Data Acquisition, Conditioning and Processing System for a

Wearable-based Biostimulation

Leonardo Sestrem

, Raul Kaizer

, Jo

ao Gonc¸alves

, Paulo Leit

, Jo

ao Paulo Teixeira

Jos

e Lima

, Tiago Franco

and Jos

e Augusto Carvalho

Research Center in Digitalization and Intelligent Robotics (CeDRI), Instituto Polit

ecnico de Braganc¸a,

Campus de Santa Apol

onia, 5300-253 Braganc¸a, Portugal

Keywords:

Wearable Bioestimulation, Data Acquisition, Electromyography, Artiﬁcial Intelligence.

Abstract:

Data acquisition by electromyography, as well as the muscle stimulation, has become more accessible with

the new developments in the wearable technology and medicine. In fact, for treatments, games or sports, it

is possible to ﬁnd examples of the use of muscle signals to analyse speciﬁc aspects related, e.g., to disease,

injuries or movement impulses. However, these systems are usually expensive, does not integrate data ac-

quisition with the muscle stimulation and does not exhibit an adaptive control behaviour that consider the

pathology and the patient response. This paper presents a wearable system that integrates the signal acquisi-

tion and the electrostimulation using dry thin-ﬁlm titanium-based electrodes. The acquired data is transmitted

to a mobile application running on a smartphone by using Bluetooth Low Energy (BLE) technology, where it

is analysed by employing artiﬁcial intelligence algorithms to provide customised treatments for each patient

proﬁle and type of pathology, and taking into consideration the feedback of the acquired electromyography

signal. The acquired patient’s data is also stored in a secure cloud database to support the physician to analyse

and follow-up the clinical results from the rehabilitation process.

1 INTRODUCTION

In 2030, according to the World Health Organisation,

it is expected that the number of people aged 60 years

and older will be approximately 1.4 billion. With the

coming of age, movement difﬁculties and patholo-

gies, such as arthritis, become frequent, consequently

increasing the workload of the physiotherapists. As

these pathologies make difﬁcult, expensive and bur-

densome to constantly leave home to receive the treat-

ment, cheap and easy-to-use alternatives must be de-

veloped to support the home healthcare.

One of the rehabilitation treatment methods makes

use of the electrostimulation, namely the Functional

Electrical Stimulation (FES), which uses low fre-

https://orcid.org/0000-0002-9344-3075

https://orcid.org/0000-0001-9273-2257

https://orcid.org/0000-0002-3502-7444

https://orcid.org/0000-0002-2151-7944

https://orcid.org/0000-0002-6679-5702

https://orcid.org/0000-0001-7902-1207

https://orcid.org/0000-0001-8574-4380

https://orcid.org/0000-0002-6074-8112

quency electrical currents to provoke the muscles con-

traction, increasing the muscle activity, along with ex-

ercises, contributing signiﬁcantly to improve the abil-

ity to carry out functional activities. Aiming to sim-

plify the everyday activities and allow its remote ex-

ecution, the functions required to carry out the treat-

ment may be implemented in a wearable system.

Wearable systems can be used by patients to per-

form health treatments and the continuous monitoring

locally at home, specially not requiring other devices

like screens and peripherals. In a wearable system, the

collected muscle data by biosignals measurements,

e.g., cardiac, neurological and musculoskeletal, are

biological samples in time and space of biological

events. The musculoskeletal signal, or the myoelec-

tric signal, is a result of the bioelectricity phenomenon

that occurs when a potential difference exists between

the internal and external sides of the muscle cell mem-

brane (Ferraz et al., 2021).

The eletromyography (EMG) is the electrodiag-

nostic medicine technique used to monitor the electri-

cal activity produced by skeletal muscles (Robertson

et al., 2013). The usage of electrodes placed on the

skin to acquire this type of signal is known as surface

Sestrem, L., Kaizer, R., Gonçalves, J., Leitão, P., Teixeira, J., Lima, J., Franco, T. and Carvalho, J.

Data Acquisition, Conditioning and Processing System for a Wearable-based Biostimulation.

