Brain signal changes of sensory cortex according to surface roughness of boneless corsets

Abstract

The tactile perception of fabric surface properties in the brain has always been a major research issue in the study of the contact comfort of fabrics. This study introduced the concept of PSC (percent signal change) obtained by fMRI (functional Magnetic Resonance Imaging) technology to explore the perceptual brain regions concerned with the roughness of the fabric surface. Firstly, the PSCs of human somatosensory cortexes, including the SI (primary somatosensory cortex) and SII (secondary somatosensory cortex), under the stimulation of boneless corsets with different surface roughness, were extracted and analyzed. As the surface roughness of the fabric increased gradually from smooth to rough, the brain region in which the maximum PSC occurred gradually transited from the SI to the SII, which indicated that the SI brain region paid more attention to the contact of smooth fabrics, while the SII brain region laid emphasis on the fine tactile perception of rough fabrics. Furthermore, the result of sub-regions in the SI and SII showed that, with the increase of the roughness of the fabric surface, the PSC of the OP2 (Operculum Parietal 2) brain region increased significantly, and the brain region concerned with the fabric surface tactile stimulation was gradually transferred from the slow adaptive sensory projection Brain Area 3a in the SI to the fast adaptive sensory projection Brain Area 1 in the SI, and finally moved to OP2 in the deep cortex of the SII.

Keywords

fabric surface tactile brain perception percent signal change primary somatosensory cortex secondary somatosensory cortex

When human skin is stimulated by contact with a fabric surface, contact and friction stimulate the tactile receptors in the skin, thus forming the action potential; the action potential is transmitted to the somatosensory centers of the brain along the nerve fibers; the conduction direction, amplitude and time information carry stimulus information of the fabric surface, resulting in the somatosensory cortex of the brain showing a series of response perception corresponding to the fabric surface contact.^1–3 The purpose of this was to use advanced biomedical technology, fMRI (functional Magnetic Resonance Imaging), to extract the human brain signal in the sensory cortex related to the perception of attention, BOLD-PSC or PSC (percent of blood oxygen level dependent signal change, or percent signal change), as a new quantitative evaluation index of the fabric surface roughness perception of brain signals, to make up for theoretical guidance defects^4,5and quantitative description loopholes^6–8 of the existing textile industry fabric sensory evaluation systems.

Because the current fabric contact comfort evaluation system only includes traditional psychological evaluation,⁹, physical evaluation,¹⁰ effector evaluation (e.g., heart rate,¹¹ skin temperature,¹² electrocatdiogram,¹³ electromyogrphy^14–16 and so on) and evaluation of simple brain perception using medical instruments with low spatial resolution, such as the spatial resolution of EEG (electroencephalogram), which is only about 5–15 cm¹⁷, none of them can achieve in situ and direct evaluation. However, the spatial resolution of fMRI using electromagnetic wave imaging could be up to 25–100 µm.¹⁸ In addition, in terms of the sensitivity and specificity of body sensation, as early as 1994, Hammeke et al.¹⁹ demonstrated that fMRI had sufficient sensitivity to locate the sensory cortex under the tactile stimulation at the 1.5 T magnetic field intensity. Currently, the sensitivity could be as high as 0.01–1 mm,¹⁸ and the specificity of body sensation can reach up to 88%. The most direct example is that for some fine sensory information, such as the sensory information of low-pressure stimulation in the forearm, can only be detected by fMRI.²⁰ Therefore, it is of great significance to study the mechanism of fabric surface contact comfort evaluation by using fMRI technology, a leading technology in the field of medical brain molecular imaging, to capture brain signals under the stimulation of fabric surface contact.

The so-called PSC was calculated through the time series, which could be simply equivalent to dividing the average value of a certain section (for block design) or the value of a certain point (for event-related design) of the entire time course by the average value of the entire time course, and multiplying by 100. This experiment was a block design experiment. Therefore, the average value of BOLD signal in a certain period (pressure stage) was selected for calculation. The calculation formula²¹ was

PSC = \frac{β_{task} \times max (HRF) \times 100}{β_{const}}

where PSC is the percent signal change,

β_{task}

is a signal estimate of a task time series in the region of interest (ROI),

max (HRF)

is the maximum regression of task events and

β_{const}

is an estimated constant value for the signal of the particular ROI within the entire range of the sequence.

Obviously, the PSC itself was calculated relative to the baseline (which could be the mean of the entire time period), and thus contained the concept of the baseline itself.^22,23 Therefore, compared to the activation intensity obtained by the comparison calculation, the calculation of the PSC need not control the condition, which was more intuitive and understandable. The practical significance was that in the brain region, the BOLD signal values of all voxels in the task time period accounted for the percentage of BOLD signal values of all voxels in the whole time series. What was worth noting was that the PSC did not represent neural excitation, but merely the extent to which the stimulus condition affected the brain region.²⁴

Actually, there is a long history of taking advantage of the PSC to explore the corresponding correlation between various functional brain areas of human beings and human movements,^25,26 human perceptions,²⁷ memories,²⁸ semantic processings,²⁹ emotions,^30,31 etc. Rao et al.³² had made use of the PSC to discover the congruent relationship between the finger movement rate and the functional magnetic resonance signal change in the human primary motor cortex. In 1999, the PSC was even used as an evaluation standard to judge whether the behavior interleaved gradients technique could interfere with hearing. In recent years, the PSC has been involved in the research field of the fabric contact case more and more. The most representative case among them is that in which Wang et al.³³ compared the PSCs generated by two fabrics with significantly different physical and mechanical characteristics, namely, silk and linen, and found that the brain region for identifying microscopic properties of fabrics (such as roughness) was located in the secondary somatosensory cortex (SII), while the brain region for identifying the macroscopic properties of fabrics (such as pliability) was located in the primary somatosensory cortex (SI), which provided evidence to guide the introduction of PSC indicators into this subject.

