The Application of High-Frequency Ultrasonography in Post-Therapeutic Assessment of Actinic Keratosis After Photodynamic Therapy

Simple Summary

Actinic keratosis (AK) is one of the most frequent reason for consultations in dermatological offices. Many treatment options are available, one of them being photodynamic therapy (PDT), an in-office method with high treatment efficacy and acceptable cosmetic effects. This study aimed to evaluate the changes observed in a non-invasive skin imaging method—high-frequency ultrasonography (HFUS)—after PDT. We observed the decrease in SLEB after therapy and showed that this parameter maybe useful in monitoring the effects of treatment and, if not reduced completely, possibly indicating a potential risk of relapse. Additionally, for the first time, we propose the use of new USG parameters in this setting, i.e., LEP, HEP, MEP, homogeneity, and EPI, which present the possibility of overall assessment of patients after PDT, taking into account skin analysis at all levels. Our results show the improvement in skin texture mirrored in the analyzed parameters corresponding to the clinical pictures.

Abstract

Objectives: Actinic keratoses (AKs) are one of the most common reasons for consultation in the elderly population. This study aimed to assess the efficacy of 5-ALA PDT in AK treatment using high-frequency ultrasonography (HFUS) to evaluate skin layer changes during therapy. Methods: In our study, we included 44 AK patients aged 53 to 89 years. All patients had lesions clinically evaluated with the Olsen and AKASI scale. HFUS imaging was performed on seemingly healthy skin and lesions before and at 4, 8, and 12 weeks of therapy. Ultrasound markers such as skin thickness, echogenicity, and pixel intensity were measured. 5-ALA was applied under occlusion for 3 h. After removing the occlusive dressing, 5-ALA was removed with a saline solution and a directed therapy with a BF-200 lamp. Full follow-ups of 56 markers of suitable quality were selected. Results: The thickness of SLEB significantly decreased in the following weeks compared to the pre-therapy results, reaching its lowest values after 12 weeks. The average pixel intensity significantly increased in each skin layer after therapy (p < 0.01). For SLEB, there were statistically significant differences in LEP, MEP and contrast. The AKASI score before and after treatment was determined for the 39 patients who underwent follow-up at week 12. The median AKASI score was 3.2 (1.2–8.6) before treatment and 0.6 (0–2.8) after. Conclusions: According to the literature data, this is the first study describing the ALA-PDT treatment efficacy in different AK severities evaluated in HFUS. HFUS provides a valuable non-invasive tool for monitoring the efficacy of PDT in AK treatment, showing significant improvements in skin texture and structure.

1. Introduction

Actinic keratoses (AKs) are one of the most common reasons for consultation in the elderly population [1]. They are associated with extensive, cumulative sun exposure, predominantly in phototypes I–II on the Fitzpatrick scale [2,3]. Ultraviolet (UV) radiation affects the cells, leading to oxidative stress, and this impacts tumor suppression protein, especially p53 [4], which might contribute to the progression of Squamous Cell Carcinoma (SCC). A risk factor that may increase the transformation to malignancy is immunosuppression. Therefore, this group of patients should be under a regular follow-up [5].

Clinically, AKs present as red, scaly, well-circumscribed lesions that are located on the scalp, face (most frequently), and extremities [6]. Usually, they are graded according to the Olsen scale, which is based on an assessment of the thickness of the lesion and the presence of scales. Lesions classified as grade 1 are invisible but palpable; in grade 2, they are visible and palpable; in grade 3, they are very thick and hyperkeratotic [7]. In recent years, many non-invasive skin imaging techniques have allowed for a precise diagnosis of AKs without taking unnecessary biopsies. The application of dermatoscopy or reflectance confocal microscopy allows for the determination of an accurate diagnosis before the treatment [8,9].

High-frequency ultrasonography (HFUS) is a widely used, non-invasive method that has been used in dermatology for many years. Higher frequencies (18–20 MHz) allow us to visualize tumor infiltration before surgical excision, especially in melanoma or basal cell carcinoma [10,11,12], while the applications of this method in AKs are limited. The tumors usually manifest as anechoic or hypoechoic oval structures, with the possibility of subepidermal low-echogenic band formation underneath the entry echo (known as SLEB). This parameter also allows the monitoring of treatment efficacy in some other dermatologic entities such as psoriasis, mycosis fungoides, or atopic dermatitis [13,14,15,16].

Treatment modalities for AKs start with proper photoprotection such as sunscreen and protective clothing; however, they are usually combined with various therapeutic methods. There are many available options that feature in or out of office settings. The treatment efficacy varies between methods, and is dependent on the number of lesions, age, and patient compliance [1]. It is usually assessed with an AKASI score that objectively allows monitoring treatment outcomes with different modalities [17].

The currently available therapeutics include 5-fluorouracil cream, cryosurgery, diclofenac 3% gel, imiquimod, photodynamic therapy (PDT), and, the newest method, tirbanibulin [18]. The main aim is to prevent progression to SCC. Since the clinical presentation or thickness does not predict the risk of transformation, prompt treatment is recommended [19].

One of the currently recommended options is photodynamic therapy (PDT), an in-office method that relies on the application of a photosensitizer under an occlusion. Usually, 5-aminolevulinic acid (5-ALA) or methyl-aminolevulinic (MAL) are used. After the occlusion is removed, the lesions might be evaluated with ultraviolet-induced fluorescence dermatoscopy to assess the fluorescence due to the presence of Protoporphyrin IX [20]. Then, the lesions are exposed to red (630 nm) or blue (417 nm) light [21,22]. A daylight option may also be chosen if weather conditions allow [23].

So far, there is not much information on the application of HFUS in AK—there has only been one study assessing its features with a 22–50 MHz transducer in 54 lesions. The most commonly described features for AKs have been the irregular basal border of the lesion and a regular surface [24,25].

So far, there are single reports on the use of HFUS in the assessment of AKs, including the use of this imaging method in the evaluation of the effects of therapy. Among others, the thickness and echogenicity of SLEB have been analyzed as parameters related to the presence of atypical keratinocytes, although it is known that inflammation or elastosis may affect the formation of this band [13,24,25]. Therefore, the aim of this study was to examine the usefulness of HFUS in the assessment of AKs using machine learning-based feature extraction analysis and to show how the skin affected by AK changes as a result of the use of PDT in a 3-month assessment.