DOI: 10.5220/0011002300003123

In Proceedings of the 15th International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC 2022) - Volume 1: BIODEVICES, pages 223-230

ISBN: 978-989-758-552-4; ISSN: 2184-4305

223

electromyography (sEMG), and presents the advan-

tage of being a non-invasive EMG technology that is

capable of monitoring the nerve-muscle interactions

signals (Ergeneci et al., 2018a). The disadvantages of

sEMG are the complexity to obtain accurate data sig-

nals and the difﬁculty of electrostimulation, namely

the need of the exact positioning of the electrode, the

concern with noisy electrical environments and the

patient biological conditions (Rodrigues et al., 2020).

The commercial wearable systems usually present

the EMG signal acquisition or electrostimulation

functions separately, due to the complexity and dif-

ﬁculty to integrate without compromising the efﬁ-

ciency, signal accuracy and comfort. To address

these requirements, the employed hardware needs to

present a greater and reliable performance, increasing

the solution investment.

Having this in mind, the paper describes the devel-

opment of a wearable electroestimulation system to

support the remote and autonomous rehabilitation of

the vastus medialis muscle in elderly people. The de-

veloped solution uses dry thin-ﬁlm electrodes on the

skin surface to acquire the EMG signal, and consid-

ers the electrostimulation system integrated with the

signal acquisition and conditioning, stimulating the

muscle during a rehabilitation session according to an

adaptive control based on the biofeedback from the

EMG biosignals and artiﬁcial intelligence (AI) algo-

rithms. A continuous remote monitoring system al-

lows the patient to get the treatment sessions at home,

with the gathered biofeedback data being transmitted

through Bluetooth Low Energy (BLE) technology to

a smartphone which serves as mediator to the cloud,

where the physician can visualise the results of the

sessions and prescribe new sessions.

The rest of the paper is organised as follows: Sec-

tion 2 presents the related work and Section 3 de-

scribes the proposed architecture for the signal ac-

quisition, conditioning and processing system for the

wearable-based biostimulation. Section 4 presents

the implementation of the proposed system and Sec-

tion 5 analyses the preliminary experimental results.

Finally, Section 6 rounds up the paper with the con-

clusions and points out the future work.

2 RELATED WORK

The wearable electronic technology has revolu-

tionised the traditional medical diagnosis methods as

well as current medical devices by providing conve-

nient, remote, wearable and portable functions (Lou

et al., 2020). The progress in electrical measure-

ments and electronics has enabled the development

of wearable medical systems, which have real-time

monitoring capabilities, ﬂexibility, easy mobility, and

compatibility with the data and signal processing

ﬁelds (Chun et al., 2019; Trung and Lee, 2016).

There is a tendency to include extra functionali-

ties in the wearable medical systems aiming the in-

tegration with other systems, e.g., non-invasive real-

time monitoring and analysis of vital human param-

eters (Lou et al., 2020). Current solutions rely on

medical monitoring systems including wearable sen-

sors, wireless communication modules, displays, and

efﬁcient power supply for improved performance and

reliable data collection (Koh et al., 2016). The com-

bination of intelligent sensors and algorithms can not

only provide these extra functions to the system but

also offers processed data and independent work-

ing mode with self-healing, self-evaluation, and self-

calibration features (Lou et al., 2020).

Several commercial solutions for wearable sys-

tems based on sEMG data acquisition are currently

available, namely Biometrics DataLOG (Biometrics

Ltd, 2021), Bitalino (Da Silva et al., 2014) and Delsys

Trigno Lab (Delsys, 2021), providing a reliable and

accurate EMG signal acquisition, but they are usually

expensive and mainly focused in the signal acquisi-

tion functions. In the same manner, several academic

works are available for this purpose, e.g., (Park et al.,

2021; Zhu et al., 2021; Ergeneci et al., 2018a; Chen

et al., 2017; Li et al., 2021). In spite of their ben-

eﬁts, mainly focused in the signal acquisition, they

are missing important features for the development of

reusable wearable biostimulation based on the EMG

signal acquisition. As examples, the solutions pro-

posed by (Park et al., 2021; Biometrics Ltd, 2021)

present a lack of wearability that do not allow sufﬁ-

cient motion freedom and others do not use wireless

data transmission which are not suitable for general

use outside the laboratory environments (see (Park

et al., 2021; Chen et al., 2017)). In the same mannner,

e.g., the setup proposed in (Ergeneci et al., 2018a)

uses the Wi-Fi technology to stream the acquired data,

which in spite of its suitability for real-time monitor-

ing, the energy consumption is a drawback to wear-

able systems, being required a low power solution to

an extensive treatment period.