Materials and methods

Subjects

Seven female volunteers without history of mental illness or metal implantation were selected as the volunteers for this project. All of them were in good physical health and had a similar age (mean age = 25 ± 2 years old) and a similar body shape (mean height =1.54 ± 0.05 cm, mean weight = 48.74 kg, mean body mass index = 20.6 ± 1.11 kg/m², mean belly fat thickness = 0.95 ± 0.18 cm, mean lower chest circumference = 74 ± 2.33 cm, mean waist circumference =68.57 ± 2.13 cm, mean abdominal circumference =73.86 ± 3.09 cm). Due to the fact that tactile perception would be influenced by memory, personality, expectation and other psychological factors,^34,35 after full knowledge and understanding of fMRI scanning, they accepted simple training in scanning procedures for this project, and eventually they all signed informed consent voluntarily. The study was approved by the ethics committee of Donghua University.

Materials and apparatus

Three types of ordinary metal-free boneless corsets were used as the experimental samples. The length and width of the corsets were 50 and 25 cm, respectively. When the fixed elongation was 20%, the elastic recovery rates were 94.75%, 91.45% and 93.19%, respectively, which proved all of them had a good elastic recovery. The two ends of the samples were a hook and a loop, respectively. The experimental corsets could obtain continuous and uniform changes in clothing pressure by stretching or relaxing.

A YG141N digital fabric thickness gauge was utilized to measure fabric thickness and a YG026MB - 250 universal electronic fabric strength tester was applied to test tensile properties, load-bearing tensile properties, elongation deformation properties, stress relaxation properties, fatigue resistance and bursting properties of the fabrics. The KES-FB System was used to measure bending properties, surface friction properties and touch feelings of warmth or coolness of the fabrics. Clothing pressures were measured by an AMI 3037 Air-pack Type Contact Pressure Measurement System, ranging from 0 to 34 kPa, output voltage 0–3.4 V, accuracy ± 0.2–0.45 kPa. FMRI scans were conducted on an Ingenia 3.0T medical fMRI equipment. Time of echo = 30 s, time of repetition = 3 s and layer thickness = 3 mm in functional and structural images and the total functional and structural images scanning times were 190 s (pre-scanning time = 10 s and scanning time = 180 s) and 300 s, respectively. The 3D-GRE T1WI sequence structure image scanning was from left to right.

Tests of basic performance and physical and mechanical properties of fabric samples

Firstly, the basic structure and surface morphology of fabrics in the transverse stretching direction were observed with the handheld USB sample 2.0MP Microscope, and the magnification was 40 times. The physical and mechanical properties of fabrics were tested in the constant temperature (20 ± 2℃) and humidity (65 ± 5% relative humidity (RH)) physics laboratory of Donghua University. All samples were pre-conditioning for 24 h before testing. Tensile testing was proceeded according to GB/T39231:2013 The first part of the fabric tensile property —— the measurement of fracture strength and elongation at break (strip method).³⁶ Five samples were taken for each sample fabric, with an effective size of 10 cm × 5 cm, 150 mm away from the cloth edge. The rising speed was 100 mm/min and the descending speed was 300 mm/min. According to FZ/T 70006:2004 Test method for elastic resilience of knitted fabric,³⁷ the elongation at constant force, the force at constant elongation, the elastic recovery and plastic deformation rate of repeated tension under constant load, the elastic recovery and plastic deformation rate at constant elongation and repeated stretching, the stress relaxation rate and the dynamic fatigue resistance were measured. According to the method shown in GB/T 19976:2005 Steel ball method for determination of textile bursting strength,³⁸ the measurements of bursting performance of the fabric samples were conducted.

FMRI scanning

As for fMRI experiments, a block design³⁹ was adopted. Each subject was asked to lie flat with her eyes closed but remaining awake. To avoid the interference of noise, the subjects were also given earplugs. Because the comfort contact pressure of the human waist would be obtained after stretching fabric 20%,⁴⁰ after resting for 30 s, 20% stretch fabric surface tactile stimulation to the waist of the human body was applied and lasted for 30 s, repeating each process three times.

Data analysis

SPM12 (Statistical Parametric Mapping) was used for image preprocessing, individual analysis and group analysis.⁴¹ Anatomy was used to select the brain region as a ROI.⁴² The so-called ROI referred to a mask file, which was used to filter out and remove all the activated areas that were not in this brain region. The remaining activation clumps within the mask brain region were analyzed.⁴³ In other words, the entire analysis of this topic were only conducted in this ROI. Finally, MarsBaR software was utilized to extract the PSC of all subjects under different fabric contact pressures.

Results and discussion

Results of basic performance and physical and mechanical properties of fabric samples

When the fabric came into contact with the human body, the first thing that affected the human body was the stimulation from the surface of the fabric. The observation results of the fabric samples are shown in Figure 1.

As can be seen clearly from Figure 1, the fabric structure of Sample 1 was of flat knit organization with high density and fine yarn, so the fabric surface was very smooth and basically hairless. Sample 2 was also weft knitted fabric with the yarns arranged neatly, but the yarns were thick and dense, so the surface was medium rough and slightly hairy. The fabric of Sample 3 was warp knitted with a relatively sparse arrangement density, and the pores between the yarns could be clearly seen after magnification by 40 times, and the yarns were very thick, so the surface was very rough.

