Computational modeling and histologic analysis of 6.78- and 2-MHz monopolar radiofrequency-induced thermal reactions

Abstract

Monopolar radiofrequency (RF) devices are widely used for skin tightening, wrinkle reduction, and body contouring. However, frequency-dependent differences in energy absorption and tissue remodeling remain insufficiently characterized. This study aimed to compare the thermal distribution and histological responses induced by 6.78‑ and 2‑MHz monopolar RF, as well as their sequential combination. Finite-element computational modeling was used to simulate electric field propagation and heat diffusion in multilayered skin with varying subcutaneous fat thicknesses and fibrous septa configurations. In vivo experiments were conducted on porcine skin treated with 6.78‑, 2‑, and dual‑frequency RF modes. Histologic changes were evaluated using hematoxylin and eosin, Masson’s trichrome, and Verhoeff-van Gieson staining. Apoptotic cell death was assessed using the terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay to evaluate adipocyte viability. Computational modeling demonstrated that 2‑MHz RF produced broader and deeper thermal effects within the adipose tissue, whereas 6.78‑MHz RF generated more localized heating along the fibrous septa. Dual-frequency RF combines these effects, creating pronounced thermal reactions at the dermosubcutaneous junction. Histological analysis revealed significant collagen and elastin remodeling across the dermis and fibrous septa in dual-frequency-treated specimens, with no evidence of adipocyte apoptosis. Moreover, remodeling changes were more extensive and persistent at later time points, suggesting that dual-frequency treatments have a greater tissue remodeling potential compared with single-frequency applications. Dual-frequency monopolar RF effectively promoted extracellular matrix remodeling in the dermis and subcutis while preserving adipocyte viability, suggesting its use as a safe and versatile modality for skin rejuvenation and contouring.

Supplementary Information

The online version contains supplementary material available at 10.1007/s10103-025-04746-8.

Introduction

Non-invasive monopolar radiofrequency (RF) devices have been used to treat facial wrinkles, scars, skin laxity, facial skin tightening, and body contouring [1–5]. Using a monopolar mode, alternating current and medium- or high-frequency RF currents flow from the active electrodes toward the grounded electrode and generate thermal tissue reactions in the entire dermis, fibrous septa of the subcutaneous fat, and fibromuscular tissues [6, 7]. Therein, the amount of energy absorbed mainly depends on the RF frequency, power, and conduction time, and frequency is regarded as the most important parameter [8]. However, the precise and predictable regulation of energy absorption across the layered skin and subdermal structures with abundant appendages requires further investigation.

Radiofrequency-induced electrical responses and thermal tissue reactions can be investigated through in vivo and ex vivo studies in humans and animals, as well as through phantom tissue models and computational simulations [8–10]. A previous study developed computational modeling for simulating RF-induced voltage distributions, electric fields, electric power absorption, and thermal reactions in the skin and subdermis, which are composed of three layers: skin, subcutaneous fat, and muscle [8]. Therein, the greater electric power density of 1-MHz monopolar RF was found to be absorbed by the fibrous septa of subcutaneous fat compared with the adipocytes in the subcutaneous fat; accordingly, the tissue temperature was markedly elevated along the fibrous septa of subcutaneous fat [8]. Nonetheless, compared with in vivo human and animal experiments, computational modeling studies can only reflect immediate tissue reactions, and the effects of various skin appendages on the dermis cannot be simulated.

In the present study, we compared the characteristics of energy absorption and distribution according to the RF frequencies, including 6.78 MHz, 2 MHz, and combined 6.78 MHz and 2 MHz, and the patterns of operating pulses using the computational modeling. Skin and subdermal structures with or without fibrous septa of subcutaneous fat were developed to simulate RF-induced thermal reactions, and various thickness settings of subcutaneous fat were also used. Additionally, a histological study was performed using an in vivo minipig model after monopolar RF treatment. The immediate and late RF-induced thermal reactions in the collagen, elastic fibers, and adipocytes of the dermis, subcutaneous fat, and fibrous septa of the subcutaneous fat were evaluated at different operating frequencies and pulse patterns.

Materials and methods

Non-invasive monopolar RF device and computational modeling

A non-invasive monopolar RF device (XERF; Cynosure Lutronic Inc., Goyang, Korea) was used in this study, and three different patterns of RF pulses were used for experimental purposes; 1) type-S RF pulse consisting of 6.78-MHz, 12 square-patterned 100-ms identical sub-pulses over 1930 ms, 2) type-M RF pulse consisting of 6.78-MHz, 6 square-patterned 200-ms identical sub-pulses over 1950 ms, and 3) type-D RF pulse consisting of 6.78-MHz, 5 square-patterned 150-ms identical sub-pulses and 2-MHz, 7 square-patterned 120-ms identical sub-pulses over 2400 ms (Fig. S1). In each experimental setting, RF treatments were administered incorporating parallel-contact gas-cooling bursts.

