Measurement of Forearm Muscle Activity and Elbow Angle of the Elderly in Using Different Types of Cane Handles through Electromyography (EMG) and Kinovea Software

. Elderly individuals often experience physical limitations, leading to a decline in their walking ability and an increased risk of falling. Falls among the elderly are a critical issue that can be minimized by using walking aids such as a Cane. However, the use of a cane poses other risks, such as injuries like carpal tunnel syndrome (CTS), caused by repetitive hand movements and changes in hand angles. The objective of this research is to select the appropriate cane handle to minimize the strain on the wrist by analyzing forearm muscle signals measured using EMG (Electromyography) combined with angle measurements using Kinovea software. The study involved 8 elderly subjects who regularly used a cane during walking gait tests. The distribution of the load during walking with the cane was assessed. The results, in the form of EMG data, were segmented based on time and angles, allowing the extraction of features using the root mean square (RMS) method. The next step involved conducting a three-way ANOVA statistical test for experimental design and identifying interactions between each factor. The findings revealed that the cane handle with a fritz-type grip, positioned posteriorly, and used during the initial phase of walking had the lowest values, measuring 50.3 73 μV for the fritz cane, 52.03 μV for the posterior position, and 56.74 μV for the initial angle. This indicates that a cane with a fritz handle distributes the load on the forearm towards the wrist more minimally, reducing the risk of CTS.


Introduction
The elderly are community members synonymous with a dynamic quality of life influenced by both individuals and the environment.Maintaining the quality of life for the elderly is crucial.However, their quality of life is at risk of decline due to the aging process, which can have direct impacts in the form of physical disability limitations.[1].
Physical limitations in the elderly can lead to disabilities, including a decrease in their ability to walk.This reduced walking capacity can disrupt balance and stability, thereby increasing the risk of falling.Falls among the elderly can result in a diminished quality of life due to traumas and injuries [2].In Indonesia, the prevalence of fall-related injuries in individuals aged over 55 years has reached 49.4%, and this figure increases to 67.1% for those over 65 years old [3].The elderly aged over 65 have a high incidence of falls, and this rate continues to rise each year, making falls among the elderly a significant global health concern [3].
The risk of falls among the elderly is a critical issue that can be minimized by using a walker.The purpose of using a walker, such as a cane, for the elderly is to help maintain their quality of life [4].When the elderly use canes while moving, it leads to changes in the cane's * haidarimam16@gmail.composition relative to the body-either in front of the body (anterior), parallel to the body (middle), or behind the body (posterior) [5].These changes in position cause repetitive flexion-extension and abduction-adduction movements in the upper extremities, particularly in the wrist [6], resulting in alterations in the distribution of load on the elbow and forearm [7].
Handle canes play a crucial role in minimizing the distribution of forearm load, as improper selection of handle canes can increase the risk of hand injuries like carpal tunnel syndrome [6].This injury occurs due to repetitive changes in the distribution of load on the forearm, particularly in the lower elbow, leading to swollen tendons that pinch the median nerve in the carpal tunnel, resulting in wrist pain [8,9].The prevalence of carpal tunnel syndrome (CTS) injury is 1.55%, and it increases by 5% in the elderly over 65 years of age [10,11] To address this issue, selecting the appropriate cane handle can help minimize the distribution of forearm load below the elbow, consequently reducing the risk of CTS injury.The selection of handle canes for the elderly can be achieved by measuring the distribution of forearm muscle load.
The amount of forearm muscle load distribution can be measured using electromyography (EMG).This device records the contraction of the elbow forearm muscles at regular intervals, allowing it to detect body biosignals, which are then processed into root mean square (RMS) values.These RMS values indicate the physiological activity level of the muscles [12].In a study conducted by Bertolacini et al. (2017), EMG was utilized to measure the amount of elbow forearm load distribution while using different cane handles.The analysis of the forearm load distribution using EMG provides precise and accurate information, enabling researchers to obtain detailed and reliable results [13].
In this study, an EMG device is utilized to measure the muscle activity and load distribution in the elderly's elbow forearm when using a cane.The primary objective of this research is to understand how the elbow forearm muscles respond when supporting body weight on the cane.The study aims to identify the differences in load distribution across several cane handles while walking and changing the position of the cane relative to the body.
The research adopts a factorial experimental design, focusing on various types of cane handles, including fritz handle, crook handle, offset handle, ergonomic handle, palm handle, and derby handle.By analyzing the amount of elbow forearm load distribution through EMG measurements, the study seeks to find the most appropriate cane handle that can enhance the quality of life for the elderly and reduce the risk of carpal tunnel syndrome (CTS) By gaining insights into the optimal cane handle for the elderly, this research has the potential to improve the overall well-being of the elderly population, ensuring safer and more comfortable mobility while minimizing the risk of CTS.

