Ergonomic glove pattern drafting method for hand assistive devices: considering 3D hand dimensions and finger mobility

Abstract

Recently, interest has surged in glove-type assistive devices for relieving hand muscle stiffness caused by brain lesions. This study aims to develop an ergonomic method for drafting glove patterns intended for hand-assistive devices. To facilitate pattern development, we acquired three-dimensional (3D) scan data from the four hemiplegic patients while their hands were in a relaxed posture, which was subsequently transformed into two-dimensional (2D) data. Based on the 3D shape data, we analyzed the finger joint range of motion (ROM) and change ratio of skin surface length resulting from flexion and extension movements of the paralyzed hand. Incisions were strategically applied to regions displaying significant variations in these parameters. These flattened 2D patterns were then integrated into revised pattern blocks to enhance the shading data related to the 3D shape, resulting in the development of four glove patterns. We found that gloves prototyped using this innovative pattern-drafting method did not impede joint ROM when worn. Changes in clothing pressure inside the glove at the joints corresponded to the bending angles of the fingers, and the pressure did not exceed the discomfort threshold during hand flexion and extension movements. Importantly, participants provided positive subjective feedback concerning the comfort of the gloves. Our findings yield fundamental data for developing a foundational glove design for hand-assisted devices for patients with paralysis, achieved through the utilization of this novel ergonomic glove pattern-drafting method.

Introduction

Stroke is a global health concern, with the World Health Organization (WHO, 2022) reporting that 15 million people worldwide experience a stroke every year. 50% of these individuals endure chronic disabilities, positioning stroke as the second-leading cause of death and the third-leading cause of death and disability combined worldwide (World Stroke Organization, 2022). Upon the onset of disability, stroke survivors often grapple with the inability to move one side or part of their body (The Stroke Foundation, 2022). One prominent manifestation of stroke-related impairment is hemiplegic disorder, characterized by muscle weakness, abnormal muscle activity, skin reflexes, and upper motor neuron syndromes such as spasticity resulting from muscle paralysis. These factors collectively hinder voluntary movement control (Cauraugh et al., 2000). Hemiplegia entails pathologies such as muscle weakness, spasticity, paralysis, and abnormal hand flexion movements (Dombovy et al., 1986). Particularly, the hand, being the most complex anatomical structure in the human body, performs various delicate functions; however, a stroke can weaken its functionality, restricting daily life activities (Nasir et al., 2014).

Recent efforts to address hypertonicity in affected hands have led to a surge in research focused on glove-type wearable devices (Lee et al., 2022a). As an alternative to rigid and bulky hand assistive devices, such glove-type wearable devices, made of flexible fabrics, help minimize damage to the human hand, possibly caused by excessive pressure and friction, especially when the devices are actuating (Ge et al., 2020). Particularly, for the hands of disabled patients with weakened skin and muscles, devices in the form of fully covered gloves could be desirable in providing necessary protection to the wearers from physical and chemical hazards of the driven actuators (Cauraugh et al., 2000; Dombovy et al., 1986). Cappello et al. (2018) used an actuator based on pneumatic technology to assist patients with paralysis in gripping or manipulating objects. They designed a soft fabric-based glove to enhance portability and user comfort. Yap et al. (2017) introduced active support for finger flexion and extension in stroke patients, incorporating an origami-type actuator within a lightweight and soft glove-type device. Wang et al. (2017) and Yoon et al. (2018) also developed glove-type rehabilitation devices for hand flexion and extension exercises using pneumatic actuators for patients with paralysis. Additionally, an elastic elastomer actuator was incorporated into glove designs for rehabilitation exercises (Chen et al., 2021). These initiatives have resulted in the creation of lightweight and flexible glove-type hand-assist devices for patients with hand paralysis, minimizing the use of rigid materials. Although many studies have suggested that hand length, width, circumference, or a combination of hand length and circumference are the most appropriate measurement items when developing glove-type products, such as gloves or hand-assist devices (Özer et al., 2007; Robinette & Annis, 1986), a standardized set of hand-size criteria for hand-related products remains lacking.