Comparison results of basic performance and physical and mechanical properties of fabric samples

Considering that the physical and mechanical properties of the fabric would also affect and interfere with the tactile perception of the fabric to some extent, a summary and comparison table of the physical and mechanical properties of various sample garments obtained by the test is given in Table 1.

Table 1.

Summary and comparison table of basic performance and physical and mechanical properties of samples

Fabric Properties		Sample 1	Sample 2	Sample 3	Sample 1/Sample 2	Sample 1/Sample 3
Basic	D/mm	1.83	2.06	1.5	100/113	122/100
	Pt/(yarn/10 cm)	370	260	130	142/100	285/100
	Pw/(yarn/10 cm)	370	210	100	176/100	370/100
	W/(g/m²)	608.33	711.11	638.33	100/117	95/100
Surface	MIU	0.1205	0.2675	0.423	100/222	100/351
Surface	MMD	0.0055	0.0185	0.0197	100/336	100/358
	SMD/µm	1.005	3.59	9.895	100/357	100/985
	Q_max/( $J / (c m^{2} \cdot s)$ )	0.1383	0.0843	0.0713	164/100	194/100
Tensile	P_b/(N/mm)	6.24	2.00	4.44	312/100	141/100
	ε_b/%	254.14	312.52	223.58	123/100	114/100
	E₁/(N/mm²)	0.22	0.20	0.17	100/110	129/100
	E₂/(N/mm²)	4.70	1.59	5.56	100/296	100/118
Constant load	ε/%	91.3	229.1	150.1	251/100	100/164
Constant load	Re/%	90.97	89.54	92.25	100/120	99/100
Constant elongation	F/N	3.33	4.23	5.20	127/100	100/156
Relaxation	R/%	7.57	9.33	6.20	123/100	122/100
Bending	B/(gf·cm²/cm)	0.5609	0.6867	0.3146	122/100	178/178
Bending	2HB/(gf·cm/cm)	0.3268	0.4935	0.6002	151/100	100/184

Note: see the Appendix for the full abbreviation.

It is not difficult to see from Table 1 that the surface performance of the samples showed the most striking difference compared with the other physical and mechanical properties. The surface friction performance parameters of Sample 2, MIU, MMD, SMD, reached 2.22, 3.36 and 3.57 times those of Sample 1, respectively. Furthermore, comparing Sample 1 with Sample 3, the differences of surface properties were more obvious. The values of MIU, MMD and SMD of the latter reached 3.51, 3.58 and 9.85 times those of the former, respectively. Although the Q_max values of the cold–hot sensation parameter of clothing in contact with the skin of Samples 2 and 3 were 1.64 times and 1.94 times that of Sample 1, respectively, which was still largely caused by the smoothness of the fabric surface, the fabric with a high surface smoothness would produce a higher contact cold sensation.⁴⁴ In addition, even for the most cold-tactile Sample 1 fabric, the Q_max value was only 0.1383 $J / (c m^{2} \cdot s)$ , which was very close to that of wool fabric, 0.134–0.159 $J / (c m^{2} \cdot s)$ .⁴⁵ So the warm and cold touch of the three samples produced little stimulation to interfere with the effect of surface roughness on brain signals.

Production results of ROIs (including ROI-SI and ROI-SII)

In order to avoiding the interference effect of other brain regions, the brain regions of SI and SII were made into ROIs, respectively. All analyses were only performed in the ROI, rather than in the whole brain. The ROI of the SI and SII shown in Figure 2 was made, calculated and depicted by Anatomy, MarsBaR⁴⁶ and BrainNet Viewer,⁴⁷ respectively.

The definition diagrams of ROI-SI and ROI-SII are as shown in Figure 2; coordinates of the center of mass in the MNI space were (2.15, 28.3, 49.2) and (1.67, 16.6, 14.8), respectively. Voxel sizes were 32,992.00 and 15,494.00 mm, respectively. The maximum/minimum in the X, Y, Z directions were (14/80, –65/67, –52/–1) and (–68/70, –33/4, 3/33), respectively.

Results and discussion of PSCs in ROIs

Limited by experiment time and interference by individual differences, in order to avoid excessive experimental errors, the extraction results of the PSC of seven subjects were presented one by one, instead of using the average method for analysis. By eliminating individual outliers, the experimental results of most subjects were selected for analysis. The Extraction results of PSCs in ROI-SI and ROI-SII are shown in Figure 3, Table 2 and Figure 4.

Table 2.

The results of extracting the percent signal changes (PSCs) of regions of interest stimulated by fabric samples

Subjects	Sample 1		Sample 2		Sample 3
Subjects	PSC_SI	PSC_SII	PSC_SI	PSC_SII	PSC_SI	PSC_SII
1	0.197	−0.353	−0.006	−0.234	0.1510	0.3890
2	−0.641	−0.22	0.516	0.166	0.4520	0.9140
3	−1.467	−2.255	0.437	−1.333	0.4770	0.2680
4	0.258	−0.384	0.306	0.512	0.2240	0.4230
5	0.159	0.076	−0.715	0.054	−0.2330	−0.5390
6	0.237	0.374	0.99	0.596	0.4600	0.7050
7	0.095	0.563	−0.822	0.126	−0.0260	0.8190
Average	−0.166	−0.314	0.101	−0.016	0.215	0.426
Standard deviation	0.654	0.929	0.664	0.645	0.273	0.488

