Andrea Mongardi

Mr. Andrea Mongardi

Department of Electronics and Telecommunications, Politecnico di Torino
orcid

Research Focus & Metrics

Academic Career Length

8Years (2019-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
3 October 2022 in IEEE Transactions on Biomedical Circuits and Systems
Verified
Journal Article
Open access
Andrea MongardiAndrea Mongardi
Andrea Mongardi
Fabio RossiFabio Rossi
Fabio Rossi
Andrea PrestiaAndrea Prestia
Andrea Prestia
PRPaolo Motto Ros
Paolo Motto Ros
Massimo Ruo RochMassimo Ruo Roch
Massimo Ruo Roch
DDDanilo Demarchi
Danilo Demarchi
Hand gesture recognition has recently increased its popularity as Human-Machine Interface (HMI) in the biomedical field. Indeed, it can be performed involving many different non-invasive techniques, e.g., surface ElectroMyoGraphy (sEMG) or PhotoPlethysmoGraphy (PPG). In the last few years, the interest demonstrated by both academia and industry brought to a continuous spawning of commercial and custom wearable devices, which tried to address different challenges in many application fields, from tele-rehabilitation to sign language recognition. In this work, we propose a novel 7-channel sEMG armband, which can be employed as HMI for both serious gaming control and rehabilitation support. In particular, we designed the prototype focusing on the capability of our device to compute the Average Threshold Crossing (ATC) parameter, which is evaluated by counting how many times the sEMG signal crosses a threshold during a fixed time duration (i.e., 130 ms), directly on the wearable device. Exploiting the event-driven characteristic of the ATC, our armband is able to accomplish the on-board prediction of common hand gestures requiring less power w.r.t. state of the art devices. At the end of an acquisition campaign that involved the participation of 26 people, we obtained an average classifier accuracy of 91.9% when aiming to recognize in real time 8 active hand gestures plus the idle state. Furthermore, with 2.92mA of current absorption during active functioning and 1.34mA prediction latency, this prototype confirmed our expectations and can be an appealing solution for long-term (up to 60 h) medical and consumer applications.

9 August 2021 in IEEE Sensors Journal
Verified
Journal Article
Open access
Fabio RossiFabio Rossi
Fabio Rossi
Andrea MongardiAndrea Mongardi
Andrea Mongardi
PRPaolo Motto Ros
Paolo Motto Ros
Massimo Ruo RochMassimo Ruo Roch
Massimo Ruo Roch
MMMaurizio Martina
Maurizio Martina
DDDanilo Demarchi
Danilo Demarchi
Device wearability and operating time are trending topics in recent state-of-art works on surface ElectroMyoGraphic (sEMG) muscle monitoring. No optimal trade-off, able to concurrently address several problems of the acquisition system like robustness, miniaturization, versatility, and power efficiency, has yet been found. In this tutorial we present a solution to most of these issues, embedding in a single device both an sEMG acquisition channel, with our custom event-driven hardware feature extraction technique (named Average Threshold Crossing), and a digital part, which includes a microcontroller unit, for (optionally) sEMG sampling and processing, and a Bluetooth communication, for wireless data transmission. The knowledge acquired by the research group brought to an accurate selection of each single component, resulting in a very efficient prototype, with a comfortable final size (57.8mm x 25.2mm x 22.1mm) and a consistent signal-to-noise ratio of the acquired sEMG (higher than 15 dB). Furthermore, a precise design of the firmware has been performed, handling both signal acquisition and Bluetooth transmission concurrently, thanks to a FreeRTOS custom implementation. In particular, the system adapts to both sEMG and ATC transmission, with an application throughput up to 2 kB s-1 and an average operating time of 80 h (for high resolution sEMG sampling), relaxable to 8Bs-1 throughput and about 230 h operating time (considering a 110mAh battery), in case of ATC acquisition only. Here we share our experience over the years in designing wearable systems for the sEMG detection, specifying in detail how our event-driven approach could benefit the device development phases. Some previous basic knowledge about biosignal acquisition, electronic circuits and programming would certainly ease the repeatability of this tutorial.

