AI-Assisted Imaging Tracks Success of Low-Fluence QSRL Laser for Melasma

A low-fluence fractional 694-nm Q-switched ruby laser (QSRL) demonstrated meaningful improvements in melasma with no cases of post-inflammatory hyperpigmentation (PIH), rebound pigmentation, or hypopigmentation during follow-up in a small retrospective study of Asian patients.¹ Investigators also reported that an artificial intelligence (AI)-assisted imaging system detected treatment-related pigment changes that may not be captured by traditional clinical assessment tools.

Background and Study Design

The study addressed 2 ongoing challenges in melasma management: identifying safe laser protocols for patients with darker skin types and developing more objective methods to measure treatment response. Melasma remains one of the most difficult pigmentary disorders to treat because of its chronic, relapsing course and its tendency to recur despite therapy. Although topical agents and strict photoprotection remain the foundation of treatment, laser therapy has become an important option for selected patients. However, conventional high-fluence laser treatments have historically been associated with a greater risk of PIH, particularly in individuals with Fitzpatrick skin types III through VI.²

To evaluate whether a lower-energy fractional approach could improve outcomes while minimizing adverse events, investigators retrospectively reviewed patients treated at a dermatology practice in Taiwan between May 2022 and May 2023. The study included patients with mixed-type melasma that had persisted for more than 10 years. Eligible participants had not received laser therapy or topical bleaching agents during the 3 months before treatment. Patients underwent 4 to 5 sessions of fractional 694-nm QSRL delivered at 3- to 8-week intervals using fluence settings of 0.77 to 0.9 J/cm². Of the 18 eligible patients, 16 completed treatment and were included in the final analysis. All participants were women with a mean age of 49 years. Nearly all had Fitzpatrick skin type IV, while 1 patient had skin type III. All cases were classified as mixed melasma with both epidermal and dermal pigment involvement. 

AI-Assisted Imaging vs. Traditional Scoring

Unlike conventional scoring systems such as the Melasma Area and Severity Index (MASI), investigators used an AI-assisted imaging platform designed to objectively evaluate pigment changes over time. The system analyzed standardized digital photographs using automated alignment technology to ensure the same facial regions were evaluated at each visit. The imaging platform incorporated SURF feature alignment and RANSAC algorithms to precisely register regions of interest on the cheeks. It then quantified pigmentation using luminosity measurements and Gray-Level Co-occurrence Matrix (GLCM) texture analysis, allowing investigators to measure changes in pigment intensity and distribution.

Key Efficacy Outcomes

Pigmentation was categorized into 3 levels corresponding to skin types III, IV, and V within the lesions. Type V