An Assessment of the Efficacy of Erbium Glass Laser Therapy in Acne Scar Treatment Using Image Analysis and Processing Methods

Abstract

Background: Acne scars significantly impact skin texture and esthetics, necessitating effective treatment modalities. This study evaluates the efficacy of erbium glass laser therapy in improving atrophic acne scars using advanced image analysis techniques. Materials and methods: Twenty patients with mild to moderate atrophic scars underwent two sessions of 1550 nm erbium glass laser treatment. The clinical photographs were analyzed using a Gray-Level Co-occurrence Matrix (GLCM) to assess changes in contrast and homogeneity across the grayscale and RGB channels. The analysis revealed statistically significant improvements post-therapy, including reduced contrast and increased homogeneity, indicating a smoother and more uniform skin texture. The blue and green channels demonstrated the greatest sensitivity to surface-level textural changes, while the red channel exhibited the smallest differences, reflecting its deeper penetration and reduced sensitivity to surface alterations. Conclusions: These findings underscore the value of quantitative imaging techniques in dermatology for objectively evaluating therapeutic outcomes and optimizing treatment strategies. Erbium glass laser therapy emerges as a non-invasive and effective solution for acne scar management.

1. Introduction

The skin, as the body’s largest organ, serves not only as a protective barrier but also as a key indicator of systemic health and well-being. Its appearance is influenced by the interplay of structural components and chromophores, such as melanin and hemoglobin, which contribute to pigmentation and vascular tone. These chromophores, along with collagen and elastin in the dermis, define the texture, elasticity, and overall visual quality of the skin [1,2].

Acne scars, which result from chronic inflammation of sebaceous glands, form due to insufficient production of fibrous connective tissue during the healing process. This leads to depressions and irregularities in the skin’s surface, significantly altering its texture and appearance. The formation of these scars involves the epidermis, dermis, and subcutaneous tissue. Prolonged inflammation within the pilosebaceous unit causes tissue damage, leading to the contraction and weakening of collagen fibers, as well as a reduction in subcutaneous fat under the influence of inflammatory mediators. These processes create visible deficits in the skin [3,4,5].

Atrophic acne scars present in various forms depending on their shape and depth. Ice-pick scars are narrow and deep, often extending into the subcutaneous tissue, while boxcar scars have sharply defined edges and deep, flat bases. Rolling scars, in contrast, feature shallow depressions with soft, wavy contours and indistinct edges [6,7,8].

In recent years, laser therapy has revolutionized the management of acne scars, offering a precise, minimally invasive, and highly adaptable treatment option. Among the available laser technologies, the erbium glass laser (1550 nm) has attracted significant attention for its ability to provide non-ablative treatment. In contrast to ablative lasers, which cause fractional tissue vaporization, erbium glass lasers denature proteins without inducing their vaporization, allowing for collagen remodeling and texture improvement while preserving the epidermis. This approach minimizes downtime, reduces the risk of side effects, and ensures better patient comfort during and after the procedure [9,10,11].

The efficacy of erbium glass lasers lies in their ability to target water molecules in the dermis, delivering controlled heat to initiate neocollagenesis and improve the skin’s elasticity. This makes them particularly suitable for treating atrophic scars, where collagen loss plays a central role in scar formation. Despite its widespread clinical use, objective and reproducible methods to assess treatment outcomes remain limited, often relying on subjective visual assessments.

In clinical dermatology, the assessment of scar severity and treatment outcomes often relies on a visual examination [12]. However, this approach is inherently subjective, influenced by individual perception, variations in skin tone, and subtle differences in scar texture, leading to potential inconsistencies and misinterpretations. Although visual assessments provide a general overview, they lack the required precision for consistent and reproducible evaluations.

