Content area
Assessing the pliability of burn scars represents a key clinical challenge for monitoring the healing process and optimising rehabilitation treatments. Currently, most of the methods used to assess scar quality are based on subjective evaluations, limiting the accuracy and reproducibility of the results. This study presents the development of an innovative scanner based on the Digital Image Correlation (DIC) technique to objectively measure changes in scarred skin deformation. The designed system integrates a hardware configuration with an image processing algorithm able to detect with high accuracy the deformations of the skin subjected to a known stress, allowing the calculation of the pliability value. By comparing two values obtained in two successive sessions, it is possible to verify the improvement in the health of the scar. Preliminary tests have shown that the system is able to provide reliable and reproducible data, confirming the technology’s potential for scar quality assessment. The implementation of this technology could help change the management of burn scars, improving patients’ quality of life through more accurate diagnoses and more effective treatments.
Introduction
The evaluation of the health status of a burn scar is a critical step in making an accurate diagnosis and identifying the proper therapy for the patient [1]. In modern clinical practice, the health of a scar is evaluated subjectively by assigning scores to several physical aspects of the lesion [2]. The combination of these scores results in a global rating that describes the scar’s overall condition [3, 4].
The Vancouver Scar Scale (VSS) and the Patient and Observer Scar Assessment Scale (POSAS) are the two most commonly utilized subjective assessment scales in clinical practice today. Both measures provide scores to the identical scar characteristics [5]. The main difference is that the POSAS is further divided into two subscales: one reflecting the physician’s opinion and the other addressing the patient’s point of view, including aspects such as discomfort and itchiness [6]. It is evident that this type of evaluation is subjective, non-repeatable, and unreliable [7, 8–9].
In recent years, research has moved towards an objective evaluation of burn scar health by analyzing individual characteristics using repeatable and reliable approaches [10, 11, 12–13].
One of the significant variables studied is pliability, which refers to the skin’s capacity to change shape in response to external forces. Restoring this feature of the lesion to the level of the surrounding healthy skin as quickly as feasible is critical, as even little changes might create motor pain in the patient. If not handled promptly, these disorders may progress to disability [14].
Currently, the evaluation of this parameter is done by “wrinkling” the scar between the doctor’s thumb and index finger and assessing the perceived softness. The structure of the skin causes it to behave like a complex substrate with elastic, viscous, and plastic properties. The elastic component of the skin reflects the stretching of collagen in elastic fibers, while the viscoelastic component is related to the movement of interstitial fluid. Scar tissue is, in fact, stiffer than healthy tissue, and its hardness increases with the severity of the injury [14, 15].
A commercially available technology that meets the requirements for an objective assessment of skin pliability is the cutometer. This instrument uses a vacuum suction system applied perpendicularly to the surface of interest through a probe, whose diameter varies depending on the area to be analyzed. The probe measures the vertical deformation of the skin caused by negative pressure that draws the tissue in [16]. The pliability of the skin is linearly correlated with its deformation in response to this type of stress [17].
The cutometer can operate in two modes: time-deformation or stress-deformation. Generally, the former technique is preferred, in which a controlled and constant force is cyclically applied for a set period, during which deformation is measured. Subsequently, a controlled relaxation phase follows in the ensuing seconds. The behavior of the skin under stress is recorded and graphically represented by a deformation curve [18].
[See PDF for image]
Fig. 1
Curve of Skin Deformation Under the Action of a Cutometer
As shown in Fig. 1. there is a very fast initial deformation during the suction phase, which represents fully elastic behaviour, followed by a viscoelastic deformation and, finally, by a viscous behaviour. During the retraction phase, the curve shows the same elements one after the other in the opposite order. When repetitive measuring cycles are applied, the curve has the tendency to move higher due to residual deformations [19, 20].
The cutometer evaluates the elasticity of the skin based on the analysis of absolute and relative parameters (Table 1). Absolute parameters include measurements of deformation and retraction, while relative parameters include the ratios between the absolute values [19].
Table 1. Absolute and relative parameters evaluates by cutometer
Absolute Parameters | Relative Parameters |
|---|---|
Ue = immediate deformation | Ua / Uf = gross elasticity, including the viscous part of skin deformation |
Uv = delayed deformation | Ur / Uf = biological elasticity |
Ut = maximum deformation | Ur / Ue = net elasticity, excluding the viscous part of skin deformation |
Ur = immediate retraction | Uv / Ue = viscoelastic/elastic ratio |
Ua = final retraction | H = hysteresis, the difference in maximum skin deformation between the last and first suction |
R = residual deformation after the first suction | |
R8 = viscous part |
Despite the cutometer being a widely used instrument in the aesthetic field, its use in the medical field is still limited. Furthermore, the tissue surface area it can analyses is relatively small due to the precision of the probe used for analysis, further limiting its clinical applications.
Within the T3Ddy Laboratory, a research facility established through a collaboration between the Department of Industrial Engineering at the University of Florence (DIEF) and the Meyer Children’s Hospital IRCCS, a project has been launched to explore new approaches for the objective evaluation of pliability using optical methods based on digital image correlation [16, 21, 22, 23–24].
In this paper, the study of the objective assessment of the pliability of a burn scar through non-suction methods will be introduced and deepened.
Referring to what has already been mentioned above, an approach of this kind arises from the need to separate the objective assessment of the characteristics of the burn scar from the application during therapies.
To achieve this goal, a scanner prototype was developed, consisting of an image acquisition system and a system to apply a known force in a constant and repeatable manner. The comparison of two scores from two consecutive sessions is a good indicator of scar health improvement.
In the following pages, the functioning of digital image correlation (DIC) will be illustrated, followed by the description of the constructive solution, along with its subsequent implementation, of the force application system. Finally, the tests performed to validate both the software, and the hardware components will be described.
Digital image correlation
Digital Image Correlation (DIC) is an advanced, non-contact optical measurement technique widely used to assess displacement and deformation on surfaces of materials and structures. This method leverages digital images taken before and after the application of a load to track surface movements, enabling detailed full-field deformation mapping [25]. DIC is particularly appreciated for its ability to detect both minor and significant deformations on materials of various kinds, including soft and complex ones [26, 27]. The technique is based on correlation algorithms that compare changes in the texture patterns applied to the object’s surface, offering high precision without introducing physical interference. Thanks to its versatility and precision, DIC has become a fundamental tool in the mechanical characterization of materials, the validation of numerical simulations, and many other research and development applications [28, 29].