The use of wet electrodes may cause allergic reac-

tions and skin irritation (Fayyaz Shahandashti et al.,

2019), and the electrolytic conductive gel that helps to

reduce the skin-electrode impedance, dehydrate over

time, hence worsening the conductivity (Ergeneci

et al., 2018a). Another disadvantage of wet electrodes

is that they need to be changed every few or every ses-

sion while dry electrodes can be reused for many ses-

sions depending on the material that it is made (Peng

BIODEVICES 2022 - 15th International Conference on Biomedical Electronics and Devices

224

et al., 2016). Several works, e.g., (Zhu et al., 2021;

Park et al., 2021; Li et al., 2021), use wet/disposable

electrodes, which do not offer reusability.

Although dry electrodes do not require the need to

use conductive gel to work properly and are reusable,

they present higher skin-electrode impedance than the

wet ones, which may increase the biosignal noise and

demand higher voltage to stimulate the muscle. Some

solutions found in the literature employ dry sensors

that use high-cost materials in their composition such

as gold, offering a signal quality many times superior

to those obtained with the wet electrodes, but increas-

ing the price of the system (Ergeneci et al., 2018b).

The aforementioned solutions are usually based

on wearables that consider wet electrodes and only

present the acquisition function and do not include

stimulation to aid the muscle rehabilitation. Another

functionality that is usually missing in the stimulation

systems is the use of AI to assist the physician in the

diagnosis and even to adjust the treatment protocol

during an electrostimulation session according to the

pathology and the patient’s response.

3 BIOSTIMULATION SYSTEM

ARCHITECTURE

The solution proposed in this work was designed to

accomplish the requirements for the rehabilitation of

the vastus medialis muscle, particularly addressing

the elderly people.

3.1 System Requirements

The wearable healthcare systems need to be comfort-

able to wear, and an ease adaptation with skin or

human body surface, with compatibility, durability,

and abrasion resistance (Lou et al., 2020). In addi-

tion, these healthcare systems can not merely improve

the health status but also enhance the development of

medical technology allowing to detect an emergency

situation and preventing accidents (Park et al., 2021).

When designing the electrodes, the shape, size

and material are the ﬁrst aspects to take into ac-

count (Di Flumeri et al., 2019), inﬂuencing the price,

comfort, signal quality, contact stability and biocom-

patibility. The material must present unchanged char-

acteristics for long periods so the electrodes do not

have to be changed. This guarantees the right connec-

tion and easy use for elderly people in their homes.

The electrodes must also be designed in a way that

allows them to be washed together with the wearable

without having their position changed and avoiding

the risk of compromising the connection stability. To

avoid damages from the contact with sweat and other

body ﬂuids, the corrosion resistant materials must be

used but they must be materials that do not show sig-

niﬁcant changes in their electrical properties when

in contact with liquids (body ﬂuids) (Tallgren et al.,

2005) to not hurt the skin with the stimulation signals.

There is the concern with electric damaging the

skin while applying the FES system through dry elec-

trodes because of the body ﬂuids. Thus, the stimula-

tion system must be designed to be capable of stimu-

late the muscle with the lowest possible voltage. To

achieve this, biphasic square wave pulses must be

used and manipulated in terms of amplitude, dura-

tion and frequency.While maintaining the frequency

between 10 and 100 Hz and the pulse duration lower

than 999 µs, the amplitude is subjective as each treat-

ment and human body requires a different value. For

this reason, implementing biphasic pulses guarantees

to reduce the tissue damage and irritation provoked by

the stimulation in long periods of time (Tallgren et al.,

2005), and helps to avoid the muscle’s premature fa-

tigue (Lynch and Popovic, 2008).

To identify the original responses of the muscle

and compare to the current responses, the sampling

rate for the acquisition of the skeletal muscle signals

must be at least 1 kHz on a frequency bandwidth of

18 − 480 Hz. The correct acquisition of signals from

0.8mV requires the use a resolution of 12 bits in the

conversion from analogue to digital.

In terms of investment, the developed system

should provide a low-cost and competitive price when

compared to commercial solutions, e.g., (Biometrics

Ltd, 2021), and the signal acquisition and stimulation

functions should be completely integrated. In fact,

the rehabilitation routines should be adapted by AI al-

gorithms according to the biofeedback that takes into

account the patient proﬁle and the speciﬁc injury.