Study Design and Participants
At this stage, it is carried out to collect various information about the conditions of elderly cane users such as age, activities, environment and the quality of life they have.Observation also records the demographics of the number of cane users and the problems they experience when using canes, which are carried out directly with interviews and documentation at the Surakarta area Nursing Home.The results of observations according to WHOQOL-BREF obtained that the quality of life of the elderly in homes and posbindu has a good value of 73.8 with key elements as supporting aspects such as physical and social aspects have good quality of 65.1 and 73.1 and environmental and psychological aspects have very good quality with values of 82 and 75.1 as described in the figure below.After observing the research site, eight elderly subjects were selected with inclusion criteria including, having an elbow forearm circumference of 16-26 cm so that they can show the selected muscle arm area, as well as the age of elderly subjects with a range of 58 years to 85 years of lower extremity sufferers.The results of the final research subjects obtained 8 elderly women with an average age of 77 years with a height of 155 cm and a weight of 55 kg.All participants routinely use canes, especially standard canes for daily activities.All participants had functionally normal vision, either with or without additional lenses.Any potential participant who has a disease that affects balance will be excluded.

Procedure
The test protocol is structured by outlining an area of the test environment that has a flat base, is free from distractions, does not endanger the subject and there is ample space to move.Testing on 8 elderly subjects by walking with a cane for 3 tests to get 8 gait cycles when using a cane.The research subjects were given a rest period of 15 minutes after each test.Testing gait walking using a cane is done by walking three meters using a cane that has been determined, then placing the camera 2.5 meters perpendicular to the direction of the road and a tripod height of 0.5 meters.The research subject was paired with 1 EMG channel attached to the selected muscle area (extensor carpi radialis) and a joint marker to determine the elbow angle attached to the midpoint of the shoulder, elbow and wrist.
This test uses supporting tools consisting of canes with different cane handles such as fritz handle, crook handle, offset handle, ergonomic handle, palm handle and derby handle, as well as BITalino (r)evolution Plugged Kit BLE/BT microcontroller, channel sensor, transmitter and EMG electrode, then laptop, goniometer, disinfectant, cotton swab, stabilizer, joint marker and 50-60 fps camera.Software consists of Opensignal EMG, Matlab, SPSS, Microsoft Excel and Kinovea.

Determining Muscle Area
First, this stage determines and tests the response area of the elbow forearm muscle using the EMG hardware BITalino (r)evolution Plugged Kit BLE/BT by setting the sampling rate at 1000Hz.Testing on the elbow forearm muscles that play a role when grasping and holding the body weight on the cane in the subject's right hand.Testing starts from the extensor carpi radialis, flexor carpi radialis, extensor carpi ulnaris and flexor carpi ulnaris muscle areas with testing each muscle for 5 seconds.The results of the radial and ulnar muscle area response readings are read in Opensignal software in the form of biosignals.Selection of the elbow forearm muscle area is done after obtaining data in the form of signals on Opensignal Software Selection is made by biosignaling the most active muscle area based on EMG waveforms in each research subject, the selected muscle area will be tested for load distribution in the table below.The test results obtained the highest value in the extensor carpi radialis area with a value of 168 microVolts and the lowest in extensor carpi ulnaris at 40,523 microVolts.The amplitude results of the four muscle areas as described in Figure below.

Variable
Tests using full factorial design experiments, with factors or independent variables consisting of different types of cane handles, differences in the position of the hand against the body (cycle using the cane), and changes in the angle of each position when using the cane.The cane used is a standard cane with a single tip which consists of several types of cane handles such as fritz handle, crook handle, offset handle, ergonomic handle, palm handle and derby handle which are shown below.

Fig. 3. Handle Cane
The use of a cane causes changes in position and angle in the arm resulting in repetitive movements such as flexion-extension and abduction-adduction of the wrist [4].Changes in position consist of the position when the arm is in front of the body (anterior), the arm is in front of the body (middle) and when it is behind the body (posterior) [5].While the angle change consists of the initial angle and the final angle at the forearm or commonly referred to as the elbow angle [14].Confirmation of whether factors or variables regarding angle changes are used Kinovea software to take angle measurements.If the angle calculation using Kinovea software makes a difference between the initial and final angles, the factor or parameter will be used, if there is no change, the factor or parameter will be eliminated.

Data Analysis
EMG signals were analyzed in MATLAB software to produce root mean square (RMS) values which were then divided into segments based on the time of each position change and angle change according to permilidetic data from Kinovea software.The angle data from the initial and final changes were analyzed using excel to obtain the difference between the two angles.After obtaining RMS data on each treatment and angle change data, the data was then statistically analyzed using a pair t test for angle differences to determine whether there was indeed an angle change in each position, namely the initial and final angles, then a correlation analysis using Pearson on angle data and EMG as well as a three-way ANOVA statistical test on EMG data to determine the interaction between treatment on EMG data.

Tracking Position Change
Tracking position changes using Kinovea software to get millisecond time for each position change.With the framing feature in Kinovea, it is obtained per frame of elderly movement when using a stick which is described in the figure below.

Fig. 4. Tracking Position Change
The results of the angle change permutation are then collected and recorded together to obtain position segmentation data which is then put together in a table for processing at a later stage.The results analyzed and processed are on all research subjects in each cycle and also each treatment, namely on each different cane handle, each change in position and each change in angle that occurs when testing.