DIC is based on sets of images of the surface of the specimen in the undeformed (reference) and deformed states. DIC can be implemented both in a bi-dimensional (2D-DIC, with a single camera) and a tri-dimensional (3D-DIC, using two or more cameras) version. A calibration is necessary to initialize the spatial correlation processes of DIC. The images are divided into smaller sub-images (facets), and a matching algorithm is used to match the facets between the reference and deformed states. The displacement field is then completed.
By transferring this deformation evaluation method from the mechanical field to the clinical field, it is possible to develop a tool for the objective assessment of skin pliability (Fig. 2). Specifically, by comparing two images of the same scar, one in the resting configuration and the second taken while the skin is deformed with the application of a controlled force, it is possible to quantify the deformation and thus the pliability.
[See PDF for image]
Fig. 2
Application of Digital Image Correlation to the study of the pliability of burn scars
In the case of 2D-DIC, images of specimen surface in the undeformed (or reference) and deformed states are acquired by one high-spatial-resolution digital image acquisition device (such as a regular digital camera, a high-speed camera, an optical microscope). The digital images (typically in grey scale) are divided into sub-images (facets). In order to obtain an approximation of grey scale between pixels instead of discrete and independent values, the grey-scale distributions are interpolated, usually with a bicubic spline. Images of the deformed states are compared to the reference one in order to match facets and track the displacement. The degree of matching between facets is evaluated by a normalized cross-correlation function such as (1).
1
or a normalized sum-of-squared-differences such as (2):
2
where F(x, y) and G(, ) represent the grey-scale value for the pixel at the coordinate (x, y) of the reference image and the coordinate (, ) of the deformed image, respectively. N and M are the dimensions of the facet, usually square. After matching the facets, the full-field displacement is automatically computed by tracking the change in position of points on digitized images. In fact, the coordinates in the reference image (x, y) and in the deformed one (, ) describe the deformation between the two states (3):
3
where u and v represent the displacements for the facet centres in the x and y directions, respectively. Δx and Δy are the distances in the x and y directions, from the centres of the facet to the point in coordinates (x, y). The gradient terms in (2) indicate that the initial facet of (M × N) pixels will be strained to optimally match the correspondent facet in the deformed status [30, 31]. The strain tensor (4) is obtained by derivation on displacement gradients [26]:
4
In order to find the six deformation parameters (u, v, , , , ) and match the facet, an approximate-solution method is adopted. Usually, the Newton–Raphson algorithm is used because of its computational economy. Other algorithms are also adopted, such as the Levenberg-Marquardt. When the method converges, the displacement field is obtained but discontinuities might appear due to the local grey-scale value [30].
A smoothing algorithm is needed to provide a continuous displacement field and perform a strain analysis. Among the available smoothing algorithms, some are better suited than others, depending on the features of the noise to be attenuated [32]. 3D-DIC can be considered as an extension of 2D-DIC, as the operating principles are similar, but extended on a third dimension by using two or more cameras in stereoscopic vision [33].
For an optimal use of DIC, the surface of interest must have a random pattern, which deforms together with the specimen surface. If the specimen presents a natural random pattern, due to an intrinsic texture or inhomogeneity, this can be directly exploited by the DIC system [34]. In all other cases, a random pattern must be generated. To ensure accuracy and precision of the computed displacements and strains, the speckle pattern should meet some requirements [35, 36]:
Random distribution, in order to make each area of the surface of the specimen univocally identifiable;
High contrast, to allow the image correlation algorithm works effectively;
Black/white ratio of 50:50, to avoid regions that cannot be properly recognized;
Roughness should be kept at minimum, in order to avoid alteration of the surface geometry.
Probably the most important issue in biomechanical applications is the size of the speckle dots (in relation to the specimen size), in order to optimally exploit the resolution of the camera [36]. In fact, the larger the measurement window, the larger the corresponding area covered by each pixel (for a given sensor resolution) and therefore the dots of the speckle pattern. In order to obtain the best speckle pattern for the specific application, the dimension of the speckle should be different for each application. The ideal size of the speckle dots corresponds to 3–5 pixels [33]. The magnification factor, M, is defined as the ratio between the number of pixels on the long side of the camera sensor and the long side of the measurement window (M indicates how many pixels correspond to the unit length of the physical specimen). Thus, the ideal size of the speckle dots corresponds to 3–5 pixels divided by M. For example, using a camera-sensor of 5 Megapixels (2448 × 2050 pixels) on a field of view of 2 mm × 2 mm (e.g. few trabeculae), yields an optimal dimension of the speckle pattern of about 0.003 mm. The same camera-sensor applied to a larger area of interest of 2 m × 2 m (e.g. a whole human body) would require larger speckle dots, about 3.25 mm.
The black-on-white speckle pattern is most frequently used: first a uniform white background is created, on which black speckles are added. This preparation provides the optimal contrast. If the surface of the specimen itself is already of a light colour (i.e. bone), preparation of the white background could be avoided. The use of water-based paints minimizes the alteration on biological specimens [37].
In order to obtain the best results from this versatile measurement technique, two parameters must be adapted to the specific application (Fig. 3):
Subset size (dimension of the sub-image used in the computation).
Subset spacing (step between consecutive facets).
[See PDF for image]
Fig. 3
Graphical definition of Subset Spacing and Subset Radius
The values assigned to these parameters determine the accuracy, precision and spatial resolution [38]. There is no universally optimal set of parameters, due to the numerous possible uses of DIC, particularly in biomechanics. A choice must be made in relation to the specific application.
The first one parameter is the Subset size or radius. The digital images are divided into sub-images, called facets, of M × N pixels (typically squared). Each subset is represented by a grey-level distribution, which is, in most cases, interpolated by a bi-cubic spline to obtain an approximation of grey-scale between adjacent pixels. Each facet is summarized by the information about the pattern, and its location in space. The correlation algorithm identifies the best-matching region at different load steps. The facet size must be defined according to the specimen size (or the field of view), the size of the speckles, and the strain gradients expected based on the loading conditions and the anatomy [33, 37]. The facet should be larger than speckle dots, to allow detection of small displacements, in relationship to the granularity of the speckle pattern [34].
However, the facet should not be unnecessarily large, to avoid loss of resolution [39].