2. Materials and Methods

We included 44 AK patients aged 53 to 89 years (median age: 73 years; 70% male) presenting at our department from June 2023 to May 2024 with clinically and dermatoscopically evident diagnoses of AK. The lesions were located on the face (27 patients) or on the scalp (17 patients). Patients with a prior dermatological treatment, ulcerated lesions, features of invasion in dermatoscopy, allergy to photosensitizers, and other chronic dermatoses in the treatment area were excluded from PDT treatment. All patients provided informed consent for the procedure. All patients underwent only one PDT session before the follow-up visits.

All patients were clinically evaluated with the Olsen [7] and AKASI scales [17], and then each lesion was marked in a photograph for follow-up visits. The HFUS images were acquired with a linear probe (20 MHz), B-mode scan (Dermascan C^®; Hadsund, Denmark). The axial and lateral resolutions were 80 and 200 µm, respectively. If possible, in examined patients, scans of skin without clinically evident AK lesions were also carried out within close or contralateral localization to the affected region, which served as a reference for healthy skin. Afterward, 5-ALA was applied within the lesioned skin under occlusion for 3 h. If necessary, a curette was used to remove the scales a few days before the procedure.

After removing the occlusive dressing, 5-ALA was cleaned with a saline solution, and a directed therapy with a BF-200 lamp, BF-RhodoLED, Biofrontera, Leverkusen, DE (narrow-emission spectrum of 635 nm ± 9 nm) was performed according to the manufacturer’s recommendations [26].

The patients were scheduled for follow-up visits at 4, 8, and 12 weeks after the treatment procedure (example HFUS images shown in Figure 1). During all visits, the non-invasive procedures within marked areas were repeated. Contralateral skin was used as a control unless it showed clinically and dermoscopically obvious signs of photodamage [27].

This study was approved by the Local Ethics Committee (Poznan University of Medical Sciences, Protocol Number 523/23).

2.1. Feature Extraction

In [28], skin layer contours were used in the estimation of skin features from images using a constant thickness of layers located below the epidermal contour. However, in our study, the markers were located at different points on the face and scalp. This, and differences in the age of the patients, can affect the thickness of the individual layers, so it was decided to contour the skin layers separately for each image. Expert outlines, supported by the machine learning method, were then used to extract features for entry echo, SLEB, and dermis [28,29,30].

In [31,32,33,34], the authors reported the relationship between the thickness of skin layers (entry echo, SLEB, and dermis) and skin condition. Based on the contours, the thickness of each layer was determined as the average thickness measured within the layer. For the entry echo, variation in the thickness across the marker area was also obtained. The thickness variation index (TVI) was calculated as the deviation in the thickness of the entry echo layer. A low value indicates that the skin layer has a consistent thickness, while higher values suggest the presence of regions with significantly different thicknesses.

The roughness of the skin surface was described using parameters determined from the outline of the entry echo layer. The surface roughness (SR) quantifies the variability in height along the upper edge of the layer. The complexity of the entry echo outline is also described by two other parameters: the ratio of the layer’s perimeter to its area and the ratio of the layer’s area to its convex hull (the smallest geometric contour of a shape). The perimeter-to-area ratio (PAR) describes the degree of irregularity in the skin layer’s outline relative to the area of the layer. A lower value suggests a smoother, more regular surface. The area-to-convex hull ratio (ACR) measures how closely the skin layer’s shape adheres to its convex hull (the smallest geometric contour of a shape). A lower value indicates a more irregular shape, whereas a value close to 1 suggests that the layer is smoother. All three parameters represent the degree of jaggedness and irregularity of the entry echo surface.

Skin echogenicity is a parameter used in the evaluation of skin aging [28,33] and progress of therapy [31,32]. As proposed in [31], pixels were divided based on their intensity into low- (<30), medium- (50–150), and high-echogenic (>200). Echogenicity features were then calculated as the ratio of low-echogenic pixels (LEPs), medium-echogenic pixels (MEPs), and high-echogenic pixels (HEPs) to all pixels within the skin layer area [28]. Echogenicity is complemented by information on the mean pixel intensity within each layer.

In addition, pixel intensity variability (PIV) was calculated as the standard deviation of the pixel intensities within a given skin layer. A higher variability means that there is a wider range of pixel intensities, indicating more variation in brightness or texture. Complementary to this parameter is the entropy of pixel intensity (EPI), quantifying the level of disorder or randomness in the layer texture. A low value suggests that the area is more homogeneous. The following textural features were also determined using a Gray-Level Co-occurrence Matrix (GLCM): correlation, homogeneity, energy, and contrast [28,35]. Correlation measures how linearly related the pixel values are within the area. A high value indicates that the pixel intensities exhibit a strong relationship, meaning that the image has well-defined and consistent patterns or textures. Homogeneity assesses how similar the pixel values are to one another across the layer. A high homogeneity value means that the pixel intensities are nearly the same throughout the region, indicating uniformity. Energy reflects the degree of uniformity or repetition in the image’s texture. High energy signifies that the image contains strong, repeatable patterns, such as regular shapes or uniform regions. Contrast quantifies the difference in gray levels between neighboring pixels. A high contrast value suggests that there are significant variations in intensity, indicating the presence of distinct textures. They provide useful information concerning the structure of the monitored regions, indicating the appearance of texture patterns or homogeneous areas.

2.2. Statistical Analysis

First, the AK changes over successive weeks of therapy were compared. The collected data met the assumptions for the Wilcoxon Signed-Rank test, comparing the positive ranks of the differences to the negative ranks of the differences. If a statistically significant difference was detected, a one-sided test was used to determine its nature. In assuming $\alpha$ = 0.05, test power = 0.9, and effect size = 0.5, the group size was estimated to be at least 51 samples. In addition, the effect of therapy on the AKASI score of fully treated patients was examined using the Wilcoxon Signed-Rank test. All calculations were performed using G*Power 3.1.9.7 [36,37].

The results obtained after 12 weeks of therapy were then compared with healthy skin recorded in the same group of patients using a Mann–Whitney test. First, a two-sided test was used to determine whether statistical differences existed between the post-treatment values and the healthy skin. When a statistically significant difference was detected, a one-sided test was used to determine the nature of this difference. Further analysis was performed in R Studio 2022.02.3.

3. Results

During the study, we examined 108 AKI, 53 AKII, and 36 AKIII. In total, 133 markers were registered at week 4, 72 at week 8, and 126 at week 12. However, full follow-up, i.e., recorded images for 4, 8, and 12 weeks, was achieved for 63 (32%) lesions. After carefully checking the quality of the recorded images, 56 (28%) markers were included in the study (see

Table 1. Inclusion and exclusion criteria for markers.