3.2 System Architecture

The proposed system architecture, illustrated in Fig-

ure 1, ensures the adequate acquisition and condition-

ing of the EMG signal (e.g., ampliﬁcation, ﬁltering

and conversion) to support the monitoring and the

stimulation of the muscle, as well as the transmission

of the acquired data to the mobile application running

in the smartphone, and later to a secure cloud where

the physician can follow-up the clinical results. The

design of this innovative biostimulation system allows

to reach a modular, scalable and low cost solution.

The wearable system comprises a pair of dry elec-

trodes to acquire the EMG signal and to perform the

FES, that are placed in the patient skin near the vastus

medialis muscle. The ﬁrst module, i.e. data acquisi-

Data Acquisition, Conditioning and Processing System for a Wearable-based Biostimulation

225

Figure 1: System architecture focusing the acquisition, sig-

nal conditioning and stimulation system.

tion block, is responsible to the implementation of the

real-time EMG data acquisition functions, namely to

collect the muscular information through EMG sen-

sors in each 1 ms. The collected electrode signals will

be forward to a signal conditioning block that ampli-

ﬁes the signal to a higher intensity that can be trans-

mitted to avoid losses and to ﬁlter the signal to atten-

uate undesired frequencies and noise. Next, the EMG

signal is converted from analogue to digital by using

an Analog/Digital Converter (ADC), and ﬁnally trans-

mitted to a microcontroller that will store and process

locally the data by executing the control rules to im-

plement the stimulation function. These control rules

are deﬁned according to the treatment protocol previ-

ously established by the physician, and triggered ac-

cording to the biofeedback from the EMG signal (and

in this way closing the loop between the signal acqui-

sition with the stimulation functions).

The commands for the muscle stimulation, pro-

vided by the microcontroller, are forward to the elet-

rical stimulation driver that is responsible to set the

amplitude and frequency of the electrical signals in-

jected in the FES electrodes. Closing the loop, the

acquisition module will measure the EMG signal that

corresponds to the muscle stimulus occasioned by the

current ﬂow.

The system also comprises an user-friendly mo-

bile application running in a smartphone, which in-

terface enables the patient to access the clinical data

and medical plan stipulated by the physician. With

this, the patient is able to start, stop and interrupt the

treatment any time, and visualise the EMG and FES

signals in real-time. For this purpose, the acquired

data from the wearable is transmitted to the mobile

application through the BLE wireless protocol, which

provides low power consumption and implementa-

tion complexity. A microservice architecture capa-

ble of communicating with the wearable system to

perform treatments at home was developed, ensuring

the data security and privacy, mainly using a Single-

Sign On (SSO) authentication service (Franco et al.,

2021). The data collected during the session and the

report provided by the patient at the end of the session

are transmitted to a secure cloud database, where the

physician can perform the clinical follow-up.

Another important module in this system is the use

of AI algorithms running in the cloud that allows to

analyse the collected data from the treatment session

(and also the previous historical data for that and oth-

ers patients) and derive new prescriptions for the next

treatment sessions according to the patient’s proﬁle

and response, and type of pathology.

4 DEVELOPMENT OF THE

BIOSTIMULATION SYSTEM

Considering the established requirements and the de-

signed system architecture, an integrated biostimu-

lation prototype system was developed as illustrated

in Figure 2 (note that the mobile app and the secure

cloud including the AI algorithms are not detailed in

this paper). The solution comprises the data acqui-

sition and electrical stimulation modules that oper-

ate integrated and in real-time, which means that the

stimulus are derived from the feedback from the col-

lected EMG signal, and this signal is the response of

the muscle to the stimulation.

VB 0.5/2/6

+3.3 V

Ref

EMG

Sensor

Signal Conditioning

(EBZ-AD8233)

Microcontroller

(ESP32)

BLE

Electrodes

Amplifying

Block

Conditioning

Block

Step-up

DC/DC

Converter

MT 3608

OUT

+ +

+3.3 V

30N06L

ADC

12 Bits

Resolution

Electrical Stimulation Block

Data Acquisition Block

Figure 2: Data acquisition and stimulation practical setup.