Table 2 shows the calculation results of the percent BOLD signal change in each brain ROI under the stimulus of smooth, medium rough and rough fabric surfaces, respectively. According to the comparative statistical results in Figure 3(a), the abscissa represented each subject, and the ordinate represented the calculated value of the PSC. It could be seen from the observation that the PSC in the brain ROI of the subjects was basically below 0.5% and tended to be zero or negative and, in most of the subjects, a small number of positive values appeared in both the SI and SII brain areas, and the SI had the larger PSC in the majority of subjects, which indicated that the brain regions of interest (ROI-SI and ROI-SII) of different subjects paid little attention to the fabric surface tactile stimulation generated in the waist of smooth fabric, which only had a minimal impact on the SI and SII brain regions, and has a greater impact on the SI brain regions. In other words, most of somatosensory cortexes were almost unaffected and only a small amount of tactile sensory stimulation occurred in the brain regions of the SI under the condition of relatively comfortable and smooth fabric stimulation.

Compared with smooth fabric surface stimulus, as can be seen from Figure 3(b), which shows the results under the fabric stimulation with a medium rough surface, what stayed the same was that the PSCs in ROI-SI were still the maximum in most of subjects, indicating that the SI brain region still paid the most attention to the sensation of fabric tactile stimulation. However, there were two differences. Firstly, in addition to the SI, the PSCs of the SII showed more positive values. Secondly, the PSC values of the two ROIs were increased to about 0.5%. These two facts suggested that the subjects' SII began to pay more and higher attention to the surface stimulation. Surface properties of the fabric played a significant role in the tactile capture process as all parts were within the comfort threshold. In addition, the rougher surface received more attention from more brain regions.

From the histogram in Figure 3(c), all of the PSCs generated by the tactile stimulation of the fabric with a very rough surface under the same tensile action were positive in both the ROI-SI and ROI-SII of most subjects, which meant that it had gained the most and greatest attention. However, what is worth noting is that ROI-SI was no longer in the dominant role, but rather ROI-SII. That is, with the continuous increase of fabric surface roughness, the attention of all interested brain regions to the tactile perception of the fabric also increased, and the attention point from the SI brain region gradually transferred to the SII brain region. It could be inferred that for fabrics with a rougher and warmer surface, the SII brain region paid more attention, while for fabrics with a smoother and cooler surface, the SI brain region signals were most affected and played a dominant role.

Figure 4 shows that there are changes in the percentage of signal changes between ROI-SI and ROI-SII under the stimulation of the fabric with increasing contact surface roughness. When the surface roughness of the fabric was very small, in other words, the surface was smooth; the black line that represents ROI-SI is higher than the red line that represents ROI-SII, indicating that the SI paid more attention to smooth fabric than the SII. However, when fabric surface roughness SMD > 6 µm, the result was the opposite of the former; the PSC in ROI-SII was significantly elevated, which exceeded ROI-I, indicating that the SII was the dominant brain region concerning rough fabric tactile, rather than the SI.

Production results of sub-ROIs

In order to further make clear the brain area corresponding to fabric surface tactile perception, in the process of the extraction of the PSC, according to probabilistic cytoarchitecture and the functional imaging data method,⁴² the two brain regions of interest, ROI-SI^48,49 and ROI-SII,^50,51 were divided into four sub-ROIs,^52–54 respectively. The details are shown in Table 3 and Figure 5.

Table 3.

Information table of sub-regions of interest (ROIs)

ROI	Sub-ROI	Center of mass	Volume (mm)	Max/min X (mm)	Max/min Y (mm)	Max/min Z (mm)
SI	BA1	(5.86,–27.7,54.4)	10,227	−65/67	−48/–6	18/80
	BA2	(2.19,–35.9,51.5)	9393	−62/59	−52/–16	32/71
	BA3a	(–2.99,–23.4,39.9)	3897	−58/58	−44/–2	15/70
	BA3b	(0.216,–23.6,45)	9475	−62/67	−48/–1	14/78
SII	OP1	(2.32,–22.6,16)	6154	−68/70	−33/–12	7/26
	OP2	(7.3,–23.1,17.2)	1168	−42/42	−31/–16	12/21
	OP3	(8.03,–12.2,15.1)	2793	−52/52	−20/–2	6/21
	OP4	(–3.59,–10.5,12.8)	5379	−68/70	−20/4	−3/33

Note: see the Appendix for the full abbreviation.

The results and discussion of PSCs in sub-ROIs

By calculation, the PSCs of each sub-ROI (including sub-ROIs-SI and sub-ROIs-SII) under various sample stimulations were shown in the Figure 6, Table 4 and Figure 7.

Table 4.

Results of percent signal changes (PSCs) in sub-regions of interest (ROIs) of Sample 1

Subjects	PSC_SI				PSC_SII
Subjects	PSC_BA1	PSC_BA2	PSC_BA3a	PSC_BA3b	PSC_OP1	PSC_OP2	PSC_OP3	PSC_OP4
1	0.068	0.177	0.343	0.31	−0.325	−0.477	−0.386	−0.341
2	−0.83	−0.758	−0.189	−0.485	−0.295	−0.26	−0.132	−0.16
3	−2.285	−1.486	−0.303	−1.089	−2.872	−2.328	−1.137	−2.083
4	0.585	0.165	0.044	0.139	−0.384	−0.47	−0.843	−0.062
5	0.058	0.114	0.456	0.199	0.248	−0.587	−0.189	0.176
6	0.142	0.225	0.419	0.265	0.247	0.506	0.51	0.429
7	−0.137	0.24	0.34	0.107	0.391	0.33	0.703	0.77
Average	−0.343	−0.189	0.159	−0.079	−0.427	−0.469	−0.211	−0.182
Standard deviation	0.955	0.672	0.308	0.519	1.124	0.922	0.665	0.918

The PSCs of each sub-ROI under smooth Sample 1 stimulation were shown in Table 4 and Figure 7.

