like figure 4(b) after signal processing.

This shows that the device is independent not only

of unwanted movements or light variations, but also

of placement inaccuracies.

A last amplification (not shown in the figure) is

used in order to set the signal to an amplitude

compatible with the rest of the electronic circuit.

3.1.2 Pulsation Detection

Pulsation detection consists in comparing the pre-

processed signal with a reference voltage. This

reference voltage was determined after tests

conducted with healthy volunteers. In figure 5(a), a

pressure of 60 mmHg was applied to the toe. With

this pressure, arterial pulsation still exists. In

particular, maxima of the signal are greater than the

reference voltage. In this case, the green LED is

switched on. Conversely, when the pressure is 100

DIABETES COMPLICATIONS - Development of a New Tool for Obliterating Arteriopathy of the Lower Limbs Detection

mmHg, the signal is lower than the reference

voltage. Here, the red LED is switched on.

Figure 4: Comparison between the signal issued from the

sensor (a), and the signal obtained after pre-processing (b).

Figure 5: LEDs driving. (a) Pressure=60 mmHg, signal

crossing the reference voltage, green LED is switched on.

(b) Pressure=100 mmHg, the signal not crossing the

reference voltage, red LED is switched on.

3.2 Experimental Device

The device is shown in figure 6.

Toe cuff, optical sensor and cuff inflator are not

shown in the figure. These three elements were

purchased from Hokanson:

• Infra-red photoplethysmographic sensor:

COPPHO

• Toe cuff: UPC2.5

• Cuff inflator: DS400

All the electronic circuits were made using

surface mounted components in order to reduce the

size of the device.

On figure 6, we can see that we used an external

power supply. Of course, this device is not the final

one. In the last version, batteries will be used instead

of the external power supply. Also, for this

intermediate version, SMA connectors have been

included in order to monitor different signals at

strategic points of the electronic circuit. These

measurements were used to define the reference

voltage mentioned above and to tune the offsets and

the gain of the amplifier to an appropriate value.

The device can be operated in two ways for

clinical trial purposes. We mentioned above that the

monitoring yellow LED is driven only when the

pressure is zero. Conversely, when the "warning"

pressure is applied, only the red and green LEDs are

driven. This configuration corresponds to the normal

used in ambulatory medicine.

Figure 6: Experimental device.

However, clinical trials imply various

experimental configurations. For the first clinical

trial (to be described in the next section) the pressure

sensor must be disconnected. Therefore, we also

design the electronic circuit so that this

configuration can be used. In this case, the three

LEDs are driven simultaneously.

3.3 First Clinical Trials

Clinical trials for this device can be separated in

BIODEVICES 2012 - International Conference on Biomedical Electronics and Devices

three parts.

In a first time, we have to check whether or not

the measurements made with our new device are

consistent with what is observed with the gold

standard technique. In a second time, the intra-

operator and inter-operator stability of our system

will be evaluated. These two first trials are

conducted at the Besançon University Hospital, with

the assistance of the cardio-vascular surgery unit and

the diabetology unit. These trials are performed with

patients suffering of various grade of OALL.

Finally, clinical trials in physician’s office with any

kind of patients will be done and the screening result

will be compared with diagnosis made at the

hospital.

Up to now, only the first step has been passed.

Patients were first tested with conventional Doppler

technique. Their toe systolic pressure was recorded

for both toes.

Then, they were tested with the new device. For

this, discrete pressure values were applied to their

toe according to the systolic pressure measured

before. The pressures ranged from 50 mmHg above

the systolic pressure down to 50 mmHg below with

10 mmHg steps. To do this, the pressure sensor in

our device was disabled as explained above. In order

to control the pressure values, we used the Hokanson

compressor (ref. AG101) and the corresponding

pressure regulator (ref. E20). Some results are

summarized in figure 7.

Figure 7: Some examples of measurements made during

the first clinical trial step.

The colours of the cells correspond to the colour

of the LED lighting during the test. For example, for

patient 12 right toe, the red LED was on from 150 to

110 mmHg and the green was on from 100 to 50

mmHg. If we compare to the systolic pressure

measured by means of the Doppler equipment, we

see that the behaviour of our device is correct

because the LEDs switch from red to green when the

pressure steps from 110 to 100 mmHg. This is

indicated by the “OK” appearing in the cells. The

same behaviour is observed with the left toe when

the pressure steps from 110 to 100 mmHg.

However, discrepancies are sometimes observed.

This is illustrated with patient 16 right toe were the

LED should have been red for 60 mmHg. This

discrepancy is indicated by the “NNNNN” in the

cell. Up to now, we are still analysing the complete

set of data (new device and Doppler records) in

order to fully understand this aspect.