BIODEVICES 2022 - 15th International Conference on Biomedical Electronics and Devices

226

4.1 Data Acquisition System

The signal conditioning stage employs the

AD8233CB-EBZ board that contains an AD8233

heart rate monitor front end. Originally, the board

was designed to acquire ECG signals with a fre-

quency bandwidth of 7−25 Hz, but to measure EMG

signals, a modiﬁcation in the High-pass ﬁlter (HPF)

and Low-pass ﬁlter (LPF) circuits were necessary

to be performed to achieve a new bandwidth of

15 − 480 Hz. This allows to acquire and amplify the

skeletal muscle signals since the biosignals have low

amplitude levels, ranging from µV to a few mV.

The ESP32 Devkit V1 was used as microcon-

troller due to their low-cost, wireless communication

(BLE and Wi-Fi) and capability to operate with dual

core. The ESP32 has an ADC with 12 bits of reso-

lution, allowing to address the system requirements,

permitting to convert analogue values from 0.8 mV,

while e.g., the Arduino only has 10 bits of resolution

that only allows to measure a signal from 5 mV. In

spite of offering the possibility to communicate using

Wi-Fi and BLE wireless protocols, the selection was

to use BLE due to the low-power consumption that

ensures a longer battery autonomy (note that Wi-Fi is

widely employed due to its availability and capacity

of streaming but in these applications, the autonomy

of the battery is a strong requirement). Furthermore,

the dual-core functionality provided by the ESP32 is

crucial to guarantee that the acquisition and stimula-

tion can be done separately without delays and inter-

ference between the two tasks.

The acquisition system operates with a sampling

rate of 1 kHz, respecting the Nyquist theorem, to

avoid aliasing. The acquired data is transmitted to the

mobile application each 200 ms due to the limitation

of the BLE protocol to receive a maximum number of

500 bytes; each transmitted sample occupies 2 bytes,

which means that each transmitted package comprise

400 bytes. Finally, the collected data is stored in the

cloud database to be available to the physician for

clinical follow-up and to the AI algorithms for opti-

misation of the treatment for the next sessions.

4.2 Electrical Stimulation Module

The electrical stimulation block is responsible to gen-

erate the muscle stimulus by applying electrical sig-

nals according to the system commands. The com-

mands are two pulses, deﬁned in duration, amplitude

and frequency, and produced by the microcontroller

to compose the biphasic waveform, necessary to ap-

ply the stimulation. The conditioning block receives

the pulses and switch them by using a pair of MOS-

FETs, in this case 30N06L that were chosen due to

their low drain-source resistance (R

= 0.035) that

allows a higher switching speed.

The ampliﬁcation power block generates the

biphasic wave form and increases the output voltage

of the stimulation signal. This stage is composed by

a MT 3608 step-up converter with a 1 : 20 conversion

ratio, and a transformer VB 0.5/2/6 with a 6 : 230

transformation ratio. The boost converter was cho-

sen due to its low cost and ampliﬁcation rate, and the

transformer due to its small size (22/22.7/19 mm)

and the disposition of the pins, once it has four pins in

the low voltage side which enables the generation of

the biphasic pulse while the signal is ampliﬁed. These

parameters are deﬁned for the treatment applied to the

patient, established by the physician and the AI algo-

rithms.

The pulses are applied to both inputs of the low

voltage side of the transformer, and the middle points

are powered by a constant voltage ampliﬁed by the

step-up converter to reach a higher output rate. This

conﬁguration allows to generate the desired pulse

form and reach the necessary amplitude voltage to ful-

ﬁl the stimulus signal. The output amplitude is con-

trolled by using the Pulse Width Modulation (PWM)

technique where the pulses are adjusted in duration

and period, being the amplitude directly proportional

to the pulse duration.

5 EXPERIMENTAL RESULTS

Several experimental tests were performed in the de-

veloped prototype solution to verify the accomplish-

ment with the established requirements. The acquisi-

tion experiments were developed with a healthy per-

son, maintaining the same motion pattern during 3

seconds to start the movement, 4 seconds in the max-

imum muscle contraction and 3 seconds to return to

the rest. The stimulation was tested in a load to sim-

ulate the skin impedance, being able to evaluate the

generated waveform and the module performance.

5.1 Acquisition of the EMG Signal

In order to evaluate the proposed solution, the exper-

imental tests were performed using wet (Ag-AgCl)

and two distinct types of dry (Ti and TiCu) electrodes.