The spacing parameter indicates the distance between two consecutive facets. It describes the density of facets in the measurement window: the smaller the grid spacing, the larger the number of facets (at a higher computational cost). The influence of the grid spacing on the precision and accuracy of the computed displacement field is minimal [39]. Conversely, the overlap provides advantages in terms of precision and accuracy of the computed strain field. The density of measurement points should be selected based on the test details (type of specimen, field of view, pattern and strain gradient). For an expected uniform strain (e.g. long bone in bending) larger grid spacing can be preferable. Conversely, if high strain gradients are expected (e.g. specimens with complex geometry), a smaller grid spacing is necessary.
For the DIC analysis in this work, an open-source MATLAB® toolbox based on Ncorr algorithms was used [40]. The code has been implemented and adapted to the application of our project.
By comparing the values obtained in two successive sessions, it is possible to verify whether the treatment undertaken to restore the pliability of the scar is effective.
System design and implementation
To achieve this type of evaluation, it is necessary to create an image acquisition system, e.g. using industrial cameras, a system to apply a repeatable constant force to the scarred area, and, finally, to process the data through DIC analysis.
Images have to be acquired, for the analysis, in two distinct configurations: one in undeformed conditions (i.e. at rest), and the other in deformed conditions (i.e. the subject of the analysis is deformed by a controlled external force). To this end, a controlled force application system was developed, consisting of an electric actuator, a pressure sensor, and an Arduino microcontroller. The movement of the actuator controls the advancement of a plunger that creates a deformation on the scar; to ensure the repeatability of the measurement, the pressure sensor transmits information on the applied force, allowing the microcontroller to adjust the position of the plunger accordingly. Figure 4 shows a diagram of the setting just described.
[See PDF for image]
Fig. 4
Setting diagram of the image acquisition system
The following paragraphs will detail the design and implementation of the components of the evaluation system concept for burn scar pliability (Fig. 5).
The Photo Acquisition System consists of two cameras that simultaneously capture images and the Force Application System (FAS) (Table 2). The selected cameras are the Ace acA4112-20 μm USB 3.0 Monochrome from Basler [41] and the Ui 3200 se from IDS Imaging Development Systems [42].
[See PDF for image]
Fig. 5
Camera technical drawing
Camera and optics data are shown in the following tables:
Table 2. Camera and optic data
IDS UI3200-SE | ACA4112-20 μm | |
|---|---|---|
Interface | USB 3 | USB 3.0, nominal max. 5 Gbit/s (SuperSpeed) |
Sensor type | CMOS | CMOS |
Frame rate (resolution max) | 24.0 fps | 23.4 fps |
Resolution (h x v) | 4104 × 3006 | 3840 × 2160 |
Shutter | Global Shutter | Global shutter |
Optical class | 1.1” | 1.1” |
Resolution | 12.34 MPix | 12.3 MP |
Pixel size | 3.45 μm | 3.45 μm |
Sensor model | IMX304 | IMX304LLR |
Lens | CHIOPT– A1616A | C-mount |
Both cameras provide extremely high image resolution (3840 × 2160 pixel) and, thanks to the dedicated software provided by the manufacturers, all image parameters, including format, gain, and gamma, can be adjusted throw dedicated software.
These components have been integrated into a custom-designed single device tailored to the system’s requirements (Fig. 6).
[See PDF for image]
Fig. 6
Illustration of the CAD model of the photo acquisition system
The prototype was designed with a focus on evaluating scars specifically on the upper limbs. This decision was made because a significant number of burn scars, particularly in pediatric cases, are found in these areas [43].
The design of the Force Application System (FAS) comprises the Actuonix Motion Devices L12-30-100-6-I [44], a resistive force sensor (FSR) and an Arduino microcontroller. This system is connected to the main structure at the back and through a central support structure. The cameras are positioned on the rear and fixed to the main structure by means of an adjustable block for the alignment of the optical axis. An arm support has been integrated, equipped with a handle that can slide along a circumference created by a perimeter guide. In addition, a system configuration was designed for the test phase, replacing the arm support with a silicone sample holder (Fig. 7).
[See PDF for image]
Fig. 7
CAD model of the system in the image capture configuration for testing
Force application system (FAS)
The design of the entire system began with the development of the force application system. (FAS). First of all, an electric actuator was selected that met the project requirements; that is, reduced dimensions and sufficient travel to be able to satisfy (Fig. 8).
The choice was the L12-30-100-6-I model from the company Actuonix, ideal for the application in question given its small size and a stroke of 30 mm. The specifications of the device are summarized in Table 3.
Table 3. Technical specification for L12-30-100-6-I
Gearing Option | 100:1 |
|---|---|
Peack power point | 31 N − 7 mm/s |
Peak efficiency point | 17 N − 10 mm/s |
Max speed | 13 mm/s |
Max force | 42 N |
Back drive force | 22 N |
Stroke option | 30 mm |
Mass | 34 g |
Repeatability | +- 0.3 mm |
Max side load | 30 N |
Close length | 102 mm |
[See PDF for image]
Fig. 8
L12-30-100-6-I technical drawing
To receive information on the force value applied by the system on the scar, a resistive force sensor was integrated and connected along with the actuator to an Arduino interface.
The arrangement of these elements is depicted in the assembly illustration and the exploded view in Fig. 9.
[See PDF for image]
Fig. 9
Exploded view of the Force Application System
The force sensor is positioned behind the tip in a special housing connected to the electric actuator.
The resistive zone of the sensor is kept in contact with the back of the device tip through two return springs positioned on the sides of the FAS. These springs were made from PLA and their elastic properties depend on the geometry of the coils and the number of them. They were designed so that the retraction forces were aligned with the center-to-center distance of the holes, thus minimizing the snagging effect.
The tip consists of two zones: the tip, specifically a hyperbolic cone with a tapered shape and rounded point, useful for not causing further discomfort or damage to the scar, reinforced by a thin rib to prevent deformations during measurements; the second zone, which fits perfectly inside the shell, is dotted with five trapezoidal notches. The serrations increase the contact surface between the parts, better distributing the applied forces and reducing the risk of unwanted movements, allowing for greater mechanical grip, improving structural rigidity, and resistance to deformation. The same design precautions were also implemented in the sensor housing to increase the system’s stability. With the advancement of the actuator, the tip will come into contact with the scar, causing the force value recorded by the sensor to increase. This value is read in real-time by an Arduino microcontroller that adjusts the run to stabilize at a target value set by the user.