Photography is essential for assessing acne scars, serving as a standardized tool for documenting skin condition before and after treatment [13,14,15,16]. However, conventional clinical photography often lacks quantitative precision in evaluating scar texture, depth, and irregularities. To address this limitation, our study integrates advanced texture analysis methods, including the Gray-Level Co-occurrence Matrix (GLCM) with color-channel separation, to enhance the accuracy of scar assessment. By isolating the R, G, and B channels before texture extraction, we improve the detection of subtle structural differences in atrophic scars, such as ice-pick, boxcar, and rolling scars. This combination of visual assessment and quantitative image analysis provides a more precise method for monitoring treatment outcomes and contributes to the development of more effective therapeutic protocols.

Among these, image analysis techniques such as the Gray-Level Co-occurrence Matrix (GLCM) have shown significant potential for quantifying skin texture and homogeneity. By analyzing spatial relationships between pixel intensities, the GLCM provides a robust framework for assessing subtle textural changes in scarred skin, offering a quantitative measure of therapeutic outcomes [17]. In this study, we focus on leveraging quantitative image analysis to enhance the evaluation of therapeutic outcomes of acne scar treatment. Furthermore, our study analyzes individual RGB color channels, identifying which wavelength range is most sensitive to laser-induced texture modifications. This aspect is particularly novel, as prior research has not systematically explored the role of color-channel segmentation in dermatological image analysis for laser therapy evaluation. By isolating individual color channels (red, green, and blue) from clinical RGB images and applying the Gray-Level Co-occurrence Matrix (GLCM) method, we aim to identify the channel that best captures the textural changes associated with scar remodeling. This approach not only addresses the limitations of subjective assessments but also provides insights into the optical and structural properties of scars, which are crucial for optimizing therapeutic strategies.

Therefore, GLCM analysis is utilized to validate treatment efficacy by assessing texture analysis metrics, providing a reproducible and objective alternative to subjective clinical observations.

By integrating clinical photography with advanced image processing, we aim to establish a more objective framework for assessing laser therapy outcomes. This approach not only enhances treatment monitoring but also paves the way for standardized, data-driven methodologies in dermatological research.

2. Materials and Methods

2.1. Patients

This study involved 20 patients (10 men and 10 women) aged 28 to 40 years who had mild to moderate atrophic acne scars and no active acne lesions. The participants met the inclusion criteria, which excluded any recent isotretinoin use (within the last 6 months), filler injections or dermabrasion in the past year, chemical peels in the previous 3 months, or a personal history of hypertrophic scars or keloids. Pregnant or breastfeeding individuals were also excluded.

This study adhered to the principles outlined in the Declaration of Helsinki and was approved by the Ethics Committee of the Medical University of Silesia (approval number: PCN/0022/KB1/12/I/20, issued on 19 May 2020). All participants provided informed, voluntary consent to take part in the research.

2.2. Treatment

Prior to the procedure, patients cleansed their skin with mild soap. About an hour before treatment, a topical anesthetic cream (EMLA, AstraZeneca, Södertälje, Sweden) was applied to the skin, followed by disinfection of the treatment area with Kodan (Schülke & Mayr, Norderstedt, Germany). The procedure involved the use of a fractional 1550 nm erbium glass laser. Each patient underwent two sessions, carried out one month apart. The laser parameters included an impulse duration of 2.8 ms, pulse energy ranging from 30 to 35 mJ, 70 µm beam size, and a spot area of 0.0038 mm². The energy density applied was 921 J/cm² (calculated as 0.035 J/0.000038 cm²).

2.3. Clinical Photographs

A series of clinical photographs were taken using the Fotomedicus system (Elfo, Łódź, Poland). This system ensures standardized and reproducible imaging conditions, allowing for consistent documentation both before and after treatment. The system incorporates a “ghost image” feature, which overlays a previously captured image onto the screen as a reference, facilitating the alignment of new images to match the previous framing accurately. The use of fixed flash energy and a stable color temperature of light ensures faithful reproduction of skin tones and features, irrespective of external lighting conditions (Figure 1).