Electro-thermal interactions of RF energy in multilayered skin tissues were simulated using a COMSOL Multiphysics® (COMSOL Inc., Burlington, MA, USA) by defining electromagnetic and thermal properties as described in a previous study (Table 1) [8]. Briefly, the model geometry simulated a multilayered skin structure composed of the epidermis (0.08-mm thickness), dermis (2.0-mm thickness), and subcutaneous fat layers of 5-, 10-, or 15-mm thickness, with or without fibrous septa within the fat layer [8, 11]. The non-invasive electrode of the monopolar RF device was positioned at the skin surface, and a dispersive return electrode was located beneath the subcutaneous fat layer. The frequency-dependent dielectric permittivity and electrical conductivity were used to accurately represent the RF behavior within biological tissues. In subcutaneous fat, the loss tangent was identified as the dominant factor influencing RF energy absorption owing to the low dielectric constant and minimal water content of the fat tissue. Conversely, conductivity played a primary role in energy deposition within the dermis, which is consistent with its collagen-rich hydrated composition. Time-dependent finite element analysis was conducted to predict the spatial and temporal distributions of the electric fields and temperature across varying frequency and depth mode conditions [8]. The model coupled three physical interfaces: electric current, bioheat transfer, and thermal damage [8].

Numerical modeling

Electric field propagation was modeled using Maxwell’s quasi-static approximation, and heat transfer in biological tissues was described using Pennes’ bioheat transfer equation. Thermal injury dynamics were incorporated through Arrhenius-based kinetics to estimate tissue damage accumulation. All simulations assumed isotropic tissue properties and did not account for mechanical deformation effects. The overall computational approach followed previously established methods for high-fidelity radiofrequency ablation modeling [8, 12, 13].

In vivo minipig study using non-invasive monopolar RF device

Three 8-week-old female minipigs (Sus scrofa domestica) weighing 24–28 kg were used for in vivo experiments. Ethical approval was obtained from the relevant Institutional Animal Care and Use Committee. All animal experiments were performed in accordance with the National Institutes of Health Guidelines for the Care and Use of Laboratory Animals and Institutional ethical standards. General anesthesia was administered via an intramuscular bolus injection of tiletamine/zolazepam (5 mg/kg) and xylazine (2 mg/kg). Subsequently, we performed endotracheal intubation and connected the minipig to a ventilator. The lungs were ventilated with oxygen, and anesthesia was maintained with 2% isoflurane. Intravenous hydration with normal saline was maintained through the superficial auricular vein (25 ml/h). After the gentle removal of the hair on the abdomen, the lesions were cleansed with a mild soap and 70% alcohol. The skin was marked with black ink to outline the grids for each experimental setting. A non-invasive monopolar RF device (XERF) with a 2.0 cm × 3.0 cm disposable tip (Effector 60; Cynosure Lutronic Inc.) was used to deliver RF energy to the in vivo minipig skin. After applying the conductive fluid to the surface of the skin, each minipig received noninvasive monopolar RF treatments at an energy setting of 240 J using a type-S RF pulse, 360 J using a type-M RF pulse, and 360 J using a type-D RF pulse. RF treatments at each setting were performed in triplicate.

Histologic analysis and terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay

The experimental minipigs were euthanized to humanely sample the treated tissues, according to standard protocols. At baseline and 7, 15, and 30 days after treatment, full-thickness tissue specimens, including the epidermis, dermis, and subcutaneous fat, were obtained for microscopic evaluation. Each Samples were fixed in 10% buffered formalin and embedded in paraffin. Serial tissue sections of 4-μm thickness were prepared for each treatment setting and stained with hematoxylin and eosin (H&E), Masson’s trichrome, and Verhoeff-van Gieson. Apoptotic cell death was evaluated using a TUNEL Assay Kit with HRP-DAB detection (ab206386; Abcam, Cambridge, UK) according to the manufacturer’s instructions. Briefly, deparaffinized tissue sections were incubated with terminal deoxynucleotidyl transferase and biotin-labeled dUTP at 37 ℃ for 1 h in a humidified chamber. Subsequently, HRP-conjugated streptavidin was applied, and signal was developed using a DAB substrate.