As shown in Figures 3 and 4, it is possible to note that

the signals acquired using the Ti ﬁlm and the poly-

lactic acid substrate (PLA) exhibit no saturation in

the readings and the energy levels were signiﬁcantly

close to those obtained with the wet sensors, present-

ing a signal-to-noise ratio (SNR) of 17.57 dB while

Data Acquisition, Conditioning and Processing System for a Wearable-based Biostimulation

227

the disposable ones provided a value of 18.15 dB.

This shows that the Ti dry electrode is competitive

to be used in such solutions since the energy levels

were close to those recorded by the Ag-AgCl elec-

trode, simplifying the algorithm to detect the amount

and moment of the muscle contractions, as well as to

avoid the use of gel that can cause skin irritation and

reduces the number of generated waste.

0 10000 20000 30000 40000 50000

Time (ms)

1.5

1.0

0.5

0.0

0.5

1.0

1.5

Amplitude (V)

EMG filtered

Figure 3: Analysis of ﬁltered reading with Ti and wet elec-

trodes signals.

0 10000 20000 30000 40000 50000

Time (ms)

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

Amplitude (V)

Signal Energy

Figure 4: Analysis of the signal energy of Ti and wet elec-

trodes.

In therms of the TiCu electrodes, the reached sig-

nal quality is 17.85 dB, which is higher than the one

provided by the Ti ﬁlm 17.57 dB, as illustrated in

Figure 5. However, the energy levels, illustrated in

Figure 6, were lower compared to the other types of

tested sensors which ends up hindering the detection

of muscle contractions and the diagnosis of the pa-

tient. Therefore, the presence of copper in the sensor

improves the quality of the acquired signal by pro-

viding an acquisition with a more reasonable level of

noise. However, the presence of this material in the

ﬁlm makes the electrode much more susceptible to

oxidation and can cause irritation in case of a pro-

longed period of time in contact with the skin.

Another critical factor observed during the tests

was the electromagnetic noise in the environment,

which affects the measurements and turns the system

inaccurate. In order to mitigate this effect, it was em-

ployed the shielding in the circuit container to isolate

the electronics of the environment noise, attenuating

0 10000 20000 30000 40000 50000

Time (ms)

1.5

1.0

0.5

0.0

0.5

1.0

1.5

2.0

Amplitude (V)

EMG filtered

Figure 5: Analysis of ﬁltered reading with TiCu and wet

electrodes.

0 10000 20000 30000 40000 50000

Time (ms)

0.00

0.25

0.50

0.75

1.00

1.25

1.50

1.75

2.00

Amplitude (V)

Signal Energy

Figure 6: Analysis of the signal energy of TiCu and wet

electrodes.

possible magnetic coupling and high-frequency inter-

ference. Protecting the circuit from these effects al-

lowed to improve the system reliability and to reduce

the measurements uncertainty.

In terms of data transmission between the wear-

able and the smartphone application using BLE, the

wearable system in average sends the acquired data

each 206 ms, and the mobile application receives the

information basically in real-time, with a delay of

some µs. This achievement is reasonable since the

microcontroller is also performing the signal process-

ing and the generation of the stimulation pulses, and

the mobile application is plotting the received signals

and running lighter artiﬁcial intelligence trained mod-

els to adapt the stimulation rules used by the wear-

able.

5.2 Electrical Stimulation

The experimental tests related to the electrical stimu-

lation were performed to evaluate if a certain stimula-

tion voltage is achieved by adjusting the pulse vari-

ables, which are executed according to the control

rules deﬁned by the AI algorithms that are continu-

ously updated on the wearable by the mobile appli-

cation. The ampliﬁcation block was powered with a

voltage of 3.3V, the step-up converter elevates the in-

put voltage to an approximate value of 8.2V, and the

transformer generates the biphasic waveform and am-

BIODEVICES 2022 - 15th International Conference on Biomedical Electronics and Devices

228

pliﬁes the pulses according to the pulse width.

Since the objective was to achieve an output am-

plitude near 30V, the width of the pulses was adjusted

until the amplitude reached the desired value, starting

with a width value of 5 µs and increasing the width ac-

cording to the balance of the pulse. During the modi-

ﬁcation of the pulses’ width, the pulses become unbal-

anced, i.e. the positive component gradually becomes

bigger than the negative component. To correct this

issue, the width of the negative component is set with

a greater value than the positive one.