4 SHORT DISCUSSION

We still have to finish the clinical trials before

deciding of the benefit of this method. Questions or

problems to be solved are:

• understanding the discrepancies that occur

sometimes

• investigating the possibility to use digital

electronics instead of analog technique

• defining an actual “warning pressure”

• investigating the intra or inter operability

• conducting the final clinical trials.

In the case of success with these aspects, EC

labelling is foreseen before any commercialization.

However, we recall that up to now, there is no

means of screening OALL. The only devices

commercially available concern the measurement of

the systolic pressure index (SPI) which is more a

diagnosis than a screening technique. Furthermore,

the price of these systems restricts their use in

specialized centres.

We think that detecting the presence or the

absence of arterial pulsation when a “warning

pressure” is applied to the patient is a potential

alternative to the current cost effective and time

consuming techniques. The market study we ordered

shows that a device price of about 250 € would be

accepted by most of the possible users. It is very

likely that the device we propose can be sold at this

low price, as long as it is fabricated at a scale large

enough.

5 CONCLUSIONS

In this position paper, we proposed a new device

DIABETES COMPLICATIONS - Development of a New Tool for Obliterating Arteriopathy of the Lower Limbs Detection

that could be used to screen OALL in ambulatory

medicine. It is meant to be used by general

practitioners but also by podiatrists. These

professionals have been identified by a market study

we recently ordered.

The principle is not to measure the systolic

pressure index (SPI), as it is commonly done in

specialized centres that can afford the expensive

equipment required. The new concept simply

consists in applying a “warning pressure” to the

patient’s toe and to check whether an arterial

pulsation is detected or not. If not, the patient is

directed to a specialized centre for complete

diagnosis.

First clinical trials show that our screening

device is consistent with what is observed with

conventional diagnosis techniques. However, full

clinical trial programme should be concluded before

deciding of the benefit of our method.

ACKNOWLEDGEMENTS

The authors would like to acknowledge the financial

support of OSEO, the CNRS and the programme

“maturation de projets innovants”.

REFERENCES

Becker, F. 1989, Classification Clinique et

hémodynamique des artériopathies oblitérantes des

members inférieurs, PhD Thesis, Burgondy University.

Boccalon, H., 2004, Artériopathies des membres inférieurs

chez le diabétique, La revue de médecine interne,

Vol.25, pp. S337-S338, doi:10.1016/j.revmed.2004.

10.2008.

Carter, S. A., Tate, R. B., 1996, Value of toe pulse waves

in addition to systolic pressures in the assessment of

the severity of peripheral disease and critical limb

ischemia, J. Vasc. Surg., Vol. 24, pp. 258-265.

De Graff, J. C., Ubbink, D. T., Legemate, D. A., Tijssen,

J. G. P., Jacobs, M., 2003, Evaluation of toe pressure

and transcutaneous oxygen measurements in

management of chronic critical leg ischemia: a

diagnostic randomized clinical trial, J. of Vasc. Surge.,

Vol. 38, pp. 528-234.

Girach, A., Vignati, L., 2006, Diabetic microvascular

complications – can the presence of one predict the

development of another ?, J. Diabetes and its compl.,

Vol.20, pp. 228-237.

HAS report, 2006, Prise en charge de l'arthériopathie

chronique oblitérante artéroscléreuse des members

inférieurs, http://www.has-sante.fr/portail/upload/docs

/application/pdf/AOMI_fiche.pdf

Johanson, K. E. A., Marklund, B. R. G., Fowelin, J. H. R.,

2002, Evaluation of a new screening method for

detecting peripheral arterial disease in primary health

care population of patients with biabetes mellitus,

Diab. Med., Vol. 19, pp. 307-310.

Kröger, K., Stewen, C., Santosa, F., Rudofsky, G., 2003,

Toe pressure measurement compared to ankle artery

pressure measurements, Angiology, Vol. 54, pp. 39-44.

Moe, S. M., O'Neill, K. D., Duan, D., Ahmed, S., Chen,

Stephen, N. X., Leapman, S. B., Fineberg, N.,

Kopecky, K., 2002, Medial artery calcification in

ESRD patients is associated with deposition of bone

matrix proteins, Kindney Int., Vol. 61, pp. 638–647.

VALMI, 2008, Veines, Artères, Lymphatiques,

Microcirculation, http://cemv.vascular-e-learning.net/

Valmi

BIODEVICES 2012 - International Conference on Biomedical Electronics and Devices