Results

Frequency-dependent electro-thermal interactions in multilayered skin tissue

Electro-thermal interaction experiments were conducted using computational modeling of multilayered skin tissue, including the epidermis, dermis, and subcutaneous fat (5.0-mm thickness), excluding subcutaneous fibrous septa. Non-invasive RF treatments at frequencies of 6.78 MHz and 2 MHz were compared under identical total energy delivery conditions. The computational modeling data demonstrated the frequency-dependent differences in thermal distribution across the multilayered skin tissue (Fig. 1). The area of the 2-MHz RF-induced electro-thermal reaction was wider and deeper compared with 6.78-MHz RF treatment. Moreover, the 2-MHz RF treatment produced significantly higher tissue temperatures in the upper subcutaneous fat layer.

Fig. 1.

Simulated thermal distributions at 0.1 s post-RF exposure at different frequencies. (a) 6.78 MHz and (b) 2 MHz simulations were conducted with equivalent total energy input. Compared to 6.78 MHz, the 2 MHz condition resulted in broader and deeper thermal diffusion within the tissue model

Additionally, electro-thermal interaction experiments in multilayered skin tissue, including the epidermis, dermis, and subcutaneous fat with components of subcutaneous fibrous septa (20-mm thickness), were performed. Our computational modeling data presented that 6.78-MHz RF treatment preferentially elevated tissue temperature along the fibrous septa of the subcutaneous fat (Fig. 2). Moreover, perpendicularly arranged fibrous septa exhibited greater thermal reaction with 6.78-MHz RF treatment compared with horizontally arranged fibrous septa. Those electro-thermal reactions in the subcutaneous fibrous septa became apparent from post-RF 0.3 s and increased over time, while those in the subcutaneous fat seemed gradually noticeable at post-RF 0.5 s. Nonetheless, the deeper part of fibrous septa was notably affected by 6.78-MHz RF treatment compared with 2-MHz RF. Meanwhile, thermal tissue reactions by 2-MHz RF treatment developed over wider and deeper areas of the multilayered skin model with subcutaneous fibrous septa, and the temperature was found to be more rapid and highly elevated compared to 6.78-MHz RF. Moreover, the 2-MHz RF treatment preferentially generated remarkable thermal reactions in the upper to middle part of the subcutaneous fat rather than the subcutaneous fibrous septa.

Fig. 2.

Time-resolved thermal propagation patterns in anatomically representative models incorporating fibrous septa. Temperature maps simulated at 0.16 s, 0.26 s, 0.36 s, and 0.50 s for (a) 6.78 MHz and (b) 2 MHz RF exposures. Heat propagation at 6.78 MHz remained localized along fibrous septa, while 2 MHz induced widespread diffusion into adipocyte-rich subcutaneous regions

Effect of RF pulse pattern and tissue thickness on electro-thermal interactions

Electro-thermal interaction experiments were also performed to compare the RF-induced thermal reactions depending on the thickness of the subcutaneous fat layer (5, 10, and 15 mm) and the types of RF pulses (S, M, and D). All experimental models exhibited noticeable RF-induced thermal reactions, which started to appear from the contact surface of the disposable tip and progressed into the subcutaneous fat and fibrous septa (Fig. 3). The degree of tissue temperature elevation was higher in the experimental setting with a thinner subcutaneous fat layer. Moreover, RF treatments using type S and M pulses exhibited the area of highest tissue temperature at the fibrous septa of the subcutaneous fat, whereas type D RF pulse treatments generated the highest tissue temperature at the subcutaneous fat. Additionally, the type-S RF treatment produced a more rapid temperature elevation in the upper portion of the subcutaneous fat layer compared with types M or D. In contrast, the type-M RF treatment resulted in a slower but greater overall temperature rise than type S in the experimental modeling. Finally, the computational modeling of the type-D RF treatment exhibited the highest temperature elevation, with heat distribution extending into the deepest tissue layers, compared to both types S and M.

Fig. 3.