These features effectively minimize geometric distortions and variability, providing reliable and objective data for comparing pre- and post-treatment outcomes. The standardized methodology supports high-quality, repeatable photographic documentation, which is crucial for quantitative and qualitative assessments in clinical research [18,19].

The digital images captured during the study were initially stored in RAWcr2 format to preserve the highest possible level of detail and avoid data loss due to compression. This unprocessed format allows for enhanced flexibility in image post-processing and analysis. Regions of interest (ROIs) encompassing acne scars were arbitrarily delineated on the images. The images were converted into grayscale bitmaps, where each pixel is represented as a single intensity value ranging from 0 to 255. To normalize the data, each image was supplemented with two additional pixels: one black (0) and one white (255). This step ensured the full extension of the grayscale range, enhancing the accuracy of the analysis. The grayscale intensity of pixels was calculated as the average value of the R, G, and B channels:

$$\mathrm{Gray} \mathrm{Value}={\displaystyle \frac{\mathrm{R}+\mathrm{G}+\mathrm{B}}{3}}$$

From the generated bitmaps, Gray-Level Co-occurrence Matrices (GLCMs) were created.

2.4. GLCM Analysis

The GLCM (Gray-Level Co-occurrence Matrix) method is used to describe the texture of an image by evaluating the spatial relationships between pixel intensity values. It measures how often a pixel with a specific intensity occurs in a predefined spatial relationship to another pixel with a different or similar intensity. This approach allows for the extraction of textural features—such as contrast and homogeneity—from an image. In the GLCM, the spatial relationships between pixels are evaluated based on specific directions, defined by their angles, and the distances between reference pixels and their neighbors. The commonly used angles are 0°, 90°, 45°, and 135° [20,21]. In our study, we utilized a GLCM approach with a spatial relationship defined by an angle of 45° and a pixel distance of 1. This configuration was chosen to effectively capture textural features relevant to the analyzed images. For the analysis of these textural features, the Gray-Level Co-occurrence Matrix (GLCM) was calculated using the Accord.Imaging library. Accord.Imaging is part of the Accord.NET Framework, a comprehensive library for machine learning and image processing, and facilitates the efficient and accurate computation of the required metrics.

In our study, contrast and homogeneity were assessed.

Contrast is a parameter that measures the degree of variation between adjacent pixel intensities. High contrast indicates significant variability and is associated with rough or irregular textures, whereas low contrast suggests smoother surfaces. In homogeneous images, the contrast value approaches zero due to minimal intensity differences.

$$\sum _{i,j=0}^{N-1}{P}_{i,j}{(i-j)}^2$$

Homogeneity assesses the uniformity of the pixel intensity distribution within an image. Its value ranges from 0 to 1, with higher values reflecting greater similarity among the pixels. Homogeneity reaches its maximum value (1) when all pixels share identical intensity levels, signifying a completely uniform texture.

$$\sum _{i,j=0}^{N-1}{\displaystyle \frac{P_{i,j}}{1+(i-j{)}^2}}$$

where:

• $i$—the brightness of the tested pixel;
• $j$—the brightness of the neighboring pixel.

2.5. Statistical Analysis

To present the obtained results on a quantitative scale, descriptive statistical methods were used, including the arithmetic mean, median (Me), standard deviation (SD), interquartile range (IQR), minimum (Min), and maximum (Max). To assess the conformity of the distribution of the studied variables with the normal distribution, the Shapiro–Wilk test was applied. A paired Student’s t-test was used to evaluate the difference between measurements taken before and after therapy. The strength of the relationship between the studied variables was assessed using Pearson’s correlation test. Results were considered statistically significant at p < 0.05. Statistical analysis was conducted using the GraphPad Prism v.9.0 software.

3. Results

3.1. Contrast

The mean contrast before the therapy was 7.12 (±2.12), while after the therapy, it was 6.36 (±1.86). It was shown that the contrast decreased by an average of 0.76 (±0.88). This reduction was found to be statistically significant (p = 0.0010) (Figure 2a) (

Table 1. Comprehensive statistical summary of texture analysis parameters before and after erbium glass laser therapy.