A control system was developed based on a PID (Proportional-Integral-Derivative) algorithm implemented on an Arduino platform (Fig. 10). The system uses a force sensor to monitor the applied force, while the linear actuator is controlled by a PWM signal. The objective is to maintain a target force, which can be adjusted according to the experimental requirements. After careful evaluation, the PID control constants are set to 0.5 for the proportional term (Kp), 0.2 for the integral term (Ki) and 0.1 for the derivative term (Kd). These values can be further adjusted to optimize the system response.
[See PDF for image]
Fig. 10
PID control to ensure convergence to the target pressure value
The Arduino software starts by first connecting the actuator and setting it to the home position. Serial communication is enabled to allow real-time interaction with the user. During operation, if control is enabled, the system reads the measured analogue value from the force sensor and maps it to an appropriate scale. The error, defined as the difference between the target force and the actual force, is used to calculate the proportional, integral and derivative components of the PID controller. The integral of the error is limited to avoid overloading and the overall PID output is used to adjust the actuator position. An ‘s’ command deactivates the control and retracts the actuator, while a ‘g’ command activates it. A delay of 600 milliseconds was introduced between loop iterations to stabilize the system and allow more accurate adjustments. This PID control system ensures that the force exerted by the actuator remains close to the desired target value, guaranteeing precision and stability in the system’s behaviour. Timing control is manual and managed by the operator.
Once the pressure value, displayed by the Arduino script and recorded by the force sensor, stabilizes (within a maximum of 20 s), a Python™ script can be executed to simultaneously control the dedicated software for both cameras and save the images into designated folders for the subsequent analysis phase.
Although acquiring perfectly simultaneous images can be technically challenging due to potential synchronization issues between separate camera systems, in this setup the Python™ script has been optimized to minimize delays. The time difference between the two image acquisitions is consistently below one hundredth of a second, which is negligible for the purposes of deformation analysis.
Photo acquisition system (PAS)
After the implementation of the force application system, the image acquisition system was designed to support cameras and FAS (force application system), ensuring stability and the ability to adjust the acquisition configuration based on the patient’s physiognomy (Fig. 11).
[See PDF for image]
Fig. 11
Load-bearing structure of the PAS: to the right, an area dedicated to housing the chamber adjustment mechanism and the FAS. On the opposite side, track for connection to the limb support or for the positioning of the samples
The system consists of two specular L-shaped elements joined at the rear by a rectangular element and at the front by an arch connection. All elements have been lightened with openings. They are connected by means of screws and M3 threaded inserts.
In the upper part the L-shaped elements are connected by a third element, called the ‘camera plate’, to which the optical axis adjustment system of the cameras will be connected. The plate allows for a more forward or retracted positioning of the camera adjustment system and the FAS system. The other part of the structure establishes a connection to the silicone sample holder, used for testing, or the base to support the paediatric patient’s limb (Fig. 12).
[See PDF for image]
Fig. 12
Silicone sample holder
The two sections of the system are connected by a dovetail joint, while a sliding rail allows for the adjustment of the specimen support, which consists of a hollow part that moves along the base. The upper part of the support has the shape of a rectangular truncated prism, equipped with symmetric structures to facilitate sliding. The silicone sample is placed in a rectangular frame, which allows the silicone to deform under pressure, simulating the behavior of human skin. The housing is designed for samples with specific dimensions (5.5 × 9.5 × 0.4 cm), and at the bottom, it features through holes to allow vertical adjustment.
To facilitate the replacement of the sample, the support is equipped with a frame cover with a snap-fit mechanism, which allows for quick silicone replacement. The sensor tip support ensures stability and proper positioning; this structure integrates with the outer shell of the sensor and is equipped with holes for cable passage. The lower part features two rows of arches that widen at the base, allowing insertion into the main structure. There are four adjustment slots to allow movement of the base and the tip support: one for each L support and two for the rail. Furthermore, reinforcements have been added to ensure greater stability.
The tip of the sensor is fixed to the support via a front cap secured with screws, while a U-shaped clamp at the rear prevents unwanted lifting of the structure during the application of forces (Fig. 13).
In the image below, the optical axis orientation system of the cameras is shown.
[See PDF for image]
Fig. 13
Optical axis orientation system
The system consists of two identical overlapping parts, starting from a square base with holes for anchoring to the main structure and for connecting with other components via screws. On the sides of the base, there are recesses that allow the insertion of side elements with an arc track that enables the tilting of the cameras. The movement of the cameras along the arc is controlled by a worm screw that engages with a helical gear, driven by a knob (Fig. 14).
[See PDF for image]
Fig. 14
Detail of the cameras adjustment system
One last element is responsible for connecting the cameras to the orientation control system. This connection is made through M3 bolts positioned at the center of the square base. The first camera is fixed with the optical axis coinciding with the plane of symmetry of the system, while the second camera is positioned in such a way that its optical axis is coplanar with the first camera with the possibility of adjusting the angle of incidence (Fig. 15 and 16).
[See PDF for image]
Fig. 15
Detail of the cameras support system: the optical axes are coplanar; the right camera is fixed while the left has one degree of freedom
[See PDF for image]
Fig. 16
Detail of the locking system of the angle adjustment system
To ensure accuracy and adaptability to acquisition requirements, all adjustments can be monitored by measuring marks engraved directly on the device.
Software definition
Several open-source software versions are available for Digital Image Correlation (DIC) analysis. Among the options, DuoDIC developed by the Solav Biomechanical Interfaces Group was selected in this work. Since DuoDIC is developed in MATLAB, it was possible to modify the code to align it with our specific requirements [45].
DuoDIC is an open-source toolbox available for MATLAB that performs stereo Digital Image Correlation with two cameras. DuoDIC receives two sets of synchronized images captured by the two cameras: the first set consists of images of a flat checkerboard target used for calibrating the stereo camera pair; the second set includes images of an object with an applied stochastic pattern, which may experience movement and deformation. The toolbox processes the image series and integrates various camera calibration algorithms with the DIC software based on the 2D subset Ncorr [46]. This integration transforms the corresponding image points into 3D points, producing a dynamic point cloud, meshed surfaces, rigid body motion, and full-field measurements of displacement, deformation, and strain.
During calibration, functions from the MathWorks Computer Vision Toolbox are used to calculate the intrinsic and extrinsic parameters of both cameras. The script requires a series of simultaneous images of a checkerboard target captured by a pair of stereo cameras as input. The points on the checkerboard pattern are automatically detected in each image and used to calibrate the camera parameters (Fig. 17).