Inclusion	Exclusion	Number of Markers After Exclusion
Flitzpatrick skin type I–II	Previous dermatological treatment, other chronic dermatoses in the examination area. Allergy to photosensitisers.	N = 197
AKs diagnosed by dermatological examination	Inability to participate in a follow-up visit.	N = 63
	Insufficient quality of any of the images in the follow-up.	N = 56

). The number of healthy skin images recorded is 35, which is related to the clinically evident sun damage in the remaining patients.

3.1. Comparison of HFUS Skin Parameters Before Therapy and During Subsequent Follow-Up Visits

Statistical differences obtained for morphological features are presented in

Table 2. Parameters describing the morphology of the skin layers on ultrasound before therapy and at weeks 4, 8, and 12 of treatment. p-values < 0.05 are marked in bold. Small effect size is marked as *, moderate as **, and large as ***.

	Week 0	Week 4	p 0–4 Weeks	Effect Size	One-Way p	Week 8	p 0–8 Weeks	Effect Size	One-Way p	Week 12	p 0–12 Weeks	Effect Size	One-Way p
Entry echo
Thickness	0.1981 ± 0.02	0.1937 ± 0.03	0.596	0.0714 *		0.1848 ± 0.02	0.0285	0.2932 *	0.0143	0.1951 ± 0.02	0.8035	0.0338 *
TVI	0.0423 ± 0.01	0.0337 ± 0.01	0.001	0.4393 **	0.0004	0.0318 ± 0.01	0.0001	0.5853 ***	0.0001	0.0319 ± 0.01	0.0033	0.3935 **	0.0014
PAR	0.1716 ± 0.02	0.1687 ± 0.02	0.9707	0.0055 *		0.1731 ± 0.02	0.5597	0.0785 *		0.1637 ± 0.02	0.436	0.1046 *
SR	11.9893 ± 6.76	10.9937 ± 5.26	0.0568	0.2551 *		8.339 ± 5.04	0.0083	0.3532 **	0.0038	9.8148 ± 3.63	0.0747	0.2387 *
ACR	0.5084 ± 0.04	0.5503 ± 0.06	0.0695	0.2431 *		0.5804 ± 0.06	0.0035	0.3902 **	0.0015	0.5472 ± 0.06	0.0435	0.2703 *	0.0214
SLEB
Thickness	0.4039 ± 0.09	0.3208 ± 0.07	0.0001	0.6181 ***	0.0001	0.3411 ± 0.08	0.0006	0.4611 **	0.0003	0.2325 ± 0.11	0.0001	0.7423 ***	0.0001
Dermis
Thickness	1.5482 ± 0.21	1.469 ± 0.20	0.0336	0.2845 *	0.0164	1.3694 ± 0.18	0.0003	0.4829 **	0.0001	1.5124 ± 0.18	0.2946	0.1406 *

, and those for echogenicity and pixel intensity dispersion in

Table 3. Parameters describing the echogenicity and distribution of pixels before therapy at weeks 4, 8, and 12 of treatment. Values of p < 0.05 are marked in bold. Small effect size is marked as *, moderate as **, and large as ***.

	Week 0	Week 4	p 0–4 Weeks	Effect Size	One-Way p	Week 8	p 0–8 Weeks	Effect Size	One-Way p	Week 12	p 0–12 Weeks	Effect Size	One-Way p
Entry echo
LEP ratio	0.0415 ± 0.03	0.0119 ± 0.01	0.0001	0.6028 ***	0.0001	0.017 ± 0.01	0.0001	0.5472 ***	0.0001	0.011 ± 0.01	0.0001	0.6202 ***	0.0001
MEP ratio	0.3991 ± 0.09	0.281 ± 0.07	0.0011	0.4382 **	0.0004	0.3196 ± 0.07	0.0018	0.4186 **	0.0007	0.2808 ± 0.07	0.0009	0.4426 **	0.0004
HEP ratio	0.3951 ± 0.17	0.6076 ± 0.11	0.0001	0.6191 ***	0.0001	0.5616 ± 0.11	0.0001	0.5755 ***	0.0001	0.6325 ± 0.1	0.0001	0.6039 ***	0.0001
MPI	158.9785 ± 23.9	193.7715 ± 18.19	0.0001	0.6148 ***	0.0001	186.7654 ± 15.88	0.0001	0.5843 ***	0.0001	199.0126 ± 13.97	0.0001	0.6104 ***	0.0001
EPI	0.1823 ± 0.02	0.1801 ± 0.02	0.6984	0.0523 *		0.1718 ± 0.01	0.0186	0.315 **	0.0089	0.1817 ± 0.02	0.6864	0.0545 *
PIV	74.2705 ± 3.52	69.6094 ± 4.16	0.0029	0.399 **	0.0012	71.3299 ± 3.14	0.047	0.266 *	0.0232	68.1178 ± 3.71	0.0014	0.4262 **	0.0006
SLEB
LEP ratio	0.7556 ± 0.12	0.5564 ± 0.16	0.0001	0.5341 ***	0.0001	0.6123 ± 0.08	0.0006	0.4611 **	0.0002	0.4247 ± 0.2	0.0001	0.6617 ***	0.0001
MEP ratio	0.1129 ± 0.06	0.2119 ± 0.1	0.0001	0.6061 ***	0.0001	0.1923 ± 0.06	0.0001	0.5232 ***	0.0001	0.258 ± 0.13	0.0002	0.5025 ***	0.0001
HEP ratio	0 ± 0	0 ± 0	0.9653	0.002 *		0 ± 0	0.4513	0.1388 *		0 ± 0	0.3279	0.1804 *
MPI	27.069 ± 7.12	36.9444 ± 9.45	0.0001	0.5363 ***	0.0001	35.7611 ± 5.06	0.0009	0.4436 **	0.0003	41.5227 ± 12.94	0.005	0.3761 **	0.0022
EPI	0.3132 ± 0.05	0.2487 ± 0.06	0.0001	0.5973 ***	0.0001	0.2721 ± 0.06	0.0008	0.4491 **	0.0003	0.2053 ± 0.08	0.0001	0.7445 ***	0.0001
PIV	22.3682 ± 5.65	29.7001 ± 6.3	0.0003	0.4894 **	0.0001	27.7261 ± 4.56	0.0011	0.436 **	0.0004	31.1373 ± 6.6	0.0207	0.3096 **	0.01
Dermis
LEP ratio	0.5957 ± 0.11	0.4622 ± 0.13	0.0001	0.5984 ***	0.0001	0.5639 ± 0.12	0.0734	0.2398 *		0.482 ± 0.13	0.0001	0.5537 ***	0.0001
MEP ratio	0.2438 ± 0.09	0.3453 ± 0.1	0.0001	0.5777 ***	0.0001	0.271 ± 0.1	0.089	0.2278 *		0.3304 ± 0.11	0.0001	0.5526 ***	0.0001
HEP ratio	0.0008 ± 0.01	0.0064 ± 0.01	0.0001	0.514 ***	0.0001	0.0016 ± 0.01	0.1484	0.1957 *		0.0046 ± 0.01	0.005	0.3761 **	0.0022
MPI	40.5087 ± 10.02	51.7453 ± 13.39	0.0001	0.5973 ***	0.0001	42.2963 ± 11.7	0.0558	0.2562 *		50.6193 ± 13.91	0.0001	0.5635 ***	0.0001
EPI	0.7542 ± 0.05	0.7352 ± 0.05	0.0507	0.2616 *		0.7035 ± 0.05	0.0002	0.4981 **	0.0001	0.7443 ± 0.05	0.2552	0.1526 *
PIV	35.3741 ± 6.72	42.9911 ± 6.71	0.0001	0.581 ***	0.0001	38.0342 ± 6.06	0.086	0.23 *		42.7968 ± 6.08	0.0001	0.5407 ***	0.0001