	Mean	Me	Min	Max	IQR	SD	p, Student’s t-Test
Contrast in average RGB
before	7.12	6.76	3.98	11.75	3.22	2.12	0.0010
after	6.36	6.30	3.17	9.85	2.24	1.86
delta	−0.76	−0.49	−3.59	0.50	0.88	0.88
Contrast channel R
before	5.97	6.18	1.89	8.73	1.69	1.72	0.0090
after	5.57	5.76	1.63	7.61	1.65	1.53
delta	−0.40	−0.21	−2.04	0.50	0.70	0.61
Contrast channel G
before	7.98	7.49	4.35	12.88	3.69	2.43	0.0067
after	7.24	7.22	4.06	11.34	2.91	2.26
delta	−0.74	−0.58	−4.45	0.78	1.05	1.08
Contrast channel B
before	13.17	12.57	7.86	21.35	4.94	3.40	0.0023
after	11.70	11.63	7.20	16.83	4.10	2.98
delta	−1.47	−1.01	−5.42	2.25	1.53	1.87
Homogeneity in average RGB
before	0.43	0.43	0.33	0.52	0.06	0.05	0.0081
after	0.45	0.44	0.36	0.56	0.06	0.05
delta	0.02	0.01	−0.04	0.05	0.03	0.02
Homogenity channel R
before	0.48	0.45	0.36	0.81	0.07	0.12	0.0377
after	0.49	0.46	0.39	0.86	0.06	0.11
delta	0.01	0.01	−0.04	0.04	0.02	0.02
Homogenity channel G
before	0.41	0.41	0.32	0.50	0.07	0.05	0.0278
after	0.43	0.42	0.34	0.51	0.08	0.05
delta	0.01	0.01	−0.05	0.05	0.03	0.02
Homogenity channel B
before	0.35	0.35	0.26	0.42	0.05	0.04	0.0437
after	0.36	0.36	0.29	0.43	0.06	0.04
delta	0.01	0.01	−0.06	0.05	0.04	0.03

).

3.2. Contrast in Channel R

The mean contrast in the R channel before the therapy was 5.97 (±1.72), while after the therapy, it was 5.57 (±1.53). It was shown that the contrast in the R channel decreased by an average of 0.40 (±0.61). This reduction was found to be statistically significant (p = 0.0090) (Figure 2b).

3.3. Contrast in Channel G

The mean contrast in the G channel before the therapy was 7.98 (±2.43), while after the therapy, it was 7.24 (±2.26). It was shown that the contrast in the G channel decreased by an average of 0.74 (±1.08). This reduction was found to be statistically significant (p = 0.0067) (Figure 3c).

3.4. Contrast in Channel B

The mean contrast in the B channel before the therapy was 13.17 (±3.40), while after the therapy, it was 11.70 (±2.98). It was shown that the contrast in the B channel decreased by an average of 1.47 (±1.87). This reduction was found to be statistically significant (p = 0.0023) (Figure 3d).

3.5. Homogeneity

The mean homogeneity before the therapy was 0.43 (±0.05), while after the therapy, it was 0.45 (±0.05). It was shown that the homogeneity increased by an average of 0.02 (±0.02). This increase was found to be statistically significant (p = 0.0081) (Figure 4a).

3.6. Homogeneity in Channel R

The mean homogeneity in the R channel before the therapy was 0.48 (±0.12), while after the therapy, it was 0.49 (±0.11). It was shown that the homogeneity in the R channel increased by an average of 0.01 (±0.02). This increase was found to be statistically significant (p = 0.0377) (Figure 4b).

3.7. Homogeneity in Channel G

The mean homogeneity in the G channel before the therapy was 0.41 (±0.05), while after the therapy, it was 0.43 (±0.05). It was shown that the homogeneity in the G channel increased by an average of 0.01 (±0.02). This increase was found to be statistically significant (p = 0.0278) (Figure 4c).