[See PDF for image]
Fig. 17
Extract from the camera calibration phase with the Computer Vision Toolbox
The intrinsic parameters include the focal lengths, the principal point (optical center), and up to six distortion coefficients. The extrinsic parameters define the 3D position and orientation of the second camera relative to the first, with a total of six degrees of freedom. Additionally, the calculated parameters are used to determine projection errors, which represent the distance between the detected and reprojected pattern points, providing a measure of the accuracy of the estimated camera parameters.
3D Printing phase
The system was entirely made using 3D printing techniques, specifically with fused deposition modeling (FDM) technology, using PLA as the material, characterized by its low cost and mechanical properties suitable for our purpose. All components were made with a 100% fill to increase the system’s stiffness and avoid deformations during the analysis.
The printers used for the production of the components are the Prusa MK3s + printer and the Prusa Mini printer realized by Prusa3D. The following image shows the arrangement of the parts on the print bed within the dedicated software (Fig. 18).
[See PDF for image]
Fig. 18
Print bed in the PrusaSlicer™ software
The structural pieces were built with a 100% fill to ensure optimal material performance and prevent inadvertent fracture when using the gadget. The printed and constructed device may be seen in the image below (Fig. 19).
[See PDF for image]
Fig. 19
Assembled photo acquisition system
Test and results
Three types of tests were conducted to validate the system:
Software tests to optimize analysis parameters.
Repeatability tests to ensure that the machinery can provide consistent results under identical conditions.
Reliability tests aimed at evaluating the system’s response in the presence of materials that simulate the different stages of burn scars. This test is qualitative.
To test the system, silicone samples with different hardness characteristics were created. Specifically, platinum silicones Ecoflex and Dragon Skin were used, which are widely employed in applications requiring a material with characteristics like human skin, thanks to their softness, flexibility, and stability.
Ecoflex is a family of flexible and soft silicones, developed primarily to simulate soft tissues. It is commonly used in applications where a high stretch capacity and a skin-like elastic response are required. The hardness of these silicones is measured on the Shore 00 scale, which evaluates the resistance to penetration of a material. The numbers associated with the different types of Ecoflex indicate the degree of hardness, with lower values representing greater softness. Three different Ecoflex silicones were considered.
zEcoflex 10 has a Shore 00 hardness of 10, it is extremely soft and flexible, capable of high elongations, which makes it suitable for applications where it is necessary to simulate particularly soft fabrics.
Ecoflex 30: it has a Shore 00 hardness of 30, so it is slightly stiffer compared to the previous one, maintains a good stretching capacity, and offers greater resistance to deformation.
Ecoflex 50: it has a Shore 00 hardness of 50, it is the most rigid in the Ecoflex range.
Dragon Skin is another type of silicone, developed for applications that require not only elasticity but also greater resistance and hardness. In this case, the hardness is measured on the Shore A scale, which indicates greater rigidity.
Dragon Skin 10: it has a Shore A hardness of 10.
Dragon Skin 30: It has a Shore A hardness of 30, compared to the previous ones, it offers superior rigidity while maintaining good elasticity.
The characteristics of the various samples are distinguishable through the Shore hardness scales. Through the latter, it is possible to classify and order the five types of silicone based on their peculiarities, allowing them to be compared and approximated to different skin conditions (Fig. 20).
[See PDF for image]
Fig. 20
Graph of the Shore A and Shore 00 hardness scales
As can be seen from the graph, despite the Shore 00 hardness showing a non-linear trend, unlike the Shore A hardness, the silicones considered for the production of the samples and the subsequent analysis have increasing rigidity.
Silicone samples measuring 5.5 × 9.5 × 0.4 cm (the same dimensions as the support) were then created, pigmented white to facilitate image acquisition. On each sample, a random stochastic pattern was created, composed of small dots distributed irregularly. As previously mentioned, this pattern is fundamental for image processing by the software, as it has allowed for precise monitoring of deformation, analyzing the movements and distortions of the spots during stress, thus providing accurate data on overall deformation. The pattern was applied with the aid of a spray can, achieving a point size in line with the requirements for performing a correct DIC analysis.
Software parameters identification
The first test aims to identify the appropriate subset spacing and subset radius parameters for our application. In fact, as already mentioned, both parameters refer to the geometry of the reference subset used to correlate the images of the deformed configuration to the undeformed configuration. As the parameters increase, there is an increase in the computation times required for correlation, and furthermore, as explained in the literature, it also depends on the size of the pattern used.
To perform this test, a silicone sample made with Dragon Skin 10 was considered, to which the pattern was applied using a spray can. A single cycle of force application was performed by setting a target value to achieve a clearly visible deformation in the specimen. The images were therefore taken in the deformed configuration and in the undeformed configuration. These images were used as input for a series of analyses by varying the values of the Subset Radius (Sr) and Subset Spacing parameters. (Ss) (Table 4).
The cameras settings have been set as reported in the table.
Table 4. Camera settings
Parameters | Value |
|---|---|
Relsolution | 3840 × 2160 px |
Expousure Time | 0.107 ms (min 0.027– max 7580) |
Gain | 1.60x |
Black Level | 0 |
Gamma Factor | Off |
The region of interest (ROI) was kept the same in all analyses, allowing for the comparison and analysis of the trend in the results. The output parameters from the DIC analysis are the value and position of the maximum strain, the average strain, and the strain distribution. Each analysis is accompanied by a graphical representation of the deformations, as can be seen in the following image (Fig. 21).
[See PDF for image]
Fig. 21
Graphical representation of displacement distribution. The point of maximum deformation is marked with a label. There is a gap in the ROI due to the gripping of the system tip. The values are in millimeters
In this case, the deformation is uniformly distributed over the analyzed area because the silicone sample has elastic characteristics that are the same throughout the entire zone. In the case of a scar, we could observe if there are areas with abnormal behavior due to the pliability value being different from the surrounding areas. In the ROI, there is a blind spot due to the presence of the tip pressing on the skin.
The values of the DIC parameters range between 20 and 60 pixels in 10-pixel intervals for the subset radius and between 6 and 10 pixels in 2-pixel intervals for the subset spacing, obtaining the following results (Table 5).
The numerical values can be summarized in the following table.