. The thickness of the entry echo layer and dermis were significantly lower at week 8 of treatment than before (p = 0.0143 and p = 0.0001), while there was no significant difference between the thickness before and during the 12-week follow-up (see Figure 1). The thickness of SLEB significantly decreased in the following weeks compared to the pre-therapy results, reaching its lowest value after 12 weeks (p < 0.0001).

The variation in thickness of the entry echo layer was significantly lower at the 4-, 8-, and 12-week follow-up than before therapy (p = 0.0004, p < 0.0001, and p = 0.0014). As for the parameters describing the smoothness of the surface, ACR was significantly higher at 8 and 12 weeks compared to baseline (p = 0.0015 and p = 0.0214), and the surface roughness was significantly lower at the 8-week follow-up (p = 0.0038).

In the entry echo layer, both the LEP and MEP ratios were significantly lower (p < 0.0001 and p < 0.001) and HEP was higher (p < 0.0001) in each follow-up compared to before therapy (see Table 3 for more details). In the case of SLEB, the LEP ratio was significantly lower before therapy and during each follow-up (p = 0.0002 and p < 0.0001), while the MEP ratio was higher before therapy and during each follow-up (p ≤ 0.0001). There were no significant differences recorded for the HEP ratio. In dermis, the LEP ratio was significantly lower at 4 weeks and 12 weeks (p < 0.0001) compared to at the baseline. Both the MEP and HEP ratios were significantly higher at 4 weeks (p ≤ 0.0001) and 12 weeks (p < 0.0001 and p = 0.0022).

A significant increase in mean pixel intensity across all layers was observed at 4 (p < 0.0001) and 12 weeks (p < 0.0001 for entry echo and dermis, p = 0.0022 for SLEB) compared to baseline. For an 8-week comparison, statistically significant improvement was noted only for the entry echo layer (p < 0.0001) and SLEB (p = 0.0003).

In the case of the entry echo layer, pixel intensity variability decreased significantly during the following weeks of therapy (p = 0.0012, p = 0.04232, and p = 0.0006). While the ratio of LEP and MEP decreased and HEP increased, the layer became brighter and more uniform. There are some significant differences in GLCM contrast (p < 0.001) and GLCM correlation (p < 0.05), which were higher in the following weeks versus before therapy.

For SLEB, the entropy of pixel intensities decreased (p < 0.001), and the PIV increased (p ≤ 0.01) compared to the pre-treatment state. GLCM contrast increased (p < 0.001), and two of the GLCM correlation coefficients decreased significantly (p < 0.05) in the following weeks compared to the baseline. The layer became more even and brighter, with a dominance of medium-intensity pixels.

For the dermis layer, there was a significant increase in pixel intensity variability (p ≤ 0.0001) and GLCM contrast (p < 0.001) at the 4- and 12-week follow-up compared to the baseline. GLCM homogeneity decreased significantly in 4 and 12 weeks (p < 0.01), and GLCM energy coefficients decreased significantly between 12 weeks of observation and before treatment (p < 0.01). This indicates the appearance of a varied texture in the dermis layer, with an increased proportion of medium- and high-intensity pixels. Details of all calculated features are summarized in Appendix A. The determined GLCM features for the entry echo layer are summarised in

Table A1. GLCM features describing the texture of the entry echo layer before therapy and at weeks 4, 8 and 12 of treatment. Values of p < 0.05 are marked in bold. Small effect size is marked as *, moderate as ** and large as ***.