3.8. Homogeneity in Channel B

The mean homogeneity in the B channel before the therapy was 0.35 (±0.04), while after the therapy, it was 0.36 (±0.04). It was shown that the homogeneity in the B channel increased by an average of 0.01 (±0.03). This increase was found to be statistically significant (p = 0.0437) (Figure 4d).

Based on our analysis, a significant correlation was observed between the changes in contrast and changes in homogeneity (delta homogeneity) in the analyzed images. Specifically, a statistically significant negative correlation was found between contrast changes (delta contrast) and homogeneity changes (delta homogeneity) (r = −0.6030; p = 0.005).

For the variable contrast channel R’s delta, a significant negative correlation was observed with homogeneity channel R (r = −0.5919; p = 0.006). Similarly, for the variable contrast channel G, a significant negative correlation was noted with homogeneity channel G (r = −0.6623; p = 0.001). Lastly, for the variable contrast channel B, a significant negative correlation was identified with homogeneity channel B (r = −0.8469; p < 0.0001).

4. Discussion

The analysis of the outcomes of acne scar treatment using an erbium glass laser demonstrated statistically significant changes in skin texture, as evidenced by alterations in the contrast and homogeneity parameters. Across all channels—red (R), green (G), blue (B), and the combined RGB—the treatment resulted in a decrease in contrast and an increase in homogeneity. These findings are consistent with the hypothesis that laser therapy effectively reduces the irregularities in skin texture that are associated with atrophic scars, leading to smoother and more uniform skin.

The observed decrease in contrast signifies a reduction in the variability of the adjacent pixel intensities, which correlates with a flattening of irregularities and an improvement in the overall texture of the skin. Simultaneously, the increase in homogeneity indicates a more uniform distribution of pixel intensities, further emphasizing the smoothing effect of the treatment. These changes were statistically significant for each analyzed channel and the overall RGB images, reinforcing the efficacy of erbium glass laser therapy for acne scar management.

When analyzing the individual RGB channels, distinct patterns emerged. The depth of light penetration varies across the channels of visible light, with blue light penetrating the shallowest and red light reaching the deepest layers of the skin [22]. The blue channel exhibited the most pronounced changes in contrast and significant changes in homogeneity, but not the largest increase among all channels. Its shallow penetration depth makes it particularly sensitive to surface-level textural changes, allowing it to capture subtle variations in skin surface irregularities. Additionally, the blue channel effectively highlights shallow discoloration changes, making it valuable for detecting subtle pigmentation irregularities. However, its sensitivity to surface-level changes may limit its ability to capture deeper textural alterations, which could reduce its utility for assessing treatments targeting subdermal structures. The higher baseline contrast in the blue channel might also make it more responsive to treatment-induced changes.

The green channel also showed significant changes, which were comparable to or potentially exceeding those observed in the blue channel, depending on the specific parameter analyzed. The green channel is known for its balanced sensitivity to both vascular and structural components of the skin, making it a reliable indicator of textural improvements.

The red channel displayed the smallest changes in both contrast and homogeneity. Its ability to penetrate deeper into the skin layers makes it less responsive to surface-level changes but valuable for assessing deeper structural alterations. This is consistent with findings from the study by Hantash et al., who demonstrated that larger laser spot sizes produce deeper thermal lesions while maintaining better epidermal integrity [23]. Although our study utilized a 70 µm spot size, which is smaller than the maximum 140 µm referenced in their research, it can be presumed that we primarily affected the superficial layers of the skin. Despite this, we still achieved significant efficacy in reducing acne scars. This suggests that the red channel’s deeper penetration correlates with its role in identifying changes in deeper skin layers, which may not manifest as visibly pronounced surface-level improvements.