An accurate understanding of a wearer’s body size is a fundamental factor in designing ergonomic wearable products. This becomes particularly true when aiming to create gloves tailored for patients with hand paralysis, where precise measurement and characterization of hand size, accounting for spasticity-induced deformations, are paramount. In essence, as the first step in the design and development process, crafting a foundational glove for hand-assistive devices necessitates considering the hand’s precise geometry to optimize wearability. However, gloves or hand-assist devices, when worn, can impose limitations on hand performance. Moreover, the seamlines of these products may interfere with hand function, given the intricate anatomical structure of the hand, which permits more nuanced functionality than other body parts (Dianat et al., 2012; Lee et al., 2013). Thus, setting the optimal position of the actuator attached to the hand or designing it such that the position of the seamline of the material used does not interfere with the function of the hand becomes critical. Additionally, assistive mechanisms must be placed at the precise position of the hand to realize optimal wearability and desirable driving force in the device.

To this end, an increasing number of researchers have embraced three-dimensional (3D) technologies for accurate anthropometry. Notably, 3D scanning technology has been proven instrumental in enhancing the wearability of assistive devices through a comprehensive understanding of the dynamic interplay between wearable devices and the human body (Nasir & Troynikov, 2017). This technology is adept at extracting skin surface characteristics, such as curvature, by minimizing measurement error rates and obtaining detailed differences in human body characteristics, including body size and shape (Choi & Hong, 2015; Yu et al., 2013). Although previous studies proposed sizing systems and patterning methods predicated on the flattening of 3D scanned models, limitations still exist in implementing such approaches directly in hand-assistive devices. That is, the causes and symptoms of upper extremity impairment are inconsistent across patients with hand deformation. Given so, a universal standard for stratifying hemiparesis is hard to set (Stoykov et al., 2009). Thus, a customized pattern is desired for each patient because patients with hemiplegia exhibit varying degrees of hand stiffness and range of motion (ROM), even when presenting similar symptoms. Furthermore, from a practical standpoint in the context of experimental research, due to spasticity and internal rotation contracture of the affected upper limb (Namdari et al., 2011), patients often experience difficulties in maintaining the required poses during 3D body scanning. Resultantly, data loss in specific areas and the introduction of motion support devices during the 3D scanning process are inevitable. Additionally, the hands must move comfortably during daily activities, regardless of whether a wearable product is worn or not. Clothing pressure is a quantitative indicator of the comfort of a wearable product that measures how much pressure is applied at a given body location (Lee et al., 2022b).

Therefore, considering these unique clinical and technical needs for stroke patients, we developed an ergonomic pattern drafting method that could allow the customization of glove patterns to fit the hands of individual stroke patients. Specifically, we measured finger joint ROMs and changes in skin surface length by examining the hands of the recruited four patients with hemiplegia. Then, each joint region on the glove pattern was incised, guided by the hand dimensional data mentioned above, and the 3D hand shapes were flattened into two-dimensional (2D) forms. During the 3D-to-2D flattening, we interpolated the missing meshes in the hand geometry, and completed the glove block patterns. Subsequently, we prototyped a glove for each of the recruited stroke patients, and compared the finger bending angles and the degree of clothing pressure at the back of the hand on the glove. Furthermore, we conducted subjective wearability evaluations on the prototyped gloves with stroke patients.

Methods

Participants

The four hemiplegic patients (70.8 ± 3.4 years; 3 men and 1 woman) participated in this study as the four distinctive use cases. To ensure diversity among the participants, we recruited those who exhibited different ergonomic characteristics and medical histories, as shown in Table 1. Yet, to eliminate the effect of the patient’s hand stiffness level, all recruited participants were Grade 1 on the Modified Ashworth Scale (MAS). MAS is a widely accepted clinical tool for measuring muscle spasticity, and MAS Grade 1 indicates a slight increase in muscle tone with minimal resistance observed at the end of the ROM during flexion or extension of the fingers (Bohannon & Smith, 1987; Meseguer-Henarejos et al., 2017). People with cognitive impairments that hindered their comprehension of the experimental sequence were excluded. The study was approved by the participating university’s Institutional Review Board Committee (IRB approval No. 2112/002-019), and all participants provided written informed consent before participation.