Table 5. Numerical results of Subset Spacing and Subset Radius analysis
Spacing Radius (Sr) | MAX Dev (mm) | Mean (mm) | Median (mm) | Stand Dev (mm) |
|---|---|---|---|---|
Subset Spacing = 10 | ||||
20 | 1.10 | 0.20 | 0.17 | 0.10 |
30 | 0.70 | 0.19 | 0.17 | 0.08 |
40 | 0.60 | 0.19 | 0.17 | 0.08 |
50 | 0.50 | 0.19 | 0.17 | 0.08 |
60 | 0.48 | 0.19 | 0.17 | 0.08 |
Subset Spacing = 8 | ||||
20 | 1.47 | 0.20 | 0.18 | 0.10 |
30 | 1.35 | 0.19 | 0.17 | 0.09 |
40 | 0.71 | 0.19 | 0.17 | 0.08 |
50 | 0.50 | 0.19 | 0.17 | 0.08 |
60 | 0.48 | 0.19 | 0.17 | 0.08 |
Subset Spacing = 6 | ||||
20 | 1.44 | 0.20 | 0.18 | 0.10 |
30 | 1.11 | 0.19 | 0.17 | 0.09 |
40 | 0.56 | 0.19 | 0.17 | 0.08 |
50 | 0.54 | 0.19 | 0.17 | 0.08 |
60 | 0.51 | 0.19 | 0.17 | 0.08 |
Below is the graph with all the deformation distributions and the graph showing the trend of the values reported in the table (Fig. 22).
[See PDF for image]
Fig. 22
Graphical representation of the maximum deviation values, mean value, median, and standard deviation of the displacement distributions with the parameters Sr and Ss combinations
It is possible to observe that as the value of the subset radius increases, there is a decrease in the average strain value, converging to a stable value. The maximum strain value also decreases, leading to a variation in the spatial position. In fact, for low subset radius values, there are more points with low correlation, meaning the program cannot perfectly reconstruct the 3D coordinates of the undeformed and deformed configurations. Consequently, the program assigns the maximum strain value to a point that does not correspond to the real case. Increasing the value of the subset radius, the correlation increases exponentially. The values of the median and standard deviation converge towards stable values.
On the other hand, increasing the number of points taken from each image increases the computation times for performing the analysis.
From these considerations, it is possible to define an optimal value for the parameters equal to 6 pixels for the Subset Spacing and 50 pixels for the Subset Radius, thus achieving the right balance between computation time and measurement accuracy.
Repeability test
The repeatability test of the measurement was conducted by performing multiple analyses on the same sample with the same configuration. Between one analysis and the next, the necessary time was allowed for the pad deformation to return to its initial configuration (thus eliminating issues of residual deformation). In this case, the average deformation and the distribution of deformations were analyzed (Fig. 23).
Ten acquisitions were made, achieving the following results:
[See PDF for image]
Fig. 23
Displacement distribution boxplot graph of 10 analysis performed for testing system repeability
Table 6. Values of mean and median obtained from 10 tests on the same sample
Case | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 |
|---|---|---|---|---|---|---|---|---|---|---|
Mean distance [mm] | 1.30 | 1.33 | 1.44 | 1.40 | 1.36 | 1.30 | 1.47 | 1.43 | 1.47 | 1.44 |
Median [mm] | 1.10 | 1.21 | 1.27 | 1.25 | 1.22 | 1.18 | 1.27 | 1.24 | 1.26 | 1.24 |
Observing the results and the graph, it is possible to make several considerations about the comparison between the various groups and the repeatability of the results.
First of all, the medians, represented by the line inside each box and reported in Table 6, are quite similar among the different groups, remaining around a value of approximately 1.5. This indicates that the central values of the distribution are consistent, which suggests a certain uniformity in the results obtained for each group. The internal variability, indicated by the interquartile range (IQR), also appears to be constant. The sizes of the boxes, which represent the range between the 25th and 75th percentiles, are indeed similar across all groups, highlighting a rather homogeneous data distribution.
The whiskers, which extend from the quartiles to the extreme values, do not show large variations in length between the groups, which implies that there are no significant differences in overall variability.
Considering these values, it is possible to extract an average value of the mean distance values equal to 1.39 mm with a standard deviation of 0.07 mm, which confirms an excellent repeatability of the system.
Qualitative evaluation of system reliability
To ensure the accuracy of the DIC software evaluation, reliability tests were conducted on four different silicone samples, each with a distinct elastic property [164].
The aim of this test is to qualitatively investigate the effectiveness of the system to distinguish different hardnesses and thus distinguish a change in the skin pliability property. To achieve this, five different silicone specimens with different hardnesses were considered.
The silicone samples used in the tests, listed in ascending order of Shore hardness, are as follows:
Ecoflex 00–10 / 00–30 / 00–50.
Dragon Skin 10 / 30.
To compare the results from different samples, the region of interest (ROI) was kept consistent across all tests, as was the target pressure value applied by the PAS system. The values discussed in the subsequent sections for each silicone type are based on the average of five tests for silicone specimen (Table 7).
Table 7. Deformation graphs of the silicone test specimen. The first line shows the graphs of the Ecoflex silicone test specimens, the second row the graphs of the Dragon Skin silicone test specimens
[See PDF for image]
Analyzing the graph for each test, it is evident that the point of maximum deformation corresponds to the point where force is applied using the FAS system, and the distribution of deformations is uniform cause of the homogeneous elastic propriety of the specimens. It is possible to plot these results as a function of hardness to obtain the following curve, which demonstrates how the behaviour of maximum deformation aligns with predictions (Fig. 23).
[See PDF for image]
Fig. 24
Max deformation in function of hardness
It can be observed that for Ecoflex silicones, maximum deformation decreases as Shore hardness increases, and the same trend is observed for Dragon Skin silicones. It is noteworthy that the maximum deformation value obtained for Dragon Skin 10 is slightly below the value obtained for Ecoflex 50 (Table 8).
The maximum deformation values, average values, and median values are reported in the following table and graph:
Table 8. Maximum deformation values, average values, and median values for each type of silicone specimen
Silicon Tipe | Median (mm) | Mean Value (mm) | MAX Dev (mm) |
|---|---|---|---|
Ecoflex 00 10 | 3.79 | 4.32 | 16.35 |
Ecoflex 00 30 | 3.15 | 3.49 | 13.28 |
Ecoflex 00 50 | 2.55 | 2.82 | 10.96 |
Dragon Skin 10 | 2.05 | 2.33 | 9.92 |
Dragon Skin 30 | 1.10 | 1.32 | 7.50 |
[See PDF for image]
Fig. 25
Boxplot graph of displacement distributions for each type of silicone spaciment
From the results of these tests, it can be concluded that the system effectively distinguishes deformation differences among specimens with varying hardness properties. Furthermore, the system’s performance aligns with the predictions, as the DIC analysis efficiencies deformation values (and thus pliability) that correspond to the theoretical expectations (Figs. 24 and 25).