Comparative thermal effects of NMRF depth parameters in subcutaneous tissue. Temperature profiles at 2.275 s, 2.66 s, and 2.67 s following treatment in the (left) type-S RF pulse, (middle) type-M RF pulse, and (right) type-D RF pulse. Segmentation at 5 mm, 10 mm, and 15 mm revealed progressively greater thermal penetration with increasing depth, with the deep mode showing the highest temperature elevation at all levels

Effect of RF frequency and pulse pattern on in vivo minipig skin

Histological examination with H&E and Masson’s trichrome staining revealed marked thickening of the dermis and fibrous septa of the subcutaneous fat on post-RF days 7, 15, and 30 in all experimental settings of pulse patterns, including S-, M-, and D-types. Histological changes were more remarkable at 30 days post-RF than at 7 or 15 days. Moreover, Verhoeff-van Gieson-stained specimens presented increased clumped or fragmented elastic fibers on post-RF day 7, and numerous thick and elongated elastic fibers on post-RF days 15 and 30 in the lower dermis and fibrous septa of the subcutaneous fat compared with the untreated control specimens (Fig. 4). Among the three pulse patterns, D-type RF treatment generated more pronounced and remarkable histological changes in the collagen and elastic fibers that were distributed into deeper parts of the dermis, dermo-subcutaneous junction, and subcutaneous fibrous septa than type M and type S. Nonetheless, the specimens exhibited preserved adipocyte morphology with no evidence of structural disruption at 7, 15, or 30-days post-treatment (Fig. 5a). Additionally, complementary TUNEL assays revealed no remarkable apoptotic cells in the subcutaneous adipocytes or fibrous components (Fig. 5b).

Fig. 4.

Histological evaluation of extracellular matrix remodeling 30 days post-XERF treatment using Verhoeff–Van Gieson (VVG) staining. Tissue sections from baseline, type-S RF pulse, type-M RF pulse, and type-D RF pulse demonstrated progressive remodeling of collagen and elastin fibers, with the most pronounced changes observed in the deep mode near the dermo-subcutaneous junction

Fig. 5.

Longitudinal histological assessment of adipose tissue following NMRF treatment. (a) Hematoxylin and eosin (H&E)-stained subcutaneous tissue at days 7, 15, and 30 in type-D RF pulse showed preserved adipocyte morphology without structural damage. (b) Terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL) assay confirmed the absence of apoptosis in subcutaneous fat, supporting the non-cytotoxic profile of NMRF

Discussion

In this computational modeling study, we comparatively demonstrated the patterns of energy absorption and distribution in the epidermis, dermis, subcutaneous fat, and fibrous septa according to the RF frequencies of 6.78 MHz, 2 MHz, and combined 6.78 MHz and 2 MHz. In the experimental settings of equal energy input, the 2-MHz RF treatment generated broader and deeper thermal reactions compared with the 6.78-MHz RF. These findings result from the equation for predicting the penetration depth of RF energy, which is inversely proportional to the square root of the frequency [14]. Additionally, 2-MHz RF energy was found to volumetrically increase tissue temperature over subcutaneous fat tissue, whereas 6.78-MHz RF energy preferentially generated thermal reactions through the subcutaneous fibrous septa in the current pathways. Our findings are likely attributable to differences in the dielectric properties and loss tangent characteristics between adipocytes and fibrous tissues [15–17]. We also suggested that 6.78-MHz RF efficiently generates thermal reactions primarily through the fibrous septal structures within the subcutaneous tissue. In contrast, 2-MHz RF treatment appears to induce bulk heat accumulation in the adipose tissue, potentially stimulating adipose-derived stem cells directly for skin rejuvenation, while also indirectly transferring thermal energy to the fibrous septal structures to promote skin contouring.

In the present study, we additionally analyzed the energy absorption and distribution according to pulse patterns, including types S, M, and D, using computational modeling. Our computational modeling study showed that RF-induced thermal reactions were generated in broader and deeper tissues in the type-M RF treatment compared to type S, under identical delivery of total energy. Both types were composed of 6.78-MHz RF subpulse packs; however, subpulse packs were composed of 12 square-patterned 100-ms identical subpulses in type S and 6 square-patterned 200-ms identical subpulses in type M. These findings suggest that the number and RF subpulse packs could have affected the generation and propagation of RF-induced thermal reactions.

Our study used the experimental setting of dual-frequency RF treatment (type-D), which consisted of 6.78-MHz, five square-patterned 150-ms identical sub-pulses and 2-MHz, seven square-patterned 120-ms identical sub-pulses. By sequentially delivering dual frequencies, notable bulk thermal reactions were simulated in a computational modeling study that presented both 6.78-MHz RF-induced tissue reactions in the fibrous septa of the subcutaneous fibrous septa and 2-MHz RF-induced reactions in the subcutaneous adipose tissue. Through a combination of time-resolved computational modeling and histological validation, we confirmed that the type S, type M, and type D modes of RF delivery generated distinct energy propagation and thermal reaction patterns. Moreover, type-D RF delivery generated broad thermal reactions across the tissue layers, with pronounced heating observed at the dermosubcutaneous junction. Although 2-MHz RF has been used in thermal lipolysis platforms to induce adipocyte apoptosis [18, 19], our experimental settings of monopolar RF treatment could not demonstrate histological evidence of lethal damage to subcutaneous adipocytes in an in vivo minipig study. These findings suggest that dual-frequency RF treatments can be used to efficiently generate therapeutic thermal reactions in aged or photoaged skin to induce the regeneration of collagen and elastic fibers without undesirable fat volume loss.