Table 1 Physical characteristics of participants

Ergonomic measurements

As the baseline information, we measured and analyzed joint ROMs and change ratios of skin surface length of each participant. The hemiplegic side of the participant’s hand was scanned in the two hand postures, including relaxed (refer to Fig. 1a and b) and flexed postures (Fig. 1c), using a 3D scanner (Artec Eva, Artec3D, Luxembourg). We focused on the dorsal side of the hand, rather than the palm side, because the skin on the former side tends to stretch, while the palm contracts in most hand movements (Nasir & Troynikov, 2017), and further the actuator is attached to the dorsal side of the hand rather than the palm side to avoid restricting hand movements. To minimize measurement errors, 24 landmarks were marked on the hands before 3D scanning (Fig. 1a). The reference points for landmarks were the endpoints of each finger (L1, L5, L9, L13, L17), the center point at each finger joint (L2–3, L6–8, L10–12, L14–16, L18–20), lateral and inner points of the hand (L4, L21), and inner, center, and lateral points of the wrist (L22–24). Additionally, for measuring the skin surface length at the finger joints, 28 landmarks were additionally attached along a continuous line from each landmark point (L1–L24) based on the creases on both sides of the center point of the finger joints (Fig. 1b). For instance, with reference to the joint wrinkle centered on the landmark L2, we set L2-1 and L2-2, the additional landmarks on each side of L2 with the same length. We then analyzed ROMs and change ratios of skin surface length at each finger joint—i.e., the distal interphalangeal joint (DIP), proximal interphalangeal joint (PIP), and metacarpophalangeal joint (MCP)—in both relaxed and maximum flexed postures using reverse engineering software (Geomagic Design X, 3D Systems, Inc., USA) (Fig. 1c). The change ratio of skin surface length was calculated using the following formula (Wang & Wang, 2015):

Fig. 1

Landmark areas: a for joint ROM, b skin surface length and c measurement method of joint ROM

$$\text{Change ratio }(\text{\%}) = (\text{b}-\text{a})/\text{a }\times 100$$

(a = surface length in relaxed posture; b = surface length in flexed posture).

Glove pattern drafting and prototyping

Utilizing the ergonomic data of the participants in joint ROMs and change ratios of skin surface length, we determined the finger joint regions that would benefit from the application of the stretching material. We then drew splines on the participant’s 3D hand shapes using Geomagic Design X, and flattened them into 2D forms in a 3D modeling program (Rhino 7, Rhinoceros 3D, USA) (Fig. 2). The splining of 3D scanned shapes was performed only on the backs of the patients' hands because the hands of hemiplegic patients tend to roll in for spasticity, making it difficult to acquire the 3D shape of the inside of the palm due to the shadow. The revised glove pattern block proposed by Sokolowski and Griffin (2020) was applied to reinforce the missing data of the flattened 2D shape. Referring to the drafting method of their performance glove’s pattern block, a box was drawn on the 3D shape using the metacarpal phalangeal joint width (relaxed posture) and hand length. A center line was drawn to divide the palm and back of the hand, and the four fingers, excluding the thumb, were divided into four equal parts to complete the pattern block. During 3D scanning of the patients’ hands, the palm area, which is prone to missing parts, was drafted identically to the sideline of the back of the hand, referring to the revised pattern block. In order to provide extra allowance between the fingers, ease amounts were set with a pattern length corresponding to the pattern length of each finger. In particular, seamlines were added to apply elastic materials to joints where the ROM and the change ratio of skin surface length are large. This pattern block can compensate or refine by referring to the lines of the drafted pattern block for missing parts in the derived 3D shape when performing accurate scan postures is difficult for the participant because of the overlap between human bodies during scanning, or rigidity and shivering of the body during scanning. Personalized glove patterns were developed for each participant by placing their respective flattened 2D shapes on the baseline of the pattern block. Finally, four gloves were prototyped employing the patterns developed using the 2D flattened data for each participant.