As shown in Figure 7(a), in sub-ROIs-SI, except when the PSCs of the second and third subjects were negative when stimulated by Sample 1 with a smooth surface, the PSCs of the other subjects were very small, which were basically below 0.5%. Besides, the largest PSCs of almost all subjects except subject 4 were located in Brain Area 3a (BA3a), which can also be verified by Figure 6. In contrast to the PSCs of the SII in Figure 7(b), only a small number of signal changes occurred in the SII of subjects 6 and 7, suggesting that the SII of most subjects paid no attention to contact stimulation from smooth fabrics, but only slow adaptation sensory BA3a in the SI showed slight attention.

The PSCs of each sub-ROI under medium rough Sample 2 stimulation are shown in Table 5 and Figure 8.

Table 5.

Results of percent signal changes (PSCs) in sub-regions of interest (ROIs) of Sample 2

Subjects	PSC_SI				PSC_SII
Subjects	PSC_BA1	PSC_BA2	PSC_BA3a	PSC_BA3b	PSC_OP1	PSC_OP2	PSC_OP3	PSC_OP4
1	0.057	0.028	0.073	−0.142	−0.27	−0.175	−0.2	−0.218
2	2.647	−0.244	−0.397	−0.021	0.303	0.201	−0.472	0.332
3	1.573	0.106	−0.34	−0.027	−1.05	−1.638	−1.843	−1.328
4	−0.569	0.559	0.901	0.583	0.461	0.313	0.63	0.557
5	−0.743	−0.886	−0.379	−0.636	−0.141	−0.435	0.107	0.447
6	1.542	0.89	0.726	0.686	0.759	0.592	0.428	0.471
7	−1.67	−0.411	−0.225	−0.541	−0.033	−0.083	0.033	0.451
Average	0.405	0.006	0.051	−0.014	0.004	−0.175	−0.188	0.102
Standard deviation	1.548	0.596	0.547	0.505	0.588	0.728	0.817	0.681

According to Figures 6, 8(a) and 8(b), for fabric stimulus with moderate surface roughness, the SI brain area was still dominant, especially in fast adaptive sensory projection Brain Area 1 (BA1), where the largest PSC was up to 2.5%, but the SII brain area also appeared to have a percentage change of about 0.5%, especially in the superficial Operculum Parietal 1 (OP1) and OP4 brain areas of the surface of the parietal lobe. This was because even though the most affected brain region of a slightly rough warm fabric stimulation on the human was still located in the SI, especially in the rapid adaptation of BA1, it also had a certain effect on the superficial cortex of the SII brain area.

The PSCs of each sub-ROI under rough Sample 3 stimulation are shown in Table 6 and Figure 9.

Table 6.

Results of percent signal changes (PSCs) in sub-regions of interest (ROI) of Sample 3

Subjects	PSC_SI				PSC_SII
Subjects	PSC_BA1	PSC_BA2	PSC_BA3a	PSC_BA3b	PSC_OP1	PSC_OP2	PSC_OP3	PSC_OP4
1	0.0250	0.2870	0.3410	0.0790	0.3900	0.3980	0.4390	0.3580
2	0.7360	0.5010	−0.1110	0.3270	1.0090	0.3200	0.6760	1.0870
3	0.6630	−0.0250	0.8920	0.6520	0.3850	0.1810	0.1880	0.1760
4	−0.0710	0.4890	0.1090	0.2580	0.1820	0.8240	1.0340	0.2600
5	−0.2490	−0.2280	−0.2950	−0.1990	−0.4930	−0.1200	−0.1710	−0.9570
6	0.4890	0.5610	0.0940	0.4730	0.5470	0.7730	0.8390	0.8280
7	−0.2820	0.1180	0.1480	0.0380	0.9920	0.8560	0.6250	0.6930
Average	0.187	0.243	0.168	0.233	0.430	0.462	0.519	0.349
Standard deviation	0.432	0.300	0.378	0.286	0.513	0.371	0.407	0.663

According to Figures 6, 9(a) and 9(b), for the same fabric stretching, the tactile stimulation to human skin from the rough surface fabric would cause a more significant PSC in the SII brain regions, especially the OP2 and OP3 where were the largest signal response change located. Except for subject 5, all the PSCs of sub-ROIs-SII of the other subjects were positive, most of which were about 0.5%, while some of subjects (such as subjects 4 and 6) even reached nearly 1%. Moreover, the signal responses of sub-ROIs-SII of all the subjects were slightly higher than those of sub-ROIs-SI.

Correlation analysis

After the above individual analysis, a Pearson correlation analysis, shown in Table 7, was conducted on the physical and mechanical properties of the fabrics and the PSCs of ROIs and sub-ROIs to further explore the localization of the brain areas where the attention of the brain area changed with the surface properties of the fabric.

Table 7.

