Introduction

Sitting is a frequent activity of the day that needs the postural control skill to maintain the body’s position in space, which is exhibited in the second trimester of age1. Technically, it is defined as the ability to maintain centre of gravity (COG) within the base of support (BOS)2,3,4. During sitting position in healthy adults, the lowest body load on seat cushion is 51.1% and 55.1% of body weight for normal and non-normal positions, respectively5. Inability to maintain their body in the proper position may expose physically disabled clients to abnormal postural alignment of the spine, pelvis, lower extremities and pressure sores6,7,8. Seat cushion modification is one of the postural support devices (PSD) that could provide firm support and comfortable weight distribution, and relieve pressure underneath the bony prominences during sitting8,9,10.

In the research on and clinical assessment of dynamic postural control, a pressure mapping device is used to measure the reaction between the human body loading and the support surface, known as the interface pressure, during static and dynamic seated postural balance11,12. It provides the data set of maximum pressure, average pressure, force, total contact area, the trajectory of the center of pressure (COP), etc. during static and dynamic states.

Commercial flexible pressure mapping systems are being increasingly used to assess seating. Examples are the Tekscan COMFORMat pressure mapping device (USA), used to investigate the effects of seat cushion contours and sitting posture on the seat pan interface pressure distribution and subjective comfort perception11 and the Force Sensory Array pressure mapping system (Canada), used to investigate the feasibility of using pressure mapping for the characterisation of the seated posture of children with cerebral palsy12. The Xsensor pressure mapping system (Canada) was used to determine the sensitivity of pressure mapping extracted variables from the pressure mapping analysis and postural trunk control functional test13. The Pliance seat sensor system (Germany) was used to predict unsatisfactory seating pressure in patients with spinal cord injuries14.

The originality of the research is to develop a pressure mapping device to assess weight distribution, which benefits improving postural alignment during sitting. The device is designed specifically for fabrication and fitting the seating in children with motor disabilities in pediatric rehabilitation centers in Thailand. Although the data of commercial pressure mapping devices are useful and provide high feasibility for evaluating the ability for postural control in real-time in seating assessment, commercial flexible pressure mapping systems are very expensive and hard to acquire in low-resourced countries. The force sensing resistors sensor (FSRs) is benefit for this research. It is piezoresistive sensors that change resistance depending on the force applied over the sensing area. FSRs are low-cost and reliable and are commonly used to measure reaction force in the flexible contact area15,16,17,18. Thus, this preliminary study aims to develop a reliable, accurate, and low-cost pressure mapping system to evaluate the force distribution for assessment and speed up the process of fabrication and fitting of seat cushion modification. The system will then be practical to use in pediatric rehabilitation centers.

Methods

Participants

This study investigated sample size of thirty volunteers whose hip width was less than 40 cm and who could sit still for at least one minute. None of the participants had musculosketal and neurological symptoms in the past six months. The ethics committee of the Mahidol University Institutional Review Board (COA No. MU-CIRB 2021/162.0707) approved the study protocol and informed consent was attained from all participants before the commencement of the study. The informed consent was obtained to publish the information and images in the online access publication.

Instrumentation

The prototype of the low-cost pressure mapping device is called a Smart mat (Fig. 1). It was developed by using force sensing resistors, FSRs, model 406 (Interlink Electronics, Inc., USA) which has a thickness of 0.2 mm and a width and length of 43.69 mm. For the FSR resolution, the voltage divider was used to amplify the output which ranges between 10 and 100 N. Sixty-four FSRs were arranged as the matrices of 8 rows and 8 columns on the 40 × 40 cm2 vinyl material. Each FSR was calibrated from 10 to 100 N for three times of increments of 10 N. The means of voltage output were used to find the equation of voltage output conversion through the Force Measurement Unit (IMADA, Inc., Japan) (Fig. 2). The graph of the calibrated force equation is shown in Fig. 3.

Fig. 1
figure 1

The developed pressure mapping device (Smart mat).

Full size image
Fig. 2
figure 2

The force measurement unit.

Full size image
Fig. 3
figure 3

Graph of calibrated force equation.

Full size image

Arduino IDE software for Windows app (Win 8.1 or 10) was used to write and upload the code to the eight Wemos Lolin 32 (Espressif Systems, China) for data acquisition. The Smart mat transmitted data through Wi-Fi to the laptop with a 20 Hz sampling rate. Microsoft Visual Studio C# program was used to create the graphical user interface (GUI) for displaying acquisition data in real time and saving the data as an Excel file on the laptop. The Smart mat could continuously collect data for up to 2 h by using the portable power bank 20,000 mAh.