GLCM	Week 0	Week 4	p 0–4 Weeks	Effect Size	One-Way p	Week 8	p 0–8 Weeks	Effect Size	One-Way p	Week 12	p 0–12 Weeks	Effect Size	One-Way p
Contr. 1	106.5664 ± 31.8	121.4188 ± 25.42	0.0721	0.2409 *		125.8546 ± 30.04	0.0623	0.2496 *		122.6773 ± 24.68	0.1643	0.1864 *
Contr. 2	307.6188 ± 146.20	476.6654 ± 162.24	0.0009	0.4458 **	0.0003	495.5858 ± 164.01	0.0002	0.4927 **	<0.0001	499.4119 ± 156.8	0.0005	0.4676 **	0.0002
Contr. 3	252.9212 ± 117.49	428.2498 ± 165.32	0.0007	0.4545 **	0.0002	484.8061 ± 151.88	<0.0001	0.5712 ***	<0.0001	432.2026 ± 120.73	<0.0001	0.5439 ***	<0.0001
Contr. 4	300.1705 ± 127.87	436.0123 ± 200.93	0.0005	0.4687 **	0.0002	509.565 ± 186.46	<0.0001	0.5832 ***	<0.0001	470.5519 ± 139.63	0.0001	0.5189 ***	<0.0001
Corr. 1	0.8275 ± 0.05	0.8694 ± 0.04	0.0004	0.4742 **	0.0001	0.8742 ± 0.04	0.0003	0.4818 **	0.0001	0.885 ± 0.04	0.0048	0.3772 **	0.0021
Corr. 2	0.5304 ± 0.06	0.5989 ± 0.06	0.0246	0.3009 **	0.0119	0.5563 ± 0.07	0.2451	0.1559 *		0.5997 ± 0.06	0.0323	0.2867 *	0.0158
Corr. 3	0.6188 ± 0.05	0.6273 ± 0.06	0.5543	0.0796 *		0.5924 ± 0.05	0.1202	0.2082 *		0.6242 ± 0.04	0.6804	0.0556 *
Corr. 4	0.6065 ± 0.09	0.5722 ± 0.08	0.6274	0.0654 *		0.5236 ± 0.07	0.0953	0.2235 *		0.5811 ± 0.06	0.8098	0.0327 *
Energy 1	0.5521 ± 0.11	0.5173 ± 0.10	0.1108	0.2136 *		0.4461 ± 0.12	0.0156	0.3237 **	0.0074	0.481 ± 0.07	0.0479	0.2649 *	0.0237
Energy 2	0.4751 ± 0.14	0.4166 ± 0.13	0.1072	0.2158 *		0.3461 ± 0.16	0.0119	0.3368 **	0.0055	0.3816 ± 0.1	0.0816	0.2333 *
Energy 3	0.4834 ± 0.15	0.4198 ± 0.13	0.086	0.23 *		0.3383 ± 0.15	0.0085	0.3521 **	0.0039	0.3893 ± 0.1	0.0537	0.2583 *
Energy 4	0.4852 ± 0.14	0.4161 ± 0.13	0.0937	0.2245 *		0.3248 ± 0.15	0.0053	0.3728 **	0.0024	0.3854 ± 0.09	0.0285	0.2932 *	0.0139
Homo. 1	0.8222 ± 0.04	0.84 ± 0.04	0.538	0.0828 *		0.8239 ± 0.05	0.6625	0.0589 *		0.828 ± 0.03	0.3503	0.1254 *
Homo. 2	0.7344 ± 0.08	0.742 ± 0.07	0.4217	0.1079 *		0.6578 ± 0.1	0.0357	0.2812 *	0.0175	0.7174 ± 0.06	0.4554	0.1003 *
Homo. 3	0.7528 ± 0.08	0.7403 ± 0.07	0.306	0.1373 *		0.6709 ± 0.11	0.0194	0.3128 **	0.0093	0.7178 ± 0.05	0.2871	0.1428 *
Homo. 4	0.7545 ± 0.07	0.7349 ± 0.09	0.2017	0.1711 *		0.6711 ± 0.11	0.0111	0.3401 **	0.0051	0.7109 ± 0.06	0.1744	0.182 *

, for the SLEB layer in

Table A2. GLCM features describing the texture of SLEB before therapy and at weeks 4, 8 and 12 of treatment. Values of p < 0.05 are marked in bold. Small effect size is marked as *, moderate as **.

GLCM	Week 0	Week 4	p 0–4 Weeks	Effect Size	One-Way p	Week 8	p 0–8 Weeks	Effect Size	One-Way p	Week 12	p 0–12 Weeks	Effect Size	One-Way p
Contr. 1	45.1717 ± 16.03	71.0067 ± 35.1	0.001	0.4393 **	0.0004	67.0024 ± 20.43	0.0045	0.3804 **	0.002	86.7034 ± 47.24	0.0014	0.4273 **	0.0006
Contr. 2	60.7874 ± 29.33	103.3784 ± 59.6	0.0009	0.4426 **	0.0004	99.8159 ± 39.14	0.0018	0.4175 **	0.0007	121.2807 ± 81.58	0.0028	0.4 **	0.0012
Contr. 3	61.0687 ± 24.64	99.9934 ± 62.47	0.0004	0.4774 **	0.0001	93.4511 ± 41.82	0.0008	0.448 **	0.0003	124.7702 ± 76.91	0.0009	0.4436 **	0.0003
Contr. 4	67.3266 ± 26.37	107.7238 ± 65.68	0.0009	0.4436 **	0.0003	102.0267 ± 45.02	0.001	0.4393 **	0.0004	136.0699 ± 73.51	0.0007	0.4556 **	0.0002
Corr. 1	0.5398 ± 0.05	0.5619 ± 0.04	0.4804	0.0948 *		0.5682 ± 0.04	0.1476	0.194 *		0.565 ± 0.06	0.9577	0.0076 *
Corr. 2	0.3428 ± 0.07	0.3373 ± 0.09	0.4855	0.0937 *		0.3373 ± 0.06	0.5931	0.0719 *		0.3579 ± 0.1	0.2227	0.1635 *
Corr. 3	0.4007 ± 0.07	0.3678 ± 0.1	0.0418	0.2725 *	0.0206	0.3821 ± 0.08	0.0905	0.2267 *		0.3569 ± 0.11	0.0139	0.3292 **	0.0066
Corr. 4	0.3883 ± 0.07	0.327 ± 0.1	0.0479	0.2649 *	0.0237	0.3208 ± 0.07	0.047	0.266 *	0.0232	0.3252 ± 0.09	0.0036	0.3891 **	0.0016
Energy 1	0.3113 ± 0.13	0.3476 ± 0.13	0.8225	0.0305 *		0.3025 ± 0.11	0.9642	0.0065 *		0.3098 ± 0.15	0.9317	0.012 *
Energy 2	0.2386 ± 0.13	0.2671 ± 0.14	0.9253	0.0131 *		0.2403 ± 0.11	0.6804	0.0556 *		0.2434 ± 0.16	0.6332	0.0643 *
Energy 3	0.247 ± 0.13	0.2736 ± 0.14	0.88	0.0207 *		0.242 ± 0.11	0.6217	0.0665 *		0.2463 ± 0.16	0.5062	0.0894 *
Energy 4	0.2498 ± 0.13	0.2749 ± 0.14	0.8035	0.0338 *		0.2403 ± 0.11	0.5488	0.0807 *		0.2444 ± 0.15	0.385	0.1166 *
Homo. 1	0.7005 ± 0.09	0.6929 ± 0.08	0.4604	0.0992 *		0.6868 ± 0.07	0.4312	0.1057 *		0.6874 ± 0.1	0.083	0.2322 *
Homo. 2	0.6363 ± 0.10	0.6238 ± 0.1	0.3216	0.133 *		0.6205 ± 0.08	0.2726	0.1472 *		0.6065 ± 0.12	0.0418	0.2725 *	0.0206
Homo. 3	0.6603 ± 0.09	0.6407 ± 0.1	0.2258	0.1624 *		0.6394 ± 0.08	0.2165	0.1657 *		0.6224 ± 0.12	0.019	0.3139 **	0.0091
Homo. 4	0.6439 ± 0.1	0.6348 ± 0.1	0.2258	0.1624 *		0.6233 ± 0.08	0.2017	0.1711 *		0.6074 ± 0.12	0.0199	0.3118 **	0.0095

and for the dermis in

Table A3. GLCM features describing the texture of dermis layer before therapy and at weeks 4, 8 and 12 of treatment. Values of p < 0.05 are marked in bold. Small effect size is marked as *, moderate as ** and large as ***.