Pearson correlation coefficients between physical and mechanical properties and percent signal changes of each brain region of interest

Pearson	SI	SII	BA1	BA2	BA3a	BA3b	OP1	OP2	OP3	OP4
D	−.114	−.282	.055	−.208	−.114	−.250	−.252	−.377	−.441*	−.163
Pt	−.283	−.418	−.196	−.330	−.016	−.296	−.431	−.502*	−.441*	−.292
Pw	−.292	−.409	−.228	−.325	.004	−.279	−.429	−.483*	−.403	−.293
W	.145	.073	.267	.076	−.125	−.011	.124	.040	−.101	.094
MIU	.285	.417	.203	.329	.012	.293	.431	.499*	.434*	.293
MMD	.287	.355	.280	.289	−.049	.220	.390	.403	.287	.269
SMD	.263	.416	.150	.324	.041	.306	.419	.508*	.473*	.282
Q _max	−.292	−.377	−.267	−.304	.035	−.242	−.407	−.434*	−.328	−.280
P _b	−.181	−.134	−.284	−.122	.117	−.034	−.184	−.115	.029	−.134
ε_b/%	−.034	−.185	.135	−.129	−.129	−.191	−.145	−.267	−.365	−.088
E ₁	−.277	−.419	−.180	−.329	−.025	−.300	−.429	−.506*	−.454*	−.290
E₂	−.006	.131	−.170	.086	.133	.156	.087	.206	.317	.049
ε	.182	.135	.285	.123	−.117	.035	.185	.116	−.028	.135
Re	.077	.238	−.094	.172	.123	.224	.202	.328	.409	.128
F	.285	.417	.202	.329	.013	.293	.431	.499*	.435*	.293
R	−.065	−.224	.105	−.160	−.125	−.215	−.187	−.312	−.397	−.118
B	−.136	−.306	.031	−.228	−.108	−.264	−.280	−.404	−.457*	−.182
2HB	.292	.407	.232	.324	−.007	.277	.428	.480*	.397	.292

The correlation was significant at the 0.01 level (bilateral).

The correlation was significant at the 0.05 level (bilateral).

As can be seen from Table 7, among all ROIs and sub-ROIs, only the PSCs of the OP2 and OP3 brain regions of the SII were significantly correlated with many physical and mechanical properties of the fabrics at the level of 0.05 or above. Among the physical factors that had significant correlation with the PSCs of the OP2 and OP3 brain regions, the surface roughness SMD of the fabric had the largest correlation. Therefore, it could be concluded that, compared with the surface roughness of the fabric, other physical and mechanical properties had less influence on the change of the brain perception signal, while the surface roughness played a major role.

Significance test

According to the correlation analysis, the surface roughness of the fabric was the main factor affecting the change of brain signal; thus, a one-way ANOVA (analysis of variance) significance test based on SPSS software was performed on the above three groups of PSC data and SMD data. After the transformation of the Z score standardization method, the test results of the homogeneity of variance are as shown in Table 8. From Table 8, p > 0.05 of the PSCs in all brain regions meant homogeneity of variance, and the one-way ANOVA method could be used for the significance test. The test results are shown in Table 9. From Table 9, only p_OP2 = 0.039 < 0.05, that is, the PSC of the OP2 brain area was significantly affected by the surface roughness of the fabric, but the effect of surface roughness on signal changes in other brain regions was not significant. It showed that the PSC of the OP2 brain region always changed significantly with the surface roughness of the fabric, whether the surface was smooth or not. However, for BA3a and BA1, which correspond to smooth and medium rough perception respectively, the signal changes were not significant in the final statistical results because they paid little attention to the perception of contact stimulus on the rough fabric surface.

Table 8.

Homogeneity test of variance

PSC	Levene	df1	df2	Significance
PSC_SI	3.288	2	17	.062
PSC_SII	.234	2	17	.794
PSC_BA1	5.409	2	16	.161
PSC_BA2	1.196	2	17	.327
PSC_BA3a	1.533	2	18	.243
PSC_BA3b	1.143	2	17	.342
PSC_OP1	.263	2	17	.772
PSC_OP2	.444	2	16	.649
PSC_OP3	.909	2	17	.422
PSC_OP4	.360	2	17	.703

Table 9.

Results of one-way analysis of variance

PSC	Sum of squares	df	Mean square error	F	p
PSC_SI	11.898	17	.700	.216	.808
PSC_SII	8.523	17	.501	1.524	.246
PSC_BA1	9.295	16	.581	.114	.893
PSC_BA2	11.326	17	.666	.589	.566
PSC_BA3a	19.640	18	1.091	.165	.849
PSC_BA3b	12.180	17	.716	.746	.489
PSC_OP1	6.183	17	.364	1.734	.206
PSC_OP2	4.209	16	.263	4.004	.039
PSC_OP3	8.859	17	.521	3.586	.050
PSC_OP4	10.923	17	.643	.337	.718

Comparing the above results, it was not difficult to find that with the increase of fabric surface roughness, the brain area with the most attention gradually changed from the slow adaptive sensory projection BA3a in the SI to the fast adaptive sensory projection BA1, and finally transferred to the OP2 and OP3 brain areas in the SII. From the degree of signal change, when the fabric surface was smooth or a little rough, the SI and SII signal changed in very slightly, just below 0.5%. With the increase of surface roughness, the SII brain regions, especially signal changes in the OP2 regions, were also gradually increased significantly, that is, the surface roughness of fabric increased tactile awareness of OP2, which was due to the top in the superficial part being activated by distinguishing the roughness.⁵⁵ In addition, the OP2 and OP3 regions, which are located in the deep cortex of the SII, were more closely associated with the region responsible for basic sensorimotor processing and motor control, while the OP1 and OP4 regions, which are located in the superficial cortex of the SII, were more closely associated with the parietal network for higher order somatosensory processing.⁵⁶