Due to the non-availability of flexible pressure mapping devices in the SMART laboratory and Mahidol University campus, the MatScan model medical sensor 3150 (Tekscan, Inc., USA) in Fig. 4 was used as the reference standard pressure mapping in this study. It is a tactile resistive sensor commonly used to assess static and dynamic postural control by force and pressure measurement, contact area, etc. Thus, MatScan was used to investigate the validity and reliability of the pressure mapping device developed in this study. The system specifications of the Smart mat and MatScan are shown in Table 1.

Fig. 4
figure 4

The MatScan.

Full size image
Table 1 Pressure mapping systems specification.
Full size table

Experiments

The experiment setting was the SMART laboratory, Mahidol University, Salaya campus. Before the commencement of the study, the anthropometric parameters, including body weight, height, hip width, and thigh and leg length of each participant were measured and recorded. According to the different sizes of the sensing area of both devices, the study parameters were modified to the total force and ischial width during sitting, which were derived as follows: the total force is the summation of force from the whole interfaced sensors and ischial width is the distance between the highest force underneath each ischial tuberosity.

The following procedure is used for determining the validity of a Smart mat and MatScan and the reliability of a developed pressure mapping device.

Procedure for pressure measurement in sitting

The Smart mat and MatScan were installed on a chair with adjustable height. The participants were asked to sit still at 90-90-90 degrees on the experimental chair, the height of which was adjusted according to the leg length of the individual. In this procedure, data from both systems could not be collected simultaneously due to the bottom surface of the Smart mat being a hard-wired data cable that might damage the top surface of MatScan. Thus, the grid lines were used to indicate the position of participants in this procedure.

The total force underneath the contact area and the ischial width of both devices were recorded for 60 s per trial for 3 trials (Fig. 5). For controlling the sitting position, the grid lines were placed on the back and lateral sides of the participants and used to indicate the sitting position during the data recording of each device. When changing the device, the participants were asked to sit in the previous position by using the reference marker on the grid lines.

Fig. 5
figure 5

Procedure for pressure measurement in sitting (a) on Smart mat, (b) on MatScan.

Full size image

The estimated ischial width of MatScan was derived from the distance between the maximum force underneath the ischial tuberosities of both sides on the x-axis (Fig. 6a). Correspondingly, an estimated ischial width of the Smart mat was calculated from the number of sensors between two maximum forces multiplied by 3.69 cm (the width of sensing area of FSR 406) which is shown in Fig. 6b.1 and b.2.

Fig. 6
figure 6

The derived ischial width (a) via MatScan, (b) via Smart mat.

Full size image

The normalised ischial width (nIW) was derived from the estimated ischial width of both systems by dividing by the hip width.

Statistical analysis

Statistical analysis was performed using the SPSS software version 25.0 (SPSS, Inc., USA). The normality was tested using the Shapiro–Wilk tests. The concurrent validity of Smart mat and MatScan was determined by normally distributed data by Pearson correlation and non-normally distributed data by Spearman correlation. The validity of estimated values was negligible correlation below 0.10, weak correlation between 0.10 and 0.39, moderate correlation between 0.40 and 0.69, strong correlation between 0.70 and 0.89, and very strong correlation between 0.90 and 119. The reliability of the Smart mat was assessed by an intraclass correlation coefficient (ICC) between trials based on the 95 percent confidence interval (CI). The considered values of ICC are poor at below 0.5, moderate at between 0.5 and below 0.75, good at between 0.75 and 0.9, and excellent at above 0.920. The percentage of a coefficient of variation (%CV) was also used to estimate the precision of the developed devices by using the percentage of the ratio of the standard deviation to the mean. The estimated %CV is excellent at below 10%, good at between 10 and 20%, fair at between 20 and 30%, and poor at above 30%21. For comparing the similarity of the parameters between Smart mat and MatScan, the paired t-test was used to determine the normally distributed data and the Wilcoxon signed ranks test was used to determine the non-normally distributed data.

Results

The thirty participants comprised 13 males and 17 females, aged between 22 and 41 years (29.57 ± 4.50), with a body mass index (BMI) ranging from 16.77 to 25.17 kg/m2 (21.36 ± 2.13) (Table 2).

Table 2 Demographic and anthropometric data of participants.
Full size table

The reliability of the Smart mat

The reliability of the developed pressure mapping device showed that the ICC of nTF in sitting was excellent (0.93) with fair CV (25.20% CV). Similarly, the ICC of an nIW was excellent (0.98) as well with good CV (12.35% CV) (Table 3).

Table 3 The ICC with 95% CI and CV values of the parameters of the smart during sitting.
Full size table

The concurrent validity of Smart mat

A comparison of the nTF of Smart mat and MatScan in sitting showed normal distribution. The Pearson correlation was used to determine the concurrent validity of Smart mat. The correlation coefficients of nTF in the sitting had a significantly strong correlation (r = 0.7**).

