Counting Grasping Action Using Force Myography: An Exploratory Study With Healthy Individuals

A survey on the state of the art of force myography technique (FMG): analysis and assessment

Precise feedback assures precise control commands especially for assistive or rehabilitation devices. Biofeedback systems integrated with assistive or rehabilitative robotic exoskeletons tend to increase its performance and effectiveness. Therefore, there has been plenty of research in the field of biofeedback covering different aspects such as signal acquisition, conditioning, feature extraction and integration with the control system. Among several types of biofeedback systems, Force myography (FMG) technique is a promising one in terms of affordability, high classification accuracies, ease to use, and low computational cost. Compared to traditional biofeedback systems such as electromyography (EMG) which offers some invasive techniques, FMG offers a completely non-invasive solution with much less effort for preprocessing with high accuracies. This work covers the whole aspects of FMG technique in terms of signal acquisition, feature extraction, signal processing, developing the machine learning model, evaluating tools for the performance of the model. Stating the difference between real-time and offline assessment, also highlighting the main uncovered points for further study, and thus enhancing the development of this technique.

A Wearable Force Myography-Based Armband for Recognition of Upper Limb Gestures

Force myography (FMG) represents a promising alternative to surface electromyography (EMG) in the context of controlling bio-robotic hands. In this study, we built upon our prior research by introducing a novel wearable armband based on FMG technology, which integrates force-sensitive resistor (FSR) sensors housed in newly designed casings. We evaluated the sensors' characteristics, including their load-voltage relationship and signal stability during the execution of gestures over time. Two sensor arrangements were evaluated: arrangement A, featuring sensors spaced at 4.5 cm intervals, and arrangement B, with sensors distributed evenly along the forearm. The data collection involved six participants, including three individuals with trans-radial amputations, who performed nine upper limb gestures. The prediction performance was assessed using support vector machines (SVMs) and k-nearest neighbor (KNN) algorithms for both sensor arrangments. The results revealed that the developed sensor exhibited non-linear behavior, and its sensitivity varied with the applied force. Notably, arrangement B outperformed arrangement A in classifying the nine gestures, with an average accuracy of 95.4 ± 2.1% compared to arrangement A's 91.3 ± 2.3%. The utilization of the arrangement B armband led to a substantial increase in the average prediction accuracy, demonstrating an improvement of up to 4.5%.

Feasibility of force myography for the direct control of an assistive robotic hand orthosis in non-impaired individuals

BACKGROUND

Assistive robotic hand orthoses can support people with sensorimotor hand impairment in many activities of daily living and therefore help to regain independence. However, in order for the users to fully benefit from the functionalities of such devices, a safe and reliable way to detect their movement intention for device control is crucial. Gesture recognition based on force myography measuring volumetric changes in the muscles during contraction has been previously shown to be a viable and easy to implement strategy to control hand prostheses. Whether this approach could be efficiently applied to intuitively control an assistive robotic hand orthosis remains to be investigated.

METHODS

In this work, we assessed the feasibility of using force myography measured from the forearm to control a robotic hand orthosis worn on the hand ipsilateral to the measurement site. In ten neurologically-intact participants wearing a robotic hand orthosis, we collected data for four gestures trained in nine arm configurations, i.e., seven static positions and two dynamic movements, corresponding to typical activities of daily living conditions. In an offline analysis, we determined classification accuracies for two binary classifiers (one for opening and one for closing) and further assessed the impact of individual training arm configurations on the overall performance.

RESULTS

We achieved an overall classification accuracy of 92.9% (averaged over two binary classifiers, individual accuracies 95.5% and 90.3%, respectively) but found a large variation in performance between participants, ranging from 75.4 up to 100%. Averaged inference times per sample were measured below 0.15 ms. Further, we found that the number of training arm configurations could be reduced from nine to six without notably decreasing classification performance.

CONCLUSION

The results of this work support the general feasibility of using force myography as an intuitive intention detection strategy for a robotic hand orthosis. Further, the findings also generated valuable insights into challenges and potential ways to overcome them in view of applying such technologies for assisting people with sensorimotor hand impairment during activities of daily living.

Wrist-worn wearables based on force myography: on the significance of user anthropometry

BACKGROUND

Force myography (FMG) is a non-invasive technology used to track functional movements and hand gestures by sensing volumetric changes in the limbs caused by muscle contraction. Force transmission through tissue implies that differences in tissue mechanics and/or architecture might impact FMG signal acquisition and the accuracy of gesture classifier models. The aim of this study is to identify if and how user anthropometry affects the quality of FMG signal acquisition and the performance of machine learning models trained to classify different hand and wrist gestures based on that data.