Angle Change
Determination of the angle at the elbow is done using Kinovea software which can display the angle formed when using a cane.The angle of the elbow that will be formed and displayed on the Kinovea software is first carried out a workzone on the anterior milliseconds of time in the first cycle to the posterior in the last cycle according to the milliseconds of time change data in the previous stage described in the figure below.

Fig. 5. Workingzone Kinovea
The workzone feature is done to provide a time limit used in determining the angle for 3 cycles.After the time limitation is done, the angle is done using the angle feature in kinovea which is described in the figure below.

Fig. 6. Angle Feature Kinovea
Determination of the elbow angle in visual gait walking using a cane by marking the shoulder area of the wrist and the elbow as the fulcrum to form the elbow angle.The elbow angle formed in the visual walking gait using a cane always changes every millisecond depending on the change in the intended position.The angle changes that occur each time are described in Figure.

Fig. 7. Graph of the Three Cycles in the First Test
The graph has a general shape with 3 mountains and 2 valleys indicating that there are 3 cycles that make up the graph.The graph also shows that there is an increase and decrease in the value of the angle indicating that there is a change in angle every millisecond both from the position, the gap between cycles, and each cycle.The detailed values of the angles formed in the position changes both anterior, middle lateral and posterior in the fritz cane angle changes of cycle 1 are shown in this table.The results of angle calculation and analysis of gait walking using a cane obtained the angle comparison value at the initial and final angles at each position with the highest value at the initial angle of 144.348° and the lowest at the final position of 142.744° explained in the figure below.Preprocessing is done to remove noise by reducing the existing frequency with the bandpass filtering method with fcut high 5 and fcut low 400 and retrification to make absolute values so that comparisons can be calculated easily.The filtering and retrification results can be seen in the figure below.

Fig. 10. Filtering and Retrification
After filtering and retrification, the next step is smoothing and windowing.Smoothing uses the RMS envelope method with a windowing of 100 ms.The purpose of smoothing and windowing is to trim signals that are too high or low and remove noise to provide the value of the actual signal.The next step is to perform normalization using the maximum voluntary contraction (MVC) value of each subject which will be the input of the max min method which will be used for normalization.

Statistical Test
Statistical tests were conducted to determine differences between factors or interactions.Statistical testing uses parametric statistics which are first tested for normality and homogeneity.Normality tests using Kolmogorov Smirnov and homogeneity using Barlett were carried out on both angle and EMG data and the results showed that the data were normally distributed and homogeneous.After the data was proven normal and homogeneous, a paired t test using SPSS was conducted on the initial and final angles whether there was a change in angle or not.The results of the difference between the initial and final angles obtained a significant value less than α 0.05.

Table 4. Paired T-Test of Initial and Final Angles
After the pair t test, the three-way ANOVA test was continued using SPSS on EMG data for each treatment factor to obtain the interaction of each factor.

Table 5. Three-Way ANOVA of EMG Data
The test results show that the significance value in all groups is below α 0.05, meaning that the significance value is smaller than the confidence level of α 0.05, which indicates that each factor and each factor interaction affects the magnitude of the EMG load distribution value on the arm to the wrist.
The next step is to test the data correlation between angle data and EMG data.The results obtained are that the angle data and EMG data are correlated because the significance value is below α 0.05 but the correlation is negative.

Conclusion
The results of testing and data processing obtained that the angle changes that occur at the beginning and end of using kinovea.The angle change is then analyzed using paired t test statistics and it is found that the difference between the initial and final angles is significant so that the factor or parameter of angle change can be included in the interaction that affects the EMG value The results of testing and processing angle data obtained correlation between angle and EMG data but negative, which means that the greater the angle formed the smaller the EMG load distribution value from the arm to the wrist.In addition, the results obtained are that the Fritz stick with the posterior position and the initial angle has a minimal load distribution from the forearm to the wrist so that it can be a selection suggestion to the elderly.In addition to stick selection advice, it can also be a benchmark for further research such as stick design or further experimental design.

Fig. 9 .
Fig. 9. EMG RAW DataProcessing of EMG consists of several preprocessing and feature extraction using MATLAB.Preprocessing is done to remove noise by reducing the existing frequency with the bandpass filtering method with fcut high 5 and fcut low 400 and retrification to make absolute values so that comparisons can be calculated easily.The filtering and retrification results can be seen in the figure below.

Fig. 11 .
Fig. 11.Smoothing Windowing and Normalization Next step is to perform segmentation and feature extraction on each segmentation.Segmentation is done to get the value of each treatment factor, namely each change in position and angle which segmentation is obtained from the permilidetic time on Kinovea software data.The results of the segmentation are described in the Figure below.

Fig. 12 .
Fig. 12. Segmentation The results of each segmentation are then subjected to feature extraction to obtain the true value of each division segment.Feature extraction is done using the time domain root mean square method.The results of each factor are shown in Figure below.

Table 2 .
Position Changes in the First Cycle

Table 3 .
Angle Change in the Anterior Position