Fig. 2

Method of flattening a 3D shape into 2D pattern

Wearability test

With the four prototyped gloves, we performed wearability tests to attest the effectiveness of the ergonomic glove pattern drafting method developed in this study. We hypothesized that the pattern drafting method is effective when (a) there is no change in the participant’s finger bending angles when wearing the prototyped glove, as compared to the bare hand; (b) the maximum pressure measured inside the glove is less than discomfort threshold of 3.30 kPa (Zakaria & Gupta, 2019); and subjective comfort perception was satisfactory (above “moderately comfortable”).

Finger joint bending angles

We measured each participant's finger joint bending angles twice: one with the bare hand and the other when wearing the prototyped glove. The participants were asked to perform their hand flexion and extension repetitively for 10 s for measurement of joint bending angles. Specifically, the patient performed the most flexed and straightened hand movements on their own without the help of a researcher, to warrant that no additional assistance was required from others. This sequence of hand movements was done three times. We performed this experiment to confirm the hypothesis that wearing gloves developed using a new pattern drafting method did not affect finger joint bending angles. We video-recorded all hand movements, and extracted motion clips at one-second intervals. The finger joint bending angles were measured at the marked landmarks using a virtual protractor (Screen Protractor, Iconico Inc., USA).

Clothing pressure

Before and after donning the gloves, we measured clothing pressure during flexion and extension of the hemiplegic hand using an air pack sensor (Kikuhime TT Meditrade, ZiboCare, Denmark) to evaluate wearability for hand movements while wearing the developed gloves. The clothing pressure sensor (sensor size = 20 mm in diagram) was attached to the MCP joint of the middle finger (Fig. 3). We selected the metacarpal bone of the middle finger because it was judged to have the most significant influence on the pressure of the glove during bending and extending movements due to its largest size (Labat & Ryan, 2019). While this pressure-sensing device has been widely used as a reliable and accurate instrument (Brophy-Williams et al., 2014), it did not allow continuous data recording. Thus, we manually measured and recorded pressure at the MCP joint of the middle finger during finger bending and extending, once every second. In fact, the manufacturer of the air pack sensor explained that it could generate pressure data up to three times per second; however, considering the patient’s slow hand movement and the sensor’s reaction speed, we collected the pressure data once per second. To ensure data reliability, we repeated this procedure three times and averaged the data. This assessment was to determine if changes in clothing pressure aligned with the variations observed in joint bending angles during hand flexion and extension. Then, the clothing pressure measured inside the glove was compared to the discomfort threshold of 3.30 kPa.

Fig. 3

Measurement region for clothing pressure

Subjective comfort measurement

To gauge the participant’s perceived comfort, we employed the standard comfort assessment scale, called the Comfort Affective Labeled Magnitude (CALM) scale developed by the US Army Natick Soldier Research, Development, and Engineering Center (Cardello et al., 2003). The CALM questionnaire contained 11 items, employing a rating scale ranging from − 100 (discomfort) to + 100 (positive comfort), with 0 indicating a neutral response.

Data analysis

Statistical analysis of anthropometric data and wearability evaluations was conducted using SPSS (Version 26.0, IBM, USA). The Shapiro–Wilk test was performed to verify the normal distribution of the datasets. Following the differences in the joint ROM and the ratio of skin length changes in the finger joints, the Kruskal–Wallis H tests were conducted. The Mann–Whitney U tests were performed to assess the differences in the finger joint bending angles before and after wearing the developed glove. The significance level for all statistical analyses was set at p < 0.050.

Results

Ergonomic hand characteristics

As seen in Table 2, the participants’ ergonomic hand dimensions varied significantly. As for the finger ROMs, for participant 1, the MCP joint exhibited a significantly larger ROM compared to the DIP and PIP joints (p = 0.000). Participants 2 and 3 displayed significantly larger joint ROMs in the PIP and MCP joints than in the DIP joint (p = 0.000), while participant 4 exhibited significantly larger changes in the DIP and MCP joints compared to other joints (p = 0.000). Notably, the MCP joint area showed the largest ROM among all patients.