20 December 2021 in Sensors
Verified
Journal Article
Open access
PDF file
Fabio RossiFabio Rossi
Fabio Rossi
FSFederica Savi
Federica Savi
Andrea PrestiaAndrea Prestia
Andrea Prestia
Andrea MongardiAndrea Mongardi
Andrea Mongardi
DDDanilo Demarchi
Danilo Demarchi
GBGiovanni Buccino
Giovanni Buccino
Action observation treatment (AOT) exploits a neurophysiological mechanism, matching an observed action on the neural substrates where that action is motorically represented. This mechanism is also known as mirror mechanism. In a typical AOT session, one can distinguish an observation phase and an execution phase. During the observation phase, the patient observes a daily action and soon after, during the execution phase, he/she is asked to perform the observed action at the best of his/her ability. Indeed, the execution phase may sometimes be difficult for those patients where motor impairment is severe. Although, in the current practice, the physiotherapist does not intervene on the quality of the execution phase, here, we propose a stimulation system based on neurophysiological parameters. This perspective article focuses on the possibility to combine AOT with a brain–computer interface system (BCI) that stimulates upper limb muscles, thus facilitating the execution of actions during a rehabilitation session. Combining a rehabilitation tool that is well-grounded in neurophysiology with a stimulation system, such as the one proposed, may improve the efficacy of AOT in the treatment of severe neurological patients, including stroke patients, Parkinson’s disease patients, and children with cerebral palsy.

1 November 2019
Verified
Conference Paper
Closed access
Andrea MongardiAndrea Mongardi
Andrea Mongardi
PRPaolo Motto Ros
Paolo Motto Ros
Fabio RossiFabio Rossi
Fabio Rossi
Massimo Ruo RochMassimo Ruo Roch
Massimo Ruo Roch
MMMaurizio Martina
Maurizio Martina
DDDanilo Demarchi
Danilo Demarchi
Latency and power consumption management are present-day hot topics for wearable devices and IoT applications. This paper presents the implementation of a low power system for hand movement recognition, based on surface ElectroMyo-Graphic (sEMG) signals, suitable for human-machine interface. Every time the sEMG signal crosses a predefined threshold an event is generated, implementing the Average Threshold Crossing (ATC) technique directly on-boards on the subject arm. Resulting quasi-digital signals, averaged over a fixed time window, are sent to an ARM Cortex-M4F processor which implements a fully-connected Neural Network (NN) able to recognize six different gestures from only three input channels. Dataset creation has involved 25 healthy people, each one performing five movements within five repeated sessions. The NN has been trained using the per-subject holdout validation method, obtaining an accuracy of 96.34%. With a maximum latency of 8.5 ms and an average power consumption of only 0.8 mW at full recognition rate, the proposed NN implementation shows promising results.

21 December 2021 in IEEE Transactions on Biomedical Circuits and Systems
Verified
Journal Article
Closed access
Andrea PrestiaAndrea Prestia
Andrea Prestia
Fabio RossiFabio Rossi
Fabio Rossi
Andrea MongardiAndrea Mongardi
Andrea Mongardi
PRPaolo Motto Ros
Paolo Motto Ros
Massimo Ruo RochMassimo Ruo Roch
Massimo Ruo Roch
DDDanilo Demarchi
Danilo Demarchi
In this work, a system for controlling Functional Electrical Stimulation (FES) has been experimentally evaluated. The peculiarity of the system is to use an event-driven approach to modulate stimulation intensity, instead of the typical feature extraction of surface ElectroMyoGraphic (sEMG) signal. To validate our methodology, the system capability to control FES was tested on a population of 17 subjects, reproducing 6 different movements. Limbs trajectories were acquired using a gold standard motion tracking tool. The implemented segmentation algorithm has been detailed, together with the designed experimental protocol. A motion analysis was performed through a multiparametric evaluation, including the extraction of features such as the trajectory area and the movement velocity. The obtained results show a median cross-correlation coefficient of 0.910 and a median delay of 800 ms, between each couple of voluntary and stimulated exercise, making our system comparable w.r.t. state-of-the-art works. Furthermore, a 97.39% successful rate on movement replication demonstrates the feasibility of the system for rehabilitation purposes.

1 October 2019
Verified
Conference Paper
Closed access
Andrea MongardiAndrea Mongardi
Andrea Mongardi
Fabio RossiFabio Rossi
Fabio Rossi
PRPaolo Motto Ros
Paolo Motto Ros
ASAlessandro Sanginario
Alessandro Sanginario
Massimo Ruo RochMassimo Ruo Roch
Massimo Ruo Roch
DDDanilo Demarchi
Danilo Demarchi
This demonstration presents a low power embedded system to classify hand movements. The surface ElectroMyo-Graphic (sEMG) signals acquired from the forearm are preprocessed using the Average Threshold Crossing (ATC) event-driven technique, which heavily reduces hardware complexity and power consumption. The quasi-digital output is sent to an ultra low power microcontroller, which implements a fully-connected Neural Network (NN). A small Arduino-based tank is used to demonstrate the real-time behavior of the system and to show the correctness of the predicted gestures 1 .