According to this, the amplitude of 30V was

reached for the values of 360 µs for the positive pulse,

550 µs for the negative pulse, and 80 ms for the pe-

riod. The pulses produced by the microcontroller,

based on the input values explained above, can be ob-

served in Figure 7. These pulses are sent to the trans-

former that composes the biphasic wave form and am-

pliﬁes its amplitude.

Figure 7: Stimulation pulses produced by the microcon-

troller.

The resulting waveform for the previous pulses is

showed in Figure 8, with the biphasic wave signal

having approximately 30V of amplitude. The result-

ing signal is close to the expected, presenting the de-

sired output voltage and a balanced shape, with the

negative component bigger than the positive compo-

nent according to the width values previously estab-

lished for the input pulses.

Figure 8: Experimental stimulation signal with approxi-

mately 30V of amplitude.

6 CONCLUSIONS

Wearable systems have a vast range of applications,

e.g., sports and healthcare, representing technologi-

cal alternatives to assist the execution of rehabilita-

tion treatments and remote monitoring of vital health

parameters.

The proposed innovative and low-cost wearable

solution considers the use of dry electrodes and an

adaptive control of the bio-stimulation based on the

sEMG feedback, as well as the use of artiﬁcial in-

telligence algorithms to adapt the treatment proto-

col according to the pathology and the patient pro-

ﬁle. Preliminary experimental tests were carried out

and clearly shows that the main system requirements

were achieved, mainly in terms of the frequency of

sampling the EMG signal and the capability to inte-

grate the data acquisition with the stimulation using

dry electrodes. Also the effective transmission of col-

lected data from EMG and FES signals for the mobile

application was veriﬁed, being the microcontroller ca-

pable to manage the signal acquisition and transmis-

sion, and to generate the stimulation pulses.

During the experiments, it was also clear the need

to ﬁx the electrodes to prevent them from changing

their position during the patient’s movements and to

apply the right pressure as this inﬂuences the noise

and may cause the EMG sensors to lose the reference

signal. The electromagnetic noise of the environment

strongly inﬂuences the quality of the acquired signal,

as well as the system accuracy, which is also affected

by the ADC resolution. For this purpose, the circuit

container was shielded to isolate the electronics from

the electromagnetic noise, with signiﬁcant results.

Future work is devoted to test the developed solu-

tion using dry electrodes with connectors that provide

a better electrical and mechanical connection and are

more malleable to better adjust to the curvature of the

leg. The implementation of the developed solution in

a PCB and the integration of AI algorithms to provide

adaptive and customised treatments will be also con-

sidered as future work.

ACKNOWLEDGEMENTS

This work was supported by the European Regional

Development Fund (ERDF) through the Operational

Programme for Competitiveness and International-

ization (COMPETE 2020), under Portugal 2020 in

the framework of the NanoStim - Nanomaterials for

wearable-based integrated biostimulation (POCI-01-

0247-FEDER-045908) project.

Data Acquisition, Conditioning and Processing System for a Wearable-based Biostimulation

229

REFERENCES

Biometrics Ltd (2021). Datalog MWX8. ttp://www.

biometricsltd.com/datalog.tm. Accessed: 26.11.2021.

Chen, X.-B., Zhou, Y.-X., Wang, H.-P., L

u, X.-Y., and

Wang, Z.-G. (2017). Design of sEMG-detecting Cir-

cuit for EMG-Bridge. In Proc. of 39th Annual Int’l

Conference of the IEEE Engineering in Medicine and

Biology Society (EMBC), pages 382–385.

Chun, S., Kim, J., Pang, C., et al. (2019). A Trans-

parent, Glue-free, Skin-attachable Graphene Pressure

Sensor with Micropillars for Skin-elasticity Measure-

ment. Nanotechnology, 30(33):335501.

Da Silva, H. P., Guerreiro, J., Lourenc¸o, A., Fred, A., and

Martins, R. (2014). BITalino: A Novel Hardware

Framework for Physiological Computing. In PhyCS,

pages 246–253.

Delsys (2021). Trigno Research+. ttps://delsys.com/trigno/

researc/. Accessed: 26.11.2021.

Di Flumeri, G., Aric

o, P., Borghini, G., Sciaraffa, N.,

Di Florio, A., and Babiloni, F. (2019). The dry rev-

olution: evaluation of three different eeg dry electrode

types in terms of signal spectral features, mental states

classiﬁcation and usability. Sensors, 19(6):1365.