METHODS

Wrist and forearm anthropometric measures were collected from a total of 21 volunteers aged between 22 and 82 years old. Participants performed a set of tasks while wearing a custom-designed FMG band. Primary outcome measure was the Spearman's correlation coefficient (R) between the anthropometric measures and FMG signal quality/ML model performance.

RESULTS

Results demonstrated moderate (0.3 ≤|R| < 0.67) and strong (0.67 ≤ |R|) relationships for ratio of skinfold thickness to forearm circumference, grip strength and ratio of wrist to forearm circumference. These anthropometric features contributed to 23-30% of the variability in FMG signal acquisition and as much as 50% of the variability in classification accuracy for single gestures.

CONCLUSIONS

Increased grip strength, larger forearm girth, and smaller skinfold-to-forearm circumference ratio improve signal quality and gesture classification accuracy.

Investigation of Regression Methods for Reduction of Errors Caused by Bending of FSR-Based Pressure Sensing Systems Used for Prosthetic Applications

The pressure map at the interface of a prosthetic socket and a residual limb contains information that can be used in various prosthetic applications including prosthetic control and prosthetic fitting. The interface pressure is often obtained using force sensitive resistors (FSRs). However, as reported by multiple studies, accuracies of the FSR-based pressure sensing systems decrease when sensors are bent to be positioned on a limb. This study proposes the use of regression-based methods for sensor calibration to address this problem. A sensor matrix was placed in a pressure chamber as the pressure was increased and decreased in a cyclic manner. Sensors’ responses were assessed when the matrix was placed on a flat surface or on one of five curved surfaces with various curvatures. Three regression algorithms, namely linear regression (LR), general regression neural network (GRNN), and random forest (RF), were assessed. GRNN was selected due to its performance. Various error compensation methods using GRNN were investigated and compared to improve instability of sensors’ responses. All methods showed improvements in results compared to the baseline. Developing a different model for each of the curvatures yielded the best results. This study proved the feasibility of using regression-based error compensation methods to improve the accuracy of mapping sensor readings to pressure values. This can improve the overall accuracy of FSR-based sensory systems used in prosthetic applications.

Investigation of Channel Selection for Gesture Classification for Prosthesis Control Using Force Myography: A Case Study

Background: Various human machine interfaces (HMIs) are used to control prostheses, such as robotic hands. One of the promising HMIs is Force Myography (FMG). Previous research has shown the potential for the use of high density FMG (HD-FMG) that can lead to higher accuracy of prosthesis control. Motivation: The more sensors used in an FMG controlled system, the more complicated and costlier the system becomes. This study proposes a design method that can produce powered prostheses with performance comparable to that of HD-FMG controlled systems using a fewer number of sensors. An HD-FMG apparatus would be used to collect information from the user only in the design phase. Channel selection would then be applied to the collected data to determine the number and location of sensors that are vital to performance of the device. This study assessed the use of multiple channel selection (CS) methods for this purpose. Methods: In this case study, three datasets were used. These datasets were collected from force sensitive resistors embedded in the inner socket of a subject with transradial amputation. Sensor data were collected as the subject carried out five repetitions of six gestures. Collected data were then used to asses five CS methods: Sequential forward selection (SFS) with two different stopping criteria, minimum redundancy-maximum relevance (mRMR), genetic algorithm (GA), and Boruta. Results: Three out of the five methods (mRMR, GA, and Boruta) were able to decrease channel numbers significantly while maintaining classification accuracy in all datasets. Neither of them outperformed the other two in all datasets. However, GA resulted in the smallest channel subset in all three of the datasets. The three selected methods were also compared in terms of stability [i.e., consistency of the channel subset chosen by the method as new training data were introduced or some training data were removed (Chandrashekar and Sahin, 2014)]. Boruta and mRMR resulted in less instability compared to GA when applied to the datasets of this study. Conclusion: This study shows feasibility of using the proposed design method that can produce prosthetic systems that are simpler than HD-FMG systems but have performance comparable to theirs.

Tactile Signatures and Hand Motion Intent Recognition for Wearable Assistive Devices

Within the field of robotics and autonomous systems where there is a human in the loop, intent recognition plays an important role. This is especially true for wearable assistive devices used for rehabilitation, particularly post-stroke recovery. This paper reports results on the use of tactile patterns to detect weak muscle contractions in the forearm while at the same time associating these patterns with the muscle synergies during different grips. To investigate this concept, a series of experiments with healthy participants were carried out using a tactile arm brace (TAB) on the forearm while performing four different types of grip. The expected force patterns were established by analysing the muscle synergies of the four grip types and the forearm physiology. The results showed that the tactile signatures of the forearm recorded on the TAB align with the anticipated force patterns. Furthermore, a linear separability of the data across all four grip types was identified. Using the TAB data, machine learning algorithms achieved a 99% classification accuracy. The TAB results were highly comparable to a similar commercial intent recognition system based on a surface electromyography (sEMG) sensing.