A comparison of the data of nIWs between Smart mat and MatScan led to the rejection of the hypothesis of normality (P ≤ 0.05). The Spearman correlation was used to determine and the results reported that the nIW was of significantly moderate correlation (r = 0.67**).

For comparing the similarity of the parameters between Smart mat and MatScan in sitting, the paired t-test was used to determine the different mean of nTF in sitting because the data is the normal distribution. The results revealed that nTF is a statistically significant difference (P ≤ 0.05). On the other hand, the Wilcoxon signed ranks test was used to determine the different means of nIW which is not a normal distribution. The results showed that the nIW is a statistically significant difference (P ≤ 0.05).

Discussion

The reliability of nTF and nIW in sitting is excellent (0.93 and 0.98, respectively) but the relative scattering of nTF is not good CV (22.20% CV). The factors that may affect this variation are hysteresis and creep which could happen by prolonged constant load over the FSRs. Normally, force calibration can help to compensate for these two factors17,22. In this study, a Force Measurement Unit was used to calibrate each FSR sensor from 10 to 100 N with 10 N increments. Stinson et al.22 found the change in pressure value in the first 6 min of 20 min of measurement of constant load over the Force Sensing Array (FSA) pressure mapping system. They recommended 2 min of resting to prevent the creep phenomenon. To prevent the consequence of load constant during data collection, the participants were asked to rest through the data export of each trial. These procedures may cause variations in force distribution outcomes. As this limitation, when using the Smart mat should calibrate before the assessment and observe the changing data over time.

Chen and Yang identified the ischial width by two peak values of seated pressure that found the difference of the pelvic and hip angle may cause the variation of measurement outcome23,24,25,26. In this study, two grid lines and a 90-90-90-degree position were set to localize the standard position of the sitting experiment. The ischial width revealed the constant distance between trails because the control location of sitting in each trial could help to weight bearing on the same contact sensors.

Concerning the correlation coefficients of nTF had a significantly strong correlation (r = 0.7**) (Fig. 7) and the nIW had a significantly moderate correlation (r = 0.67**) (Fig. 8), there are limitations relating to differences in the resolution of Smart mat and MatScan systems. Whereas smaller sensors offer higher resolution, Smart mat might require a large number of sensors and increases production costs. In this study, FSR406 sensors were selected to invent 8 × 8 matrix of pressure mapping device because this model of FSR is affordable and can be designed to fill a Smart mat system with space-free in the budget of 40,000 Thai baht.

Fig. 7
figure 7

The mean total force of Smart mat and MatScan in sitting.

Full size image
Fig. 8
figure 8

The mean ischial width of Smart mat and MatScan in sitting.

Full size image

As the low resolution of the original FSR, the voltage divider was used to amplify the FSR resolution from 10 to 100 N. Although FSR can be utilized Smart mat as the desirable size, there are gaps between the FSR406 sensor’s sensing area that may affect the accuracy of results. From the literature reviews of FSR, Lana et al.27 applied twenty-seven sensors of FSR model 408 (340 mm × 340 mm) that revealed the results of the capability of the articulated seat to change the pressure distribution under the gluteal region. Kim et al.28 applied 128 sensors of FSR model 406 (189.9 cm × 90.9 cm) that could propose an algorithm that identifies six lying postures. According to the validity of the clinical trial of Smart mat29, minor gaps between sensors might not significantly affect the system's capability to guide the customization of seat cushions.

The estimated ischial tuberosity width could be estimated via flexible pressure mapping systems12,24,30,31. The peak values were used at resolution via commercial pressure mapping devices at 16 × 16 sensors of 24.5 mm × 24.5 mm (392 mm × 392 mm of sensor surface) that found the gap between the sensor of 2.69 cm could be estimated successfully ischial width in adults neither nor children12. As the resolution issue, the Smart mat should decrease the size and increase the number of sensors at the ischial tuberosities area to increase the resolution of the Smart mat.

The limitations of the preliminary study should be a concern. Firstly, the Smart mat and MatScan were asynchronous data collection which may affect the system's validity. The future study should collect data simultaneously. Secondly, the Smart mat has low resolution especially in the ischial tuberosity area when collecting the data in healthy adults. Further study should investigate clients with motor disabilities and Smart mat should optimize the resolution regarding the anthropometry of the clients.

Conclusion

In less-resourced countries, the well-known flexible pressure mapping systems are high-cost and difficult to acquire. As the results of this study showed, the Smart mat is a low-cost pressure mapping system that can be practically used as an assessment tool for force distribution in healthy volunteers. However, data collection methods for clients with motor disabilities require further investigation. Therefore, future studies will utilize a Smart mat to assess force distribution during sitting in clients with motor disabilities.