GLCM	Week 0	Week 4	p 0–4 Weeks	Effect Size	One-Way p	Week 8	p 0–8 Weeks	Effect Size	One-Way p	Week 12	p 0–12 Weeks	Effect Size	One-Way p
Contr. 1	78.5713 ± 24.07	112.9283 ± 34.53	0.0003	0.4807 **	0.0001	86.2491 ± 22.28	0.0601	0.2518 *		107.6226 ± 26.38	0.0001	0.5243 ***	<0.0001
Contr. 2	94.3037 ± 31.44	137.602 ± 41.77	0.0006	0.4567 **	0.0002	109.845 ± 32.72	0.0747	0.2387 *		132.783 ± 34.58	0.0003	0.4818 **	0.0001
Contr. 3	78.4161 ± 25.86	107.8711 ± 36.84	0.0036	0.3891 **	0.0016	88.7475 ± 27.92	0.1108	0.2136 *		108.7588 ± 30.2	0.0011	0.4382 **	0.0004
Contr. 4	93.71 ± 29.95	132.7353 ± 44.95	0.0004	0.4774 **	0.0001	106.4666 ± 29.61	0.0537	0.2583 *		130.9523 ± 34.96	0.0001	0.5145 ***	<0.0001
Corr. 1	0.5657 ± 0.03	0.5775 ± 0.04	0.1409	0.1973 *		0.565 ± 0.05	0.6274	0.0654 *		0.5704 ± 0.04	0.2834	0.1439 *
Corr. 2	0.4679 ± 0.04	0.5025 ± 0.05	0.0547	0.2572 *		0.4595 ± 0.05	0.4456	0.1025 *		0.4705 ± 0.06	0.262	0.1504 *
Corr. 3	0.5682 ± 0.04	0.5946 ± 0.04	0.0055	0.3717 **	0.0024	0.5732 ± 0.05	0.3256	0.1319 *		0.5812 ± 0.04	0.0139	0.3292 **	0.0066
Corr. 4	0.4927 ± 0.04	0.492 ± 0.05	0.3985	0.1134 *		0.4608 ± 0.05	0.1454	0.1951 *		0.4801 ± 0.06	0.5273	0.085 *
Energy 1	0.0988 ± 0.05	0.081 ± 0.04	0.0788	0.2354 *		0.0868 ± 0.05	0.1476	0.194 *		0.0683 ± 0.03	0.0026	0.4022 **	0.0011
Energy 2	0.0822 ± 0.05	0.0623 ± 0.04	0.076	0.2376 *		0.0676 ± 0.04	0.1202	0.2082 *		0.0506 ± 0.02	0.0031	0.3957 **	0.0013
Energy 3	0.0853 ± 0.05	0.0649 ± 0.04	0.0537	0.2583 *		0.0717 ± 0.05	0.2135	0.1668 *		0.0515 ± 0.03	0.0027	0.4011 **	0.0011
Energy 4	0.0801 ± 0.05	0.0643 ± 0.04	0.0547	0.2572 *		0.066 ± 0.04	0.0905	0.2267 *		0.0499 ± 0.03	0.0025	0.4044 **	0.001
Homo. 1	0.5203 ± 0.08	0.4772 ± 0.07	0.0045	0.3804 **	0.002	0.4947 ± 0.07	0.0845	0.2311 *		0.452 ± 0.04	0.0004	0.472 **	0.0001
Homo. 2	0.4869 ± 0.08	0.4374 ± 0.08	0.0035	0.3913 **	0.0015	0.446 ± 0.08	0.0418	0.2725 *	0.0206	0.4074 ± 0.05	0.0005	0.4644 **	0.0002
Homo. 3	0.5063 ± 0.08	0.4549 ± 0.07	0.0039	0.3859 **	0.0017	0.4684 ± 0.08	0.1643	0.1864 *		0.4246 ± 0.05	0.001	0.4404 **	0.0004
Homo. 4	0.4923 ± 0.08	0.4412 ± 0.08	0.0019	0.4164 **	0.0008	0.4501 ± 0.08	0.0336	0.2845 *	0.0164	0.406 ± 0.05	0.0003	0.4785 **	0.0001

. The characteristics for which significant statistical differences were shown between the control group and week 12 are summarised in

Table A4. Features that are statistically different (p < 0.05) for seemingly healthy and 12 weeks skin. Small effect size is marked as *, moderate as ** and large as ***.