A Review of Force Myography Research and Development

Information about limb movements can be used for monitoring physical activities or for human-machine-interface applications. In recent years, a technique called Force Myography (FMG) has gained ever-increasing traction among researchers to extract such information. FMG uses force sensors to register the variation of muscle stiffness patterns around a limb during different movements. Using machine learning algorithms, researchers are able to predict many different limb activities. This review paper presents state-of-art research and development on FMG technology in the past 20 years. It summarizes the research progress in both the hardware design and the signal processing techniques. It also discusses the challenges that need to be solved before FMG can be used in an everyday scenario. This paper aims to provide new insight into FMG technology and contribute to its advancement.

An Investigation on the Sampling Frequency of the Upper-Limb Force Myographic Signals

Force myography (FMG) is an emerging method to register muscle activity of a limb using force sensors for human-machine interface and movement monitoring applications. Despite its newly gained popularity among researchers, many of its fundamental characteristics remain to be investigated. The aim of this study is to identify the minimum sampling frequency needed for recording upper-limb FMG signals without sacrificing signal integrity. Twelve healthy volunteers participated in an experiment in which they were instructed to perform rapid hand actions with FMG signals being recorded from the wrist and the bulk region of the forearm. The FMG signals were sampled at 1 kHz with a 16-bit resolution data acquisition device. We downsampled the signals with frequencies ranging from 1 Hz to 500 Hz to examine the discrepancies between the original signals and the downsampled ones. Based on the results, we suggest that FMG signals from the forearm and wrist should be collected with minimum sampling frequencies of 54 Hz and 58 Hz for deciphering isometric actions, and 70 Hz and 84 Hz for deciphering dynamic actions. This fundamental work provides insight into minimum requirements for sampling FMG signals such that the data content of such signals is not compromised.

Does force myography recorded at the wrist correlate to resistance load levels during bicep curls?

Resistance strength training is a proven method to improve bone density and muscle strength. A solution capable of automatically detecting the resistance force level exerted by a user from a wrist-based device can offer great convenience to the trainee and hence facilitate a better training outcome. In this short communication, we present our investigation aimed at exploring if force myographic (FMG) signals recorded at the wrist can predict the relative resistance levels that are associated with different weights. Specifically, we investigated the Spearman's correlations between the wrist FMG signal features and the dumbbell weights during a bicep curl exercise. 10 volunteers were recruited to perform a total of 100 curl actions, which included both the hammer and regular curls while the wrist FMG signals were being recorded. Three sets of weights ranging from 0.2 lb to 8 lb were used. For the hammer curls, a median correlation coefficient of 0.92 with an interquartile range (IQR) of 0.03 was obtained. For the regular curls, a 0.94 median correlation with a 0.02 IQR was obtained. We also used the data from the first 36 curls to generate a classifier model and applied it onto the rest of the data. An averaged validation accuracy of 88% was obtained. The results of this study showed the potential use of wrist FMG signal to detect different levels of the load during exercises; such information could potentially be used as feedback in fitness, sports, and rehabilitation activities.

Regressing grasping using force myography: an exploratory study

BACKGROUND

Partial hand amputation forms more than 90% of all upper limb amputations. This amputation has a notable effect on the amputee's life. To improve the quality of life for partial hand amputees different prosthesis options, including externally-powered prosthesis, have been investigated. The focus of this work is to explore force myography (FMG) as a technique for regressing grasping movement accompanied by wrist position variations. This study can lay the groundwork for a future investigation of FMG as a technique for controlling externally-powered prostheses continuously.

METHODS

Ten able-bodied participants performed three hand movements while their wrist was fixed in one of six predefined positions. The angle between Thumb and Index finger ([Formula: see text]), and Thumb and Middle finger ([Formula: see text]) were calculated as measures of grasping movements. Two approaches were examined for estimating each angle: (i) one regression model, trained on data from all wrist positions and hand movements; (ii) a classifier that identified the wrist position followed by a separate regression model for each wrist position. The possibility of training the system using a limited number of wrist positions and testing it on all positions was also investigated.

RESULTS

The first approach had a correlation of determination ([Formula: see text]) of 0.871 for [Formula: see text] and [Formula: see text]. Using the second approach [Formula: see text] and [Formula: see text] were obtained. The first approach is over two times faster than the second approach while having similar performance; thus the first approach was selected to investigate the effect of the wrist position variations. Training with 6 or 5 wrist positions yielded results which were not statistically significant. A statistically significant decrease in performance resulted when less than five wrist positions were used for training.

CONCLUSIONS

The results indicate the potential of FMG to regress grasping movement, accompanied by wrist position variations, with a regression model for each angle. Also, it is necessary to include more than one wrist position in the training phase.