The analysis of the combined RGB images integrates information from all three channels, providing a comprehensive overview of the treatment’s impact on skin texture. The changes in contrast and homogeneity in these images—which reflect data derived from photographs without isolating individual color channels—further highlight the smoothing and unifying effects of the therapy. These findings emphasize that evaluating the RGB composite can offer a robust, holistic measure of treatment efficacy while ensuring that no significant detail is overlooked by isolating the channels. The statistically significant reduction in contrast and increase in homogeneity for the combined RGB images further validate the overall effectiveness of the erbium glass laser therapy. This holistic approach captures a broader range of textural changes, ensuring that both subtle and pronounced effects are accounted for.

To comprehensively evaluate skin texture changes, various image analysis methods have been employed in dermatology, each with distinct advantages and limitations. GLCM is particularly effective for assessing spatial dependencies between pixel intensities, making it well-suited for quantifying scar remodeling. In medical sciences, the GLCM matrix has been used to analyze computed tomography images, determine breast tissue density, characterize the organization of fibrillar collagen and assess cellulite [24,25,26,27]. However, alternative approaches offer different perspectives on texture evaluation. Local Binary Patterns (LBP), for example, encode local intensity differences into binary patterns, providing a computationally efficient and rotation-invariant method [28]. While LBP is widely used for texture classification, it lacks the ability to capture co-occurrence relationships between pixels, making it less sensitive to subtle surface changes in acne scars. Gabor Wavelet Transform (GWT) is another tool for analyzing skin texture, capable of extracting multi-scale and multi-orientation features, which makes it particularly useful for skin disease recognition [29]. However, its high computational cost and sensitivity to parameter tuning limit its routine clinical application. Similarly, Wavelet Transform (WT) has been employed for multi-resolution texture analysis, effectively distinguishing fine skin structures, but its reliance on frequency decomposition can make interpretation challenging in clinical dermatology [28]. Compared to alternative methods, GLCM demonstrates favorable performance in terms of computation time. In the comparative study by Ou et al. [29], GLCM showed faster feature extraction than Gabor Wavelet Transform (GWT), which, although powerful in capturing multi-scale and multi-orientation features, was found to be computationally demanding.

Our findings underscore the importance of utilizing quantitative image analysis methods, such as GLCM, in evaluating the efficacy of acne scar treatments. However, GLCM analysis can have limitations, such as its dependence on high-quality, standardized images and the need for careful parameter selection to avoid misinterpretation of textural features. These challenges highlight the necessity of robust imaging protocols and expert oversight during analysis. The integration of the GLCM method in evaluating the impact of acne scar treatments aligns with findings from Wawrzyk-Bochenek et al. on hyperpigmentation reduction using microneedle mesotherapy [30]. Both studies underscore the significance of using precise quantitative techniques, such as the GLCM, to assess skin texture changes objectively. While the current study focuses on contrast and homogeneity improvements post erbium glass laser therapy, the methodology parallels the hyperpigmentation analysis, which showed reduced GLCM contrast and increased homogeneity after mesotherapy. This reinforces the broader applicability of the GLCM as a robust analytical tool for various dermatological treatments. Future studies could benefit from comparing these modalities to optimize approaches for complex skin conditions involving both scars and pigmentation changes.