Table 2 Ergonomic measurement results for participants’ hand

Regarding the change ratio of skin surface length, participant 1’s MCP joint exhibited a significantly larger change than the DIP and PIP joints (p < 0.001). Participants 2 and 3 had larger changes in the PIP and MCP joints than in the DIP joint (p = 0.001), while participant 4 displayed significantly larger changes in the DIP and MCP joints compared to other joints (p = 0.009). Although the change ratio of skin surface length of participant 2’s PIP joint tended to be larger than those of the DIP and MCP joints, no difference was observed between these joints (p = 0.345). Except for participant 3, the joint ROM and change ratio of skin surface length exhibited similar trends. These results underscore the variability in joint ROM and change ratios of skin surface length for each joint among participants with the same spasticity grade. In addition, significant differences in finger joint ROMs were observed among the DIP, PIP, and MCP joints (p = 0.046, 0.001, and 0.000, respectively) for all four participants. In other words, despite having the same spasticity grade, patients with hemiplegia exhibit significant variability in finger joint ROMs. Consequently, they may require a customized pattern-drafting approach tailored to their unique characteristics when designing gloves.

Ergonomic glove pattern drafting

As the first step toward drafting the customized glove pattern for each participant, we developed pattern blocks based on the surface lengths of the ulnar and radial sides of the 3D hand geometry in a relaxed posture (Fig. 4a). We then modified these pattern blocks individually for each participant by referencing the performance glove pattern block proposed by Sokolowski and Griffin (2020). Specifically, the finger length of the pattern block was set according to the lengths of the middle fingers on the 3D shape, and the basic seam line of each finger pattern block was determined by dividing it into quarters. Next, we overlaid the flattened 3D panels on the pattern block and delineated customized patterns (Fig. 4b).

Fig. 4

Glove pattern drafting method: (a) 3D scan data and pattern block, (b) pattern drafting using flattened 3D shape and pattern block, and (c) refining process

Guided by the ergonomic data of the participants, we determined the incision line of the glove pattern to apply elastic materials to joints with large ROMs and change ratios in skin surface length. Specifically, we isolated the MCP joint area from the backhand of the pattern because this area exhibited the largest ROM in all patients. In addition, the angle (16°–31°) between the horizontal line and the line from the ulnar and radial styloid points determined the curvatures of the MCP pattern of the four fingers (Fig. 4c). The flattened 2D patterns of the fingers, back of the hand, and wrist were refined based on the pattern block. In particular, missing data during 3D scanning were reinforced by referring to the pattern block. The arrows in Fig. 4c indicate how to supplement missing shape data and modify the pattern block to match 3D shape data.

Additionally, the glove patterns were customized following the steps. We separated the PIP panels for participants 2 and 3 and the DIP panel for participant 4 from the finger panels because the ROMs of the finger joints were greater than those of DIP and PIP joints, respectively. These panels were set based on reference points when measuring the change ratio in skin surface length (refer to Figs. 1 and 5). Because the separated pattern had a large change ratio of skin surface length, elastic material was applied to allow the glove material to expand or contract accordingly during the movement of the finger joints. The blue shaded parts in Fig. 5 are the areas where the elastic materials are applied. On the other hand, the change ratio of skin surface length between the joints of participant 3 did not differ. Despite no differences between joints, we decided to apply an elastic material because the change ratio of skin surface length between the PIP and MCP joints was not negligible. Additionally, the curvatures in the longitudinal and lateral directions of each finger in contact with the MCP joints were determined based on the gradient of each finger. It can reflect the curved part of the fingers due to muscle stiffness in patients with hemiplegia and the natural abduction and adduction movements of the fingers. Consequently, personalized glove patterns were developed for each participant that reflected hand size and shape as well as hand motion (Fig. 5).