	Healthy	Week 12	p Two-Way	Effect Size	p One-Way
Entry echo
LEP ratio	0.0458 ± 0.03	0.011 ± 0.01	0.0001	0.5388 ***	0.0001
MEP ratio	0.4396 ± 0.08	0.2808 ± 0.07	0.0001	0.5191 ***	0.0001
HEP ratio	0.3649 ± 0.12	0.6325 ± 0.1	0.0001	0.5841 ***	0.0001
EPI	0.1185 ± 0.02	0.1817 ± 0.02	0.0001	0.8321 ***	0.0001
MPI	151.2504 ± 21.22	199.0126 ± 13.97	0.0001	0.5465 ***	0.0001
PIV	73.5075 ± 3.51	68.1178 ± 3.71	0.0016	0.3309 **	0.0008
Contr. 2	702.1389 ± 296.76	499.4119 ± 156.8	0.0364	0.2198 *	0.0182
Contr. 3	641.9425 ± 289.07	432.2026 ± 120.73	0.0051	0.2942 *	0.0025
Contr. 4	628.9161 ± 300.47	470.5519 ± 139.63	0.0116	0.2651 *	0.0058
Corr. 1	0.853 ± 0.03	0.885 ± 0.04	0.0194	0.2454 *	0.0097
Corr. 2	0.1909 ± 0.11	0.5997 ± 0.06	0.0001	0.7953 ***	0.0001
Corr. 3	0.1748 ± 0.09	0.6242 ± 0.04	0.0001	0.8329 ***	0.0001
Corr. 4	0.1084 ± 0.1	0.5811 ± 0.06	0.0001	0.785 ***	0.0001
Homo. 1	0.7987 ± 0.08	0.828 ± 0.03	0.0085	0.2762 *	0.0043
Homo. 2	0.6046 ± 0.15	0.7174 ± 0.06	0.002	0.325 **	0.001
Homo. 3	0.6071 ± 0.15	0.7178 ± 0.05	0.0007	0.3566 **	0.0003
Homo. 4	0.5923 ± 0.15	0.7109 ± 0.06	0.0007	0.3566 **	0.0003
Thickness	0.1141 ± 0.02	0.1951 ± 0.02	0.0001	0.8338 ***	0.0001
TVI	0.0215 ± 0.00	0.0319 ± 0.01	0.0001	0.5268 ***	0.0001
PAR	0.2638 ± 0.04	0.1637 ± 0.02	0.0001	0.7996 ***	0.0001
SR	6.0213 ± 2.82	9.8148 ± 3.63	0.0004	0.3737 **	0.0002
SLEB
MEP ratio	0.1242 ± 0.12	0.258 ± 0.13	0.0017	0.3286 **	0.0009
MPI	28.4545 ± 19.26	41.5227 ± 12.94	0.0009	0.3483 **	0.0005
PIV	22.1524 ± 14.17	31.1373 ± 6.6	0.0012	0.3406 **	0.0006
Contr. 1	49.0758 ± 40.63	86.7034 ± 47.24	0.0015	0.3325 **	0.0008
Contr. 2	71.5932 ± 60.81	121.2807 ± 81.58	0.0038	0.3041 **	0.0019
Contr. 3	69.86 ± 61.81	124.7702 ± 76.91	0.0046	0.2972 *	0.0023
Contr. 4	67.4893 ± 65.52	136.0699 ± 73.51	0.0039	0.3033 **	0.0019
Corr. 1	0.4908 ± 0.27	0.565 ± 0.06	0.0004	0.3703 **	0.0002
Corr. 2	0.2204 ± 0.17	0.3579 ± 0.1	0.0023	0.3196 **	0.0012
Corr. 3	0.2332 ± 0.16	0.3569 ± 0.11	0.0005	0.3634 **	0.0003
Corr. 4	0.21 ± 0.13	0.3252 ± 0.09	0.0003	0.384 **	0.0001
Energy 3	0.1268 ± 0.18	0.2463 ± 0.16	0.0385	0.2173 *	0.0193
Energy 4	0.1191 ± 0.17	0.2444 ± 0.15	0.037	0.2191 *	0.0185
Homo. 2	0.4599 ± 0.32	0.6065 ± 0.12	0.0409	0.2148 *	0.0204
Homo. 3	0.4723 ± 0.32	0.6224 ± 0.12	0.0235	0.238 *	0.0117
Homo. 4	0.4553 ± 0.31	0.6074 ± 0.12	0.025	0.2354 *	0.0125
Thickness	0.1931 ± 0.16	0.2325 ± 0.11	0.0317	0.2256 *	0.0158
Dermis
EPI	0.5782 ± 0.06	0.7443 ± 0.05	0.0001	0.6499 ***	0.0001
Contr. 1	77.9284 ± 18.53	107.6226 ± 26.38	0.0003	0.3771 **	0.0002
Corr. 1	0.6438 ± 0.06	0.5704 ± 0.04	0.0049	0.295 *	0.0025
Corr. 3	0.5045 ± 0.07	0.5812 ± 0.04	0.0001	0.4695 **	0.0001
Thickness	0.9761 ± 0.15	1.5124 ± 0.18	0.0001	0.6499 ***	0.0001

.

3.2. Comparison of Skin Results at 12-Week Follow-Up with Healthy Skin

In recordings of seemingly healthy skin, the SLEB layer was present in 71% (in 25 out of 35) markers, and in the 12-week follow-up, this layer was still visible in 88% (49 out of 56) cases. In the entry echo layer, the MEP and LEP ratios were significantly lower, and the HEP ratio and mean pixel intensity were higher at the 12-week follow-up (p < 0.0001) in comparison to healthy skin.

In SLEB, the MEP ratio and mean pixel intensity were significantly higher at the 12-week follow-up (p = 0.0009 and p = 0.0005), but no differences for LEP and HEP were recorded. Divergence in pixel intensity, measured with the standard deviation of intensity and GLCM coefficients, was significantly higher at the 12-week follow-up (p = 0.0006 and p < 0.05). Pixel entropy in the dermis layer was significantly higher at week 12 (p < 0.0001). For parameters describing echogenicity and most concerning skin texture, however, there were no statistical differences between seemingly healthy skin and that at the 12-week follow-up. For more details, see Appendix A.

3.3. AKASI Before and After Therapy

The AKASI score before and during the 12 weeks of observation was determined for the 39 patients who underwent follow-up at week 12. The median AKASI score was 3.2 (1.2–8.6) before treatment and 0.6 (0–2.8). The median difference before and after treatment was equal to 2.8. The difference was statistically significant with a p-value < 0.0001 and effect size = 0.8705, i.e., large. The complete resolution of symptoms, i.e., an AKASI score of 0, was achieved in 49% (19 of 39) of patients. Clinical images of a patient before and during PDT treatment can be seen in Figure 2.

4. Discussion

According to the literature data, this is the first study describing the ALA-PDT treatment efficacy in different AK severities evaluated in the HFUS. In order to observe the skin changes occurring in the AK during treatment, for the first time, the authors performed follow-up visits and assessed objective ultrasound parameters such as SLEB thickness, echogenicity, and the thickness of individual layers as well as used new variables describing changes in the skin during PDT therapy, such as surface roughness, mean pixel intensity, and parameters describing skin texture. Furthermore, the results were compared with clinically unaffected skin.