Ergeneci, M., Gokcesu, K., Ertan, E., and Kosmas, P.

(2018a). An embedded, eight channel, noise can-

celing, wireless, wearable semg data acquisition sys-

tem with adaptive muscle contraction detection. IEEE

Transactions on Biomedical Circuits and Systems,

12(1):68–79.

Ergeneci, M., Gokcesu, K., Ertan, E., and Kosmas, P.

(2018b). An embedded, eight channel, noise can-

celing, wireless, wearable semg data acquisition sys-

tem with adaptive muscle contraction detection. IEEE

Transactions on Biomedical Circuits and Systems,

12(1):68–79.

Fayyaz Shahandashti, P., Pourkheyrollah, H., Jahanshahi,

A., and Ghafoorifard, H. (2019). Highly con-

formable stretchable dry electrodes based on in-

expensive ﬂex substrate for long-term biopotential

(EMG/ECG) monitoring. Sensors and Actuators A:

Physical, 295:678–686.

Ferraz, R. d. S., Fonseca, R. S. d., Rodrigues, I. T., Silva, C.

B. d., et al. (2021). Surface electromyographic signal

acquisition system for real time monitoring of upper

limbs muscles. Journal of Computational and Theo-

retical Nanoscience, 18(4):1147–1152.

Franco, T., Henriques, P. R., Alves, P., et al. (2021). System

Architecture for Home Muscle Rehabilitation Treat-

ment. In Submitted to the 10th World Conference on

Information Systems and Technologies.

Koh, A., Kang, D., Xue, Y., Lee, S., Pielak, R. M.,

Kim, J., Hwang, T., Min, S., Banks, A., Bastien,

P., et al. (2016). A soft, wearable microﬂuidic

device for the capture, storage, and colorimetric

sensing of sweat. Science translational medicine,

8(366):366ra165–366ra165.

Li, X., Zhou, Z., Ji, M., and Liu, W. (2021). A

wearable wireless device designed for surface elec-

tromyography acquisition. Microsystem Technologies,

27(4):1787–1795.

Lou, Z., Wang, L., Jiang, K., Wei, Z., and Shen, G. (2020).

Reviews of wearable healthcare systems: Materials,

devices and system integration. Materials Science and

Engineering: R: Reports, 140:100523.

Lynch, C. L. and Popovic, M. R. (2008). Functional elec-

trical stimulation. IEEE control systems magazine,

28(2):40–50.

Park, H.-K., Jung, J., Lee, D.-W., Shin, H. C.,

Lee, H.-J., and Lee, W.-H. (2021). A wear-

able electromyography-controlled functional electri-

cal stimulation system improves balance, gait func-

tion, and symmetry in older adults. Technology and

Health Care, (Preprint):1–13.

Peng, H.-L., Jing-Quan Liu, Tian, H.-C., Dong, Y.-Z., Yang,

B., Chen, X., and Yang, C.-S. (2016). A novel passive

electrode based on porous ti for eeg recording. Sen-

sors and Actuators B: Chemical, 226:349–356.

Robertson, D. G. E., Caldwell, G. E., Hamill, J., Kamen,

G., and Whittlesey, S. (2013). Research methods in

biomechanics. Human kinetics.

Rodrigues, M., Fiedler, P., K

uchler, N., P. Domingues, R.,

Lopes, C., Borges, J., Haueisen, J., and Vaz, F. (2020).

Dry electrodes for surface electromyography based on

architectured titanium thin ﬁlms. Materials, 13(9).

Tallgren, P., Vanhatalo, S., Kaila, K., and Voipio, J. (2005).

Evaluation of commercially available electrodes and

gels for recording of slow eeg potentials. Clinical

Neurophysiology, 116(4):799–806.

Trung, T. Q. and Lee, N.-E. (2016). Flexible and stretch-

able physical sensor integrated platforms for wear-

able human-activity monitoringand personal health-

care. Advanced materials, 28(22):4338–4372.

Zhu, L., Mao, G., Su, H., Zhou, Z., Li, W., L

u, X., and

Wang, Z. (2021). A wearable, high-resolution, and

wireless system for multichannel surface electromyo-

graphy detection. IEEE Sensors Journal, 21(8):9937–

9948.

BIODEVICES 2022 - 15th International Conference on Biomedical Electronics and Devices

230