Conclusions
This study presents the development of a Digital Image Correlation (DIC)-based scanner for the objective evaluation of burn scar pliability, representing a significant advancement over current subjective assessment methods that suffer from operator dependency and lack of standardization. The integrated system combines precision hardware components with advanced image processing algorithms, achieving sensitive detection of mechanical deformation variations in scar tissue. The implementation of 3D-printed custom components has yielded a versatile design with strong potential for clinical adoption.
Validation studies using silicone phantoms with controlled mechanical properties have established the system’s capability to reliably discriminate between different stiffness levels, as evidenced by the high repeatability demonstrated in testing (mean deformation = 1.39 mm, SD = 0.07 mm).
It is important to note that the current validation has been limited to synthetic phantoms, and clinical studies with human subjects will be essential to establish meaningful correlations with existing scar assessment scales and therapeutic outcomes. The current image acquisition protocol requires manual positioning of the scanner, which can introduce operator-dependent variability. However, reference elements are available to assist in correctly aligning the area to be examined. Looking ahead, it would be possible to implement an automatic recognition system for the area to be scanned, such as using artificial intelligence, to ensure accurate alignment and reduce operator influence. While the system has been optimized for upper limb assessment, adaptation to other anatomical regions will require additional design modifications to ensure comprehensive clinical utility.
Future research directions should prioritize clinical validation studies to quantify relationships between DIC-derived parameters and established clinical metrics. The implementation of automated positioning systems would enhance measurement consistency.
Development of AI-assisted analysis algorithms would further improve reliability and clinical relevance.
This technological approach marks a substantial step forward in quantitative scar assessment, with potential to transform current practices in burn rehabilitation monitoring. The methodology’s fundamental principles show promise for extension to other clinical applications requiring precise soft tissue mechanical characterization, including postoperative monitoring and dermatological evaluation. Continued development and clinical validation will be crucial to realize the full potential of this DIC-based system in improving treatment personalization and outcome assessment in burn care and beyond.
Author contributions
The concept of the construction solution presented in the article is shared by all the authors. F. D. M. and R. F. handled the hardware design of the device, while F. D. M. and L. G. were responsible for the software development. The prototype was manufactured using 3D printing by F. D. M. F.D.M. wrote the main manuscript. All authors reviewed the manuscript.
Funding
No funding was received for this study.
Data availability
All data generated or analysed during this study are included in this published article [and its supplementary information files].
Declarations
Ethics approval, Consent to Participate, and Consent to Publish
Not applicable
Competing interests
The authors declare no competing interests.
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
1. Sheridan RL. Burns in children. in Journal of burn care and research. Lippincott Williams and Wilkins; 2017. pp. e618–24. https://doi.org/10.1097/BCR.0000000000000536.
2. Held M et al. Jun., Alteration of biomechanical properties of burned skin, Burns, vol. 41, no. 4, pp. 789–795, 2015, https://doi.org/10.1016/j.burns.2014.09.021
3. Tyack, Z; Simons, M; Spinks, A; Wasiak, J. A systematic review of the quality of burn Scar rating scales for clinical and research use. Feb; 2012; [DOI: https://dx.doi.org/10.1016/j.burns.2011.09.021]
4. Jaspers MEH, van Haasterecht L, van Zuijlen PPM, Mokkink LB. A systematic review on the quality of measurement techniques for the assessment of burn wound depth or healing potential, Mar. 01, 2019, Elsevier Ltd. https://doi.org/10.1016/j.burns.2018.05.015
5. van der Wal MBA, Tuinebreijer WE, Lundgren-Nilsson Å, Middelkoop E, van Zuijlen PPM. Differential item functioning in the observer scale of the POSAS for different Scar types. Qual Life Res. Sep. 2014;23(7):2037–45. https://doi.org/10.1007/s11136-014-0637-4.
6. Bozkurt M, Ceran F, Guvercin E, Filinte GT. A new method for scar tissue assessment: Modified POSAS observer scale, 2017, Elsevier Ltd. https://doi.org/10.1016/j.burns.2016.04.022
7. Idriss N, Maibach HI. Scar assessment scales: A dermatologic overview, 2009. https://doi.org/10.1111/j.1600-0846.2008.00327.x
8. Van Zuijlen PPM, Kruis R, Beverwijk Z. The patient and observer Scar assessment scale: A reliable and feasible tool for Scar evaluation, 2004, https://doi.org/10.1097/01.prs.00012207.28773.56
9. Van Der Wal M, Verhaegen P, Middelkoop E, Van Zuijlen P. A clinimetric overview of Scar assessment scales. J Burn Care Res. 2012;33(2). https://doi.org/10.1097/BCR.0b013e318239f5dd.
10. DeJong H, et al. Reliable protocol to objectively assess Scar stiffness using shear wave elastography. Ultrasound Med Biol. 2020;46(7):1614–29. https://doi.org/10.1016/j.ultrasmedbio.2020.03.003. A Novel.
11. Lee KC, et al. Burns objective Scar scale (BOSS): validation of an objective measurement devices based burn Scar scale panel. Burns. 2020;46(1):110–20. https://doi.org/10.1016/j.burns.2019.05.008.
12. Promny, D; Billner, M; Reichert, B. Objective burn depth assessment of hand burns. Handchirurgie Mikrochirurgie Plastische Chirurgie; 2019; 51,
13. Verhaegen PDHM, Van Der Wal MBA, Middelkoop E, Van Zuijlen PPM. Objective scar assessment tools: A clinimetric appraisal, Plast Reconstr Surg, vol. 127, no. 4, pp. 1561–1570, Apr. 2011, https://doi.org/10.1097/PRS.0b013e31820a641a
14. Oliveira GV, Chinkes D, Mitchell C, Oliveras G, Hawkins HK, Herndon DN. Objective Assessment of Burn Scar Vascularity, Erythema, Pliability, Thickness, and Planimetry.
15. Lye, I; Edgar, DW; Wood, FM; Carroll, S. Tissue tonometry is a simple, objective measure for pliability of burn scar: is it reliable?. J Burn Care Res; 2006; 27,
16. Mura FD. Automatic strategy for objective evaluation of burn scars roughness on 3D scans.
17. Mueller B, Elrod J, Distler O, Schiestl C, Mazza E. On the reliability of Suction measurements for skin characterization. J Biomech Eng. Feb. 2021;143(2). https://doi.org/10.1115/1.4047661.