Fig. 5

Final glove patterns for participants

Prototyping of ergonomic gloves

Once the glove patterns were developed using the new drafting method, we prototyped the gloves and evaluated wearability. The choice of material for the gloves was crucial, as it needed to accommodate a considerable range of finger movements, offer comfort and flexibility, and be sufficiently thin to allow full joint ROM without causing excessive bulk. To offer desirable mobility, we placed elastic materials (Nylon 96%, Polyurethane 4%; thickness = 1.13 ± 0.00 mm) in regions with large change ratios of skin surface length and substantial ROM in each participant’s finger joints (joint areas incised in the pattern), as well as in the gaps between the fingers. In addition, we used neoprene material (Polyester 100%; thickness = 2.27 ± 0.01 mm) for the remaining areas to minimize the movement of the actuators attached to the glove for hand assistance or to reduce any potential discomfort caused by pressure or fixation of the actuators. Furthermore, we used Velcro on the sideline of the thumb area for ease of donning and doffing for the participants with hemiplegic hands. Figure 6 provides a visual representation of the participants wearing the gloves created by the developed patterns. Please note that the overall shape and finger panels are different by the participants in the prototyped gloves.

Fig. 6

Prototyped gloves using the ergonomic pattern drafting method

Wearability test

The results of the Mann–Whitney U test for the change in the maximum flexion angle of the finger joint before and after wearing the developed glove were as follows. The bending angles at each joint of the four participants tended to decrease by 0.16–4.70% after wearing the prototype gloves compared to before wearing them. However, these differences in bending angles before and after wearing the gloves were not statistically significant for any of the participants (Table 3). These results indicate that the gloves made using the newly developed pattern-drafting method do not restrict the wearer’s finger joint movement and allow for comfortable flexion and extension.

Table 3 Difference in bending angle of finger joint before and after glove use

In addition, we assessed the clothing pressure of the MCP joints of the middle fingers, when participants wore the developed gloves. In relaxed postures, the clothing pressure ranged from 0.44 ± 0.06 kPa to 0.58 ± 0.06 kPa. Notably, the clothing pressure at the MCP joint of the middle finger changed as the joint ROM changed during the flexion and extension movements (Fig. 7). These results indicate that the developed gloves effectively adapt to finger joint movements, even in areas with a significant change ratio of skin surface length. In addition, when participants flexed and extended their hands for 10 s with the gloves on, the maximum clothing pressure increased, ranging from 0.72 ± 0.11 kPa to 1.08 ± 0.25 kPa. This represented an increase of 47.14–103.25% compared to wearing gloves in a relaxed state. In other words, the maximum clothing pressure inside the glove did not exceed the discomfort threshold of 3.30 kPa during hand flexion and extension. These results indicate the satisfying level of comfort that the developed gloves can offer.

Fig. 7

Change in clothing pressure according to the MCP joint’s bending angle

Furthermore, participants provided comfort scores for the prototyped gloves on the CALM scale, with ratings ranging from 38.00 to 70.00 (Fig. 8). Specifically, P1 and P4 responded “moderately comfortable” with 38.00 points, and P2 and P3 answered “very comfortable” and “extremely comfortable” with 58.00 and 70.00 points, respectively. The average comfort score was 51.00 ± 13.67, which fell between “moderately comfortable” and “very comfortable.” This positive subjective evaluation suggests that wearers found the gloves comfortable during finger flexion and extension movements after wearing them.

Fig. 8

Comfort scores of participants on the CALM scale

Discussion

This study introduced a novel approach to designing ergonomic glove patterns for hand-assistive devices, leveraging anthropometric data. We acquired 3D-shape data of the affected hands of patients with hemiplegia and converted it into a 2D format. We then developed a modified pattern block specifically for glove pattern drafting and superimposed the flattened 2D pattern on it to address any missing data. The resulting gloves, created using this ergonomic pattern drafting method, did not impede the ROM in the finger joints, and their movements closely mimicked the wearer’s joint motions. These findings highlight the potential of this new ergonomic glove pattern drafting method can reinforce missing data for patients with hemiplegia who may struggle to maintain accurate measurement postures during 3D scanning. This method can serve as a foundational design for glove-type hand assistive devices.