Conclusion

In conclusion, the relationship between the signal response of the somatosensory cortex of the human brain under fabric surface tactile stimulation and the surface roughness of the fabric was explored by fMRI technology in this study. It was concluded that the brain region in which the maximum PSC occurred gradually transited from the SI to the SII as the surface roughness of the fabric changed from smooth to rough gradually, indicating that the SI brain region paid more attention to the tactile perception of smooth fabrics, while the SII brain region laid emphasis on the fine tactile perception of rough fabrics. In addition, the traditional probabilistic cytoarchitecture and functional imaging data method was utilized to subdivided the SI and SII brain regions into the sub-regions of BA1, BA2, BA3a, BA3b and OP1, OP2, OP3,OP4, respectively. The result showed that most of the somatosensory cortex paid little attention to the contact stimulation of smooth fabrics; only the slow adaptive projection BA3a in the SI showed slight attention. The same method was used to obtain the conclusion that BA1 of the fast adaptive sensory projection brain area paid more attention to the medium rough surface fabric tactile stimulation and that the rough fabric surface tactile stimulation had a higher influence on the OP2 and OP3 brain areas, which are located in the middle parietal cortex of the secondary sensory cortex, especially for the cortex OP2, which would be obviously affected by fabric roughness.

Footnotes

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship and/or publication of this article: This work was supported by Exemplary Projects of the Research on Mechanism of Body Injury in the Disaster Environment and Research and Development of Rescue Protection Technology and Equipment (Grant No. 2016YFC0802800) and the Natural Science Foundation of Shandong Province, China (Grant No. ZR2017BEM041).

References

Johnson

Hsiao

. Neural mechanisms of tactual form and texture perception. Ann Rev Neurosci 1992; 15: 227–250.

Srinivasan

LaMotte

. Tactual discrimination of softness: abilities and mechanisms. In: O. Franzen, R. Johansson and L. Terenius. Berlin (eds). Somesthesis and the neurobiology of the somatosensory cortex, Springer, 1996, pp. 123–135.

Chen Si. Research on the mechanism of skin friction tactile perception. PhD. China University of Mining and Technology, 2016.

Pan N. Measuring Fabric Sensory Attributes: Theory and Practice[J]. 2019.

Yuan

Chen

, et al. A potential new fabric evaluation approach by capturing brain perception under fabric contact pressure. Text Res J. 89: 3312–3325.

Pan

. Quantification and evaluation of human tactile sense towards fabrics. Int J Des Nat Ecodynam 2006; 1: 48–60.

Pan D. Roughness based on texture composition of fabric surface. Master's Thesis. Donghua University, 2011.

Roberto

B-S

Rocio Alfonsina

L-M

Roberto Alexander

B-D

. Intelligent neuro-fuzzy fabric evaluation system: a novel multi-dimensional stochastic fuzzy system and a generator of training patterns for an artificial neural network. Int J Ind Eng Theor Appl Pract 2018; 25: 215–229.

Meng

. Study on contact pressure of elastic fabrics and establishment of contact pressure comfort evaluation system, Shanghai: Donghua University, 2010.

10.

Peirce

. The “handle” of cloth as a measurable quantity. J Text Inst Trans 1930; 21: T377–T416.

11.

Wang

Chen

. Method of assessing clothing comfort from the perspective of heart rate. Shandong Text Technol 2008; 49: 36–39.

12.

Sugita

Okabe

Kioka

. Effects of girdles on comfort of aged women–Results from early responses of a heart rate, a skin temperature and an electroencephalograph. J Jpn Res Assoc Text End Use 2002; 43: 33–44.

13.

Camillieri

Bueno

M-A

Fabre

, et al. From finger friction and induced vibrations to brain activation: Tactile comparison between real and virtual textile fabrics. Tribol Int 2018; 126: 283–296.

14.

Yick

Tse

, et al. Effects of indoor slippers on plantar pressure and lower limb EMG activity in older women. Appl Ergon 2016; 56: 153–159.

15.

Guo J, Yu S, Li Y, et al. A soft robotic exo-sheath using fabric EMG sensing for hand rehabilitation and assistance. In: 2018 IEEE International Conference on Soft Robotics (RoboSoft), Livorno, Italy, 24–28 April 2018, pp.497–503. New York.

16.

Karabulut

Ortes

Arslan

, et al. Comparative evaluation of EMG signal features for myoelectric controlled human arm prosthetics. Biocybern Biomed Eng 2017; 37: 326–335.

17.

Huettel

Song

McCarthy

. Functional magnetic resonance imaging, Sunderland, Massachusetts U.S.A, MA: Sinauer Associates, 2004.

18.

Liu

, et al. Molecular imaging technology in medical imaging. Imag Sci Photochem 2018; 36: 359–366.

19.

Hammeke

Yetkin

Mueller

, et al. Functional magnetic resonance imaging of somatosensory stimulation. Neurosurgery 1994; 35: 677–681.

20.

Schirmer

Adolphs

. Emotion perception from face, voice, and touch: comparisons and convergence. Trend Cognit Sci 2017; 21: 216–228.

21.

Glascher

. Visualization of group inference data in functional neuroimaging. Neuroinformatics 2009; 7: 73–82.

22.

Liu

Nalci

Falahpour

. The global signal in fMRI: nuisance or information?. Neuroimage 2017; 150: 213–229.

23.

Kang

Sung

Y-W

Shioiri

. Estimation of physiological sources of nonlinearity in blood oxygenation level-dependent contrast signals. Magnet Reson Imag 2018; 46: 121–129.

24.

Ellis

Lesnick

Henselman-Petrusek

, et al. Limitations of topological data analysis for event-related fMRI. bioRxiv 2018; 10.1101: 457747.

25.