Our histological analysis of in vivo minipig skin revealed RF frequency- and mode-dependent patterns of extracellular matrix remodeling. Moreover, Verhoeff-van Gieson staining demonstrated a marked increase in collagen and elastic fibers throughout all layers of the skin and the subcutis across all experimental RF mode settings. In this study, we could not investigate 2‑MHz monopolar RF‑induced in vivo minipig skin reactions without combining 6.78‑MHz RF. Therefore, we could not precisely compare the frequency‑dependent tissue changes between computational modeling and in vivo minipig experiments. Nonetheless, our study demonstrated that the sequential delivery of 6.78‑ and 2‑MHz RF pulses into the minipig skin resulted in more pronounced histologic changes in collagen and elastic fibers compared to 6.78‑MHz RF treatment alone. Therein, the most remarkable histologic changes in the deep reticular dermis, dermo-subcutaneous interface, and subcutaneous fibrous septa were found in the specimens, which were treated with type-D RF mode, compared with types S and M. We suggest that 2-MHz RF-induced bulk heating in the subcutaneous fat layer, which was observed in this computational modeling study, could have resulted in marked regeneration of collagen and elastic fibers. As this study was preclinical, it does not directly provide clinical outcome data. However, based on the observed differences in thermal propagation and histologic remodeling, we proposed potential clinical indications for each RF mode. Specifically, type S RF, which showed shallower and localized effects, may be more suitable for patients with fine wrinkles and thin subcutaneous fat. Type M RF may benefit those with moderate wrinkles and average fat thickness. Type D RF, with the deepest and broadest remodeling, could be ideal for patients with coarse wrinkles or thicker subcutaneous tissue.

Our experimental settings of sequential 6.78- and 2-MHz RF delivery demonstrated that the structural integrity and viability of adipocytes in the subcutaneous fat layer were preserved in the in vivo minipig specimens immediately after treatment and at 7, 15, or 30-days post-treatment. RF treatment at lower frequencies and longer wavelengths resulted in bulk tissue heating and deeper energy penetration [14]. Thereby, RF energy at lower frequencies has been used for fat reduction, skin tightening, and rejuvenation [14, 18]. Nonetheless, in our study, we found that the sequential delivery of 6.78- and 2-MHz RF pulse packs in our experimental energy settings generated remarkable regeneration of collagen and elastic fibers throughout the entire dermis and subcutaneous fat layer, while preserving adipocyte viability. We suggest that dual-frequency monopolar RF is a safe and versatile modality for skin rejuvenation and contouring. However, our study has some limitations: (1) computational modeling cannot precisely reflect RF-induced tissue reactions in human skin; (2) the in vivo minipig model differs from human facial and neck skin in terms of collagen content, elastic fiber density, and skin appendages; (3) lack of molecular analysis; and (4) small sample size, short follow-up period, and restricted RF parameter setting.

Conclusion

In this study, we compared the characteristics of energy absorption and distribution according to RF frequencies and patterns of operating pulses using computational modeling and in vivo minipig studies. A computational modeling study demonstrated that 2‑MHz RF produced broader and deeper thermal effects within adipose tissue, whereas 6.78‑MHz RF generated more localized heating along the fibrous septa. Dual-frequency RF combines these effects, creating pronounced thermal reactions at the dermosubcutaneous junction. Histological analysis revealed significant collagen and elastin remodeling across the dermis and fibrous septa in dual-frequency-treated specimens, with no evidence of adipocyte apoptosis. Nonetheless, further studies are required to clinically validate these results and develop applicable treatment protocols for skin rejuvenation and contouring.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contribution

K, H, and S designed research; K, H performed research; K, H, J, and S analyzed data; H, W, S, and S wrote the paper. W, S, and S provided guidance and advice. All authors reviewed the manuscript.

Funding

The authors received no financial support for the research, authorship, or publication of this article.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval

Not applicable.

Clinical trial number

Not applicable.

Competing interests

The authors declare no competing interests.