There are reports on the effectiveness of Er:glass lasers for reducing acne scars; however, most of these studies rely on measurement scales that are based on subjective evaluations by researchers [9,10]. Noteworthy are the few studies that confirm the effectiveness of Er:glass laser therapy using objective, quantitative methods of evaluation. For example, Naranjo et al. [31] reported a 49.5% reduction in roughness volume, a 46% reduction in roughness-affected area, and a 24.6% decrease in maximum depth depression following five sessions of 1540-nm non-ablative fractional Er:glass laser therapy, measured using the Antera 3D system. Furthermore, Rongsaard and Rummaneethorn [32] conducted a randomized split-face clinical trial comparing a 1550-nm fractional erbium-doped glass laser with fractional bipolar radiofrequency. The effectiveness of the treatment was assessed using three complementary methods: evaluation of acne scar improvement by three independent dermatologists based on photographic comparisons, quantitative analysis of skin texture using the VISIA^® Complexion Analysis System, and patient-reported satisfaction using a graded scale. The study concluded that the erbium- glass laser is an effective and safe modality for treating atrophic acne scars in patients with Fitzpatrick skin types III to V. Specifically, the mean acne scar improvement grades rated by independent dermatologists were 2.86 ± 0.42 (on a 0–4 scale), while patients rated the improvement at 2.89 ± 0.57. Texture analysis also showed a statistically significant reduction in surface irregularities, with a mean texture score reduction of 2.94 ± 1.84 (p < 0.001) after treatment. This reliance on subjective assessments can introduce variability and limit the robustness of findings. Incorporating quantitative methods, such as GLCM analysis, offers a more objective and reproducible approach to evaluating therapeutic outcomes, enabling more precise verification of treatment efficacy. The conducted GLCM analysis is particularly useful for the quantitative assessment of acne scars. Notably, the analysis of homogeneity and contrast based on the GLCM enables the identification of even subtle changes that are imperceptible to the naked eye. Additionally, the method’s repeatability (unlike the subjective visual evaluation by a specialist) and its ability to quantitatively compare acne-related changes, such as those observed before and after therapy, provide unique opportunities for verifying the effectiveness of various scar treatment methods. However, to achieve high sensitivity and specificity and to accurately reflect the scar severity, it is crucial to capture high-quality, reproducible skin images. Proper lighting selection plays a key role in obtaining clinical photographs that are suitable for further analysis. The significant improvements observed in both the contrast and homogeneity parameters align with the clinical expectations of smoother and more uniform skin post treatment. Moreover, the distinct responses of individual RGB channels highlight the value of channel-specific analyses in capturing diverse aspects of skin texture changes. Specifically, the blue and green channels may serve as more sensitive indicators of surface-level improvements, while the red channel provides complementary information on deeper skin changes. One aspect that is worth considering in future studies is the selection of pixel distance and angular direction in GLCM analysis. While this study utilized a pixel distance of 1 and an angle of 45°, these parameters may not be universally optimal for all types of scars. For instance, increasing the pixel distance could help capture broader textural patterns. Similarly, other angles, such as 0° or 90°, might better align with the specific orientation or structure of certain scar types.

Recent advances in artificial intelligence (AI)-assisted imaging highlight the potential for further enhancing acne scar assessment. AI-based methodologies, particularly machine learning (ML) algorithms, have been successfully integrated with biomedical imaging techniques such as Light-Sheet Fluorescence Microscopy (LSFM) to improve post-imaging analysis and automate feature detection [33]. While our study relies on the GLCM method for quantitative texture analysis, incorporating AI-driven techniques could provide deeper insights into treatment efficacy. AI-enhanced image processing could refine acne scar evaluation by enabling more sophisticated feature extraction, automated classification of treatment response, and predictive modeling of long-term outcomes. Future research should explore the synergy between AI and GLCM to further enhance the objectivity and clinical applicability of texture analysis in dermatology.

Study Limitations

• Limited sample size—20 participants, reducing generalizability.
• Short follow-up—No long-term outcome assessment.
• High group homogeneity—Similar phototypes and scar types.
• No treatment comparison—Lacked evaluation against other methods.

5. Conclusions

In conclusion, erbium glass laser therapy was proven to be an effective modality for improving the texture and uniformity of atrophic acne scars. These findings highlight its potential as a key tool in clinical dermatology, offering a non-invasive solution to enhance patient outcomes. Future research could focus on optimizing treatment parameters and exploring the treatment’s efficacy across diverse patient populations, while clinicians may integrate these insights to refine therapeutic strategies and ensure tailored care for individuals with acne scars. The combination of reduced contrast and increased homogeneity across all analyzed channels provides robust evidence of the therapeutic benefits of this treatment, reinforcing the utility of quantitative image analysis in dermatological research.