Neyedli

Sampaio-Baptista

Kirkman

, et al. Increasing lateralized motor activity in younger and older adults using real-time fMRI during executed movements. Neuroscience 2018; 378: 165–174.

26.

Raffin

Richard

Giraux

, et al. Primary motor cortex changes after amputation correlate with phantom limb pain and the ability to move the phantom limb. Neuroimage 2016; 130: 134–144.

27.

Wang

, et al. Investigation of the cortical activation by touching fabric actively using fingers. Skin Res Technol 2015; 21: 444–448.

28.

Manoach

Schlaug

Siewert

, et al. Prefrontal cortex fMRI signal changes are correlated with working memory load. Neuroreport 1997; 8: 545–549.

29.

Sims

Kapse

Glynn

, et al. The relationships between the amount of spared tissue, percent signal change, and accuracy in semantic processing in aphasia. Neuropsychologia 2016; 84: 113–126.

30.

Young

Siegle

Zotev

, et al. Randomized clinical trial of real-time fMRI amygdala neurofeedback for major depressive disorder: effects on symptoms and autobiographical memory recall. Am J Psychiatr 2017; 174: 748–755.

31.

Young

Siegle

Misaki

, et al. Altered task-based and resting-state amygdala functional connectivity following real-time fMRI amygdala neurofeedback training in major depressive disorder. NeuroImage Clin 2018; 17: 691–703.

32.

Rao

Bandettini

Binder

, et al. Relationship between finger movement rate and functional magnetic resonance signal change in human primary motor cortex. J Cerebr Blood Flow Metabol 1996; 16: 1250–1254.

33.

Wang

Chen

, et al. Brain discriminative cognition on the perception of touching different fabric using fingers actively. Skin Res Technol 2016; 22: 63–68.

34.

Johnson

Yoshioka

Vega–Bermudez

. Tactile functions of mechanoreceptive afferents innervating the hand. J Clin Neurophysiol 2000; 17: 539–558.

35.

de Lafuente

Romo

. Neuronal correlates of subjective sensory experience. Nat Neurosci 2005; 8: 1698.

36.

GB/T39231:2013. Textile fabrics - tensile properties - part 1: determination of tensile strength at break and elongation at break (strip method).

37.

FZ/T 70006:2004. Test method for stretch elastic recovery rate of knitted fabrics.

38.

GB/T 19976:2005. Textile bursting strength - determination of steel ball method.

39.

Zhang

, et al. FMRI comparative study of hand motion block design and event-related design. Radiol Pract 2006; 1: 5–8.

40.

Yuan

Wang

, et al. A potential brain zone perceiving a comfortable fabric pressure touch. Text Res J. 89: 3499–3505.

41.

Ashburner

Barnes

Chen

, et al. SPM12 manual, London: Wellcome Trust Centre for Neuroimaging, 2014.

42.

Eickhoff

Stephan

Mohlberg

, et al. A new SPM toolbox for combining probabilistic cytoarchitectonic maps and functional imaging data. Neuroimage 2005; 25: 1325–1335.

43.

Zhu Y. Image cognition research based on functional magnetic resonance imaging (FMRI). Dalian Maritime University, Dalian, 2013.

44.

Sun Y. Analysis of cold feeling of fabric contact and its influencing factors. Cotton Text Technol 2009; 10: 18–21.

45.

Yang

Yao

, et al. Effects of fiber type and fabric structure on cold and warm feeling of contact. Cotton Text Technol 2015; 43: 45–48.

46.

Brett M, Anton J L, Valabregue R, et al. Region of interest analysis using an SPM toolbox. In: 8th international conference on functional mapping of the human brain. Sendai, Japan, 2–6 June 2002, 16(2): 497. Sendai.

47.

Xia

Wang

. BrainNet Viewer: a network visualization tool for human brain connectomics. PloS One 2013; 8: e68910.

48.

Geyer

Schleicher

Zilles

. Areas 3a, 3b, and 1 of human primary somatosensory cortex: 1. Microstructural organization and interindividual variability. Neuroimage 1999; 10: 63–83.

49.

Geyer

Schormann

Mohlberg

, et al. Areas 3a, 3b, and 1 of human primary somatosensory cortex: 2. Spatial normalization to standard anatomical space. Neuroimage 2000; 11: 684–696.

50.

Eickhoff

Amunts

Mohlberg

, et al. The human parietal operculum. II. Stereotaxic maps and correlation with functional imaging results. Cerebr Cortex 2006; 16: 268–279.

51.

Eickhoff

Schleicher

Zilles

, et al. The human parietal operculum. I. Cytoarchitectonic mapping of subdivisions. Cerebr Cortex 2006; 16: 254–267.

52.

Kurth

Eickhoff

Schleicher

, et al. Cytoarchitecture and probabilistic maps of the human posterior insular cortex. Cerebr Cortex 2009; 20: 1448–1461.

53.

Heimer

De Olmos

Alheid

, et al. The human basal forebrain. Part II. Primate Nervous Syst III Handbook Chemical Neuroanat 1999; 15: 57–226.

54.

Scheperjans

Hermann

Eickhoff

, et al. Observer-independent cytoarchitectonic mapping of the human superior parietal cortex. Cerebr Cortex 2008; 18: 846–867.

55.

Ledberg

Osullivan

Kinomura

, et al. Somatosensory activations of the parietal operculum of man - a pet study. Eur J Neurosci 1995; 7: 1934–1941.

56.

Eickhoff

Jbabdi

Caspers

, et al. Anatomical and functional connectivity of cytoarchitectonic areas within the human parietal operculum. J Neurosci 2010; 30: 6409–6421.