The baseline image of AK in HFUS usually presents as decreased echogenicity underneath the entry echo, possibly visualized perpendicular to the entry echo shadows due to the presence of keratin on the surface. A linear SLEB might be detected. It may be in line with tumor formation, or corresponding inflammation [13]. Histopathologically, AKs can be differentiated from field cancerization by hyperkeratosis along a concomitant parakeratosis and abnormal keratinization with a lymphocytic infiltrate, which is suspected to affect the progression of the disease [38]. Furthermore, SLEB might also represent elastosis, which histopathologically correlates with the accumulation of abnormal elastotic fibers in the upper and middle dermis [39,40]. Patients with AKs usually present with typical features for sun-exposed skin; thus, even in skin without clinically evident changes, SLEB might be visible.

The PDT treatment monitoring of AK with the use of HFUS has already been reported in one study [41]. Arisi et al. evaluated MAL-PDT efficacy in a group of 26 patients with AKII with a 50 MHz transducer and reported that MAL-PDT increased the dermal density and reduced the SLEB in the treatment of targeted and perilesional skin. This is concomitant with our observations, as the SLEB thickness decreased within the control visits at 4, 8, and 12 weeks (p < 0.0001) within lesional skin, although it did not disappear completely and was still present after 12 weeks, but to a lower extent. This might correspond to the elastosis. However, we cannot unequivocally rule out the presence of infiltrate, and further study in this area is necessary (including histopathological confirmation as well as the detailed monitoring of patients in this area) because this may be the group of patients in whom AK recurrence will be more likely than in those in whom SLEB has completely subsided.

In our results, the skin roughness in HFUS scans has also improved at the 12-week interval, which was already described clinically by Szeimies et al. [42] and Reinhold et al. [26]. The quality of the skin after BF-200 PDT was enhanced substantially. The authors reported that the number of patients without skin roughness, dryness, or scaling improved from 15 to 63% after the procedure [43]. PDT leads to improvement with regard to texture, wrinkling, skin coloration, and reduction in telangiectasia [43]. In our study, the entry echo layer became smoother in subsequent months of therapy (p < 0.05). For parameters referring to the roughness of the skin surface, significant improvement was noted at weeks 8 and 12 (p < 0.05).

Apart from SLEB, other skin parameters can be analyzed in HFUS scans, especially for skin texture analyses. Crisan et al. used other parameters to evaluate the efficacy of vitamin C treatment on skin rejuvenation and the changes occurring in the skin during antiaging therapy [31]. They used three consecutive numbers: LEP aligns with the number of low-echogenic pixels, MEP with medium-echogenic pixels, and HEP with high-echogenic pixels in the assessed scans [28,31].

In our study, the number of LEPs decreased significantly at all follow-up visits. This relationship was observed for all skin layers (p < 0.001 for entry echo, SLEB, and dermis at 4 and 12 weeks). In our study, the values were related to the total number of pixels distributed within the layers, due to the fact that their morphological parameters may depend on the measurement site and individual patient characteristics [44]. LEP is associated with quantifying the hydration degree within the skin, collagen degeneration, and elastosis, as well as the inflammation and infiltration of a malignant tumor [28,31]. The biological effect of PDT depends on a reaction of photosensitizer with a specific wavelength, which leads to the occurrence of molecular oxygen, and then a formation of singlet oxygen [45]. This molecule is very active and substantially causes oxidative damage and cell death [45], along with reduction in the histological features of actinic damage, decreased expression of Ki-67 and p53, and reduction in elastin thickness [46,47]. MEP and HEP quantify the levels of collagen, elastin, and microfibrils [28,31]. The MEP count increased in SLEB at the 4-, 8-, 12-week follow-ups, which was expected due to the reduction in the SLEB layer at follow-up. PDT treatments increase the collagen I levels, which is one of the causes of the visible clinically reversed aging [48,49], with a decrease in the elastic fibers [43]. The levels of matrix metalloproteinase-3 are also enhanced leading to the degradation and removal of old collagen fibers [48,49]. A modification within the skin proteins and the activation of skin fibroblasts can be observed [43,49], which might be visible in the MEP and HEP increase in the dermis throughout the 4-, 8-, 12-week intervals in our patients. Since there is a quantitative change in the skin texture, PDT remains a cosmetically acceptable treatment modality that does not lead to skin atrophy or scarring compared to cryotherapy [50].

Our study additionally compared skin at 12-week follow-up with scans obtained within clinically unchanged skin (without obvious, visible changes associated with sun damage) in 32 treated patients. SLEB was present in the majority of cases in both groups, but there was no statistically significant difference in the thickness of this layer. Echogenicity in this layer was comparable (regarding the LEP and HEP ratios) or better (regarding the MEP ratio and mean pixel intensity; p < 0.001) at week 12. At week 12, the contrast within the layer was also higher than for unchanged healthy skin (p < 0.05). The echogenicity of the dermis was comparable in both groups. A significant limitation to the collection of clinically unaffected skin samples in this patient group is the nature of the disease itself with the presence of field cancerization, meaning that subclinical changes might occur.

In this study, for the first time, we analyzed new parameters in a post-therapeutic assessment of PDT. Homogeneity corresponds to the uniformity of pixels, energy, describing how texture forms into visible patterns, and the EPI, which quantifies the level of randomness within pixels and correlation and shows the improvement in skin quality after therapy, corresponding to how pixels are related in the skin layers. Variations in pixel intensities are described by such parameters as contrast and PIV, with high values indicating a presence of distinctive patterns. These imply improvement in skin texture and clinically visible enhancement of skin quality.

Our study, similarly to previous ones, shows that PDT is a highly effective form of treatment that leads to a reduction in AKs along with an enhancement of skin quality [48,49]. This is visible in the significant drop in the AKASI score in our group, which objectively allows us to assess the treatment outcome [51]. We think that HFUS can be a valuable non-invasive modality in monitoring treatment efficacy after PDT. The evolution of skin layers seen in the aforementioned parameters enables us to see whether the therapy was successful or if the patients require an additional procedure.

Regardless, fast, non-invasive treatment monitoring might be useful in clinical practice; however, studies focused on selected AK grades II and III might be required. HFUS devices might not be as available as dermatoscopes; nevertheless, they allow us to visualize deeper layers of the skin, which is quite important regarding treatment monitoring as they may enable the examination of subclinical lesions and indicate which patients may require monitoring, because the risk of relapse may be higher.

5. Conclusions

Our study confirms that HFUS can be helpful in monitoring the effects of PDT and can complement the clinical-dermatoscopic assessment. We believe that HFUS might not only show the possible effect of PDT in decrease of SLEB but also detect subclinical lesions and allows us to analyze deeper layers of the skin.