We analyzed the joint ROMs and change ratios of skin surface length in patients with hemiplegia to enable the designing of a novel ergonomic glove pattern. Our findings revealed variations in joint ROM and change ratio of skin surface length among participants, despite having the same spasticity grade. Patients with paralysis and brain lesions often experience limited joint ROM due to stiffness and paralysis of the hand, and the severity varies depending on the symptoms (Lindberg et al., 2012; Parker et al., 1986). Additionally, the skin surface length of the dorsal hand can change significantly depending on the hand movement, with pronounced deformations in the joint area (Choi & Ashdown, 2011; Vergara et al., 2018). This information offers valuable insights into determining seamline locations or materials when designing hand-related products. Consequently, we proposed methods that cater to finger joint movements by applying pattern design techniques and elastic materials to joint areas with significant ROM and skin surface length change, customizing the approach for each participant's unique characteristics.

In this study, we introduced a new pattern-drafting technique aimed at addressing the challenge of missing 3D body data by flattening a 3D human body shape into a 2D representation using a 3D design program. This pattern-drafting method offers versatility and can be customized for various applications, particularly catering to individuals with unique anthropometric characteristics, including variations in body curvature and shape. Recent advancements in this field have introduced alternative techniques, such as using multilevel meshes to tackle the issue of large mesh complexities or employing 3D-modeled virtual clothing to generate 2D patterns (Liu et al., 2018; Zhang et al., 2018). However, while providing numerous advantages over traditional measurement methods such as time efficiency and high accuracy, presents a challenge in the form of missing data caused by shading (Istook & Hwang, 2001). This problem is exacerbated in areas of the body where one part casts shadows on another due to body structures (e.g., under the breasts and chin), at joints where shadows can occur because of specific postures (e.g., behind the armpit and knee), and in regions where body parts overlap (e.g., the crotch, between the toes, and between the fingers). Notably, this issue becomes more pronounced in patients with paralysis as their stiff muscles render maintaining precise measurement postures difficult. To mitigate this problem and compensate for data loss due to shading, we recommend a solution that involves adapting the performance glove pattern block proposed by Sokolowski and Griffin (2020) to suit each patient’s unique requirements. Our study presents a novel pattern drafting method that effectively addresses these challenges, offering a promising solution for capturing accurate body data in cases where conventional methods fall short.

Furthermore, we determined the curvature of the MCP pattern of the fingers for the area with the largest change in ROM and skin surface length by reflecting the angle between the horizontal line and the line from the ulnar and radial transverse points on the flattened 2D pattern (i.e., the region where the fingers and dorsal hand meet). The hand naturally forms an arched shape during flexion, curving toward the palm, and the metacarpal joint area plays a crucial role in enabling hand functionality (Labat & Ryan, 2019). In particular, when the hand is rigid and paralyzed, the flexion of the hand is greater, and it maintains this shape even in a relaxed state (Dombovy et al., 1986). The method we proposed for applying curvature to the MCP pattern panel is one of the ergonomic design approaches for creating 3D gloves that are well-suited to the curved hands of paralyzed patients. To elaborate, because the new pattern drafting method proposed in this study considers the changes in finger ROMs and skin surface lengths of the patients’ hands occurred during hand motions into the patterns, the gloves that are created using the pattern drafting method likely offer a satisfactory fit to the wearer. Resultantly, the ergonomically-shaped gloves are effective not only in facilitating the intended function of the actuator by accurately transmitting it to the right location on the wearer’s hand, but also in providing necessary wearability to the wearer, as they move along with the hand curvatures and joint movements.