18. Busche MN, Thraen A-CJ, Gohritz A, Rennekampff H-O, Vogt PM. Burn Scar evaluation using the cutometer ® MPA 580 in comparison to ‘patient and observer Scar assessment scale’ and ‘vancouver Scar scale’. J Burn Care Res. 2018;39(4):516–26. https://doi.org/10.1093/jbcr/irx009.
19. Abbas DB et al. Jun., Standardizing Dimensionless Cutometer Parameters to Determine In Vivo Elasticity of Human Skin, Adv Wound Care (New Rochelle), vol. 11, no. 6, pp. 297–310, 2022, https://doi.org/10.1089/wound.2021.0082
20. Van Nuffel M et al. Mar., Palmar skin elasticity measured by the Cutometer MPA 580 is decreased in mild Dupuytren’s disease compared to healthy controls, Hand Ther, vol. 27, no. 1, pp. 14–21, 2022, https://doi.org/10.1177/17589983211061616
21. Mura, FD; Furferi, R; Governi, L; Puggelli, L. Application of a computational method based on 3D scans for burn Scar topology characterisation. Comput Aided Des Appl; 2023; 20, pp. 37-49. [DOI: https://dx.doi.org/10.14733/cadaps.2023.S6.37-49]
22. Mura FD, Furferi R, Governi L, Puggelli L. Application of a computational method based on 3D scans for burn Scar topology characterisation, https://doi.org/10.14733/cadaps.YYYY.aaa-bbb
23. Mura FD, Carfagni M, Volpe Y. Improvement of the LPG-D probe with sensors and electronics for the objective assessment of burn scars in pediatric cases. Inst Electr Electron Eng (IEEE). Jul. 2024;482–7. https://doi.org/10.1109/metroind4.0iot61288.2024.10584172.
24. Dalle Mura F, Governi L, Furferi R, Cervo M, Puggelli L. Towards the Development of a Device for Assessing the Pliability of Burn Scars, Front Bioeng Biotechnol, vol. 10, Jun. 2022, https://doi.org/10.3389/fbioe.2022.856562
25. Gdoutos EE. Digital Image Correlation (DIC), in Solid Mechanics and its Applications, vol. 269, Springer Science and Business Media B.V., 2022, pp. 251–259. https://doi.org/10.1007/978-3-030-89466-5_12
26. Sutton, MA; Orteu, JJ; Schreier, HW; Reu, P. Introduction to digital image correlation: best practices and applications. Jan; 2012; [DOI: https://dx.doi.org/10.1111/j.1747-1567.2011.00798.x]
27. Rizzuto E, Carosio S, Prete ZD, Rizzuto CE. Characterization of a digital image correlation system for dynamic strain measurements of small biological tissues, 2014, https://doi.org/10.1007/s40799-016-00
28. Palanca, M; Tozzi, G; Cristofolini, L. The use of digital image correlation in the Biomechanical area: A review. Taylor Francis Ltd; 2016; [DOI: https://dx.doi.org/10.1080/23335432.2015.1117395]
29. Zhang, D; Arola, DD. Applications of digital image correlation to biological tissues. J Biomed Opt; 2004; 9,
30. Bruck H, McNeill S, Sutton M, Peters W. III, Digital Image Correlation Using Newton-Raphson Method of Partial Differential Correction.
31. WH Peters, RW. Digital imaging techniques in experimental mechanics. Opt Eng; 1982; 21, pp. 427-31.
32. Optimal. smoothing, Mathematics in Science and Engineering, vol. 141, pp. 1–22, Jan. 1982, https://doi.org/10.1016/S0076-5392(08)63221-X
33. Sutton W. MA, Determination of displacements using an improved digital correlation method. Image Vis Comput 1:133–9.
34. Sutton Y. A, The effect of out-of-plane motion on 2D and 3D digital image correlation measurements. Opt Lasers Eng 46:746–57.
35. Lamberti LM. Advancement of optical methods&digital image correlation in experimental mechanics, vol 3. Springer, Berlin.
36. Lin S. M-T, Advancements in optical methods & digital image correlation in experimental mechanics, vol 3. Springer, Berlin.
37. Freddi A, Olmi G, Cristofolini L. Springer Series in Solid and Structural Mechanics 4 Experimental Stress Analysis for Materials and Structures Stress Analysis Models for Developing Design Methodologies. [Online]. Available: http://www.springer.com/series/10616
38. Brugo TM, Campione I, Minak G. Investigation by digital image correlation of mixed-mode I and II fracture behavior of polymeric IASCB specimens with additive manufactured crack-like notch, Materials, vol. 14, no. 5, pp. 1–15, Mar. 2021, https://doi.org/10.3390/ma14051084
39. Lionello G, Cristofolini L. A practical approach to optimizing the Preparation of speckle patterns for digital-image correlation. Meas Sci Technol. Oct. 2014;25(10). https://doi.org/10.1088/0957-0233/25/10/107001.
40. Ncorr. V1.2. Accessed: Oct. 24, 2024. [Online]. Available: https://ncorr.com/
41. Documentation BP. acA4112-20um datasheet. Accessed: Aug. 27, 2024. [Online]. Available: https://docs.baslerweb.com/aca4112-20um
42. IDS Imaging Development Systems. Ui 3200SE. Accessed: Aug. 27, 2024. [Online]. Available: https://en.ids-imaging.com/store/ui-3200se.html
43. Vinaik R, Fish J, Jeschke MG. Burn hypertrophic Scar in pediatric patients: clinical case. in Textbook on Scar management. Springer International Publishing; 2020. pp. 517–21. https://doi.org/10.1007/978-3-030-44766-3_60.
44. Actuonix. Miniature Linear Motion Series · L12, 2019. [Online]. Available: http://www.actuonix.com/terms.asp
45. Solav D, Silverstein A. DuoDIC: 3D digital image correlation in MATLAB. J Open Source Softw. Jun. 2022;7(74):4279. https://doi.org/10.21105/joss.04279.
46. Blaber J, Adair B, Antoniou A. Ncorr: Open-Source 2D Digital Image Correlation Matlab Software, Exp Mech, vol. 55, no. 6, pp. 1105–1122, Jul. 2015, https://doi.org/10.1007/s11340-015-0009-1
© The Author(s) 2025. This work is published under http://creativecommons.org/licenses/by-nc-nd/4.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.