Although glove-type assistive devices should not interfere with hand movements, some pressure may be necessary to achieve the desired functionality. This ensures that the base glove of the assistive device can effectively follow the joint’s movement (Labat & Ryan, 2019). In particular, when the gloves to which the developed glove pattern design method was applied were worn, the clothing pressure of the participants’ MCP joints ranged from 0.72 kPa to 1.08 kPa at most during joint movements. Increased pressure between clothing and the human body can cause discomfort. We found that the clothing pressure exerted by the gloves developed using this study’s approach did not exceed 3.30 kPa, which falls within the permissible clothing pressure range for medical clothing (Zakaria & Gupta, 2019). Furthermore, hands with many bones and joints are capable of performing delicate work; however, wearing hand-related wearable products restricts finger movements and can worsen dexterity (Dianat et al., 2012). Our results demonstrated that the glove prototyped with the developed pattern-drafting method had no significant effect on participants’ joint ROM. Moreover, these gloves effectively adapted to the movements of the joints, ensuring that the assistive device didn’t impede the wearer's ability to move their hands naturally. In other words, when an actuator attached and inserted into base gloves using the developed pattern, our results indicate that the base gloves did not impede the function of the actuator intended by the developer, but it rather provided satisfactory wearability to the wearer.

However, we acknowledge that this study has several limitations. First, it focused on designing a customized glove pattern for only four patients with hemiplegia. While the findings are promising, further evaluation with a more extensive and diverse sample size is essential to generalize the results effectively. As the patients with hemiplegia who participated in this study shared the same spasticity grade (MAS grade 1 +or lower), further evaluation is required to target patient groups with varying symptoms and spasticity grades. Second, the material of the prototyped glove for evaluating the developed pattern-drafting method can affect the wearer’s joint ROM and overall wearability. To comprehensively assess the versatility and adaptability of the developed pattern-drafting method, conducting evaluations using various materials to identify the most suitable options for different scenarios and wearer preferences is essential. That is, additional evaluation should be conducted by developing gloves using materials with different degrees of elasticity to reflect the elasticity rate in the divided pattern area according to the change in bending angle of each joint and the change ratio in skin surface length of the patients.

Another limitation is related to the experimental and environmental conditions of the evaluations. There may be a slight difference in position between the reference point attached to the glove and the reference point of the finger inside the glove. To increase the reliability of measurement results, experimental methods must be designed in which landmarks can be measured at the exact same location before and after wearing gloves to minimize these differences. Although we conducted the assessments in the controlled environment of a rehabilitation treatment room, the degree of hand spasticity may vary among paralyzed patients on different days, leading to slight variations in joint ROM. To account for these potential fluctuations, conducting evaluations under various conditions and over more extended periods could provide a more comprehensive understanding of the method's effectiveness. Finally, our evaluation was limited to a controlled environment and did not consider the real-world use of the gloves. Future studies should aim to assess the developed gloves in actual use scenarios to further validate their performance and wearability. Additional evaluation efforts are desirable to verify the effect of the pattern drafting method by comparing with existing glove products available on the market. Despite these limitations, our study introduces a novel pattern drafting method for individual customization that addresses a critical gap in the field. This method holds significant promise, particularly for individuals with unique hand shapes and curvatures, such as those with disabilities. Moreover, our research creatively tackles the challenges associated with current 3D body scanning technology, specifically addressing issues related to data missing in various body regions.

Conclusions

In this study, we developed an ergonomic glove pattern-drafting method for individuals with hand deformation based on dynamic measurements such as joint ROM and the change ratio of skin surface length for hand-assistive devices. The glove pattern-drafting method we developed can positively contribute not only to the field of clothing textiles but also to the field of wearable assistive devices. A key achievement of this study is the creation of glove patterns that accurately capture the unique body characteristics of the wearer’s hand. This was realized through a comprehensive analysis of 3D hand shapes, which were subsequently transformed into 2D patterns. These patterns offer the promise of superior wearability and enhanced functionality for hand-assistive devices, ultimately improving the quality of life for individuals with hand paralysis. In addition, we proposed a novel and effective pattern-drafting method that addresses the challenges inherent in 3D technology, particularly in scenarios where data acquisition is hindered by hand stiffness and tremors, leading to inaccuracies in posture capture. In conclusion, the findings and methodologies presented in this study can serve as valuable foundational data for the development of wearable hand-assistive devices tailored to the unique needs of individuals dealing with hand paralysis. This study represents a significant step toward enhancing the accessibility and effectiveness of assistive technologies for those with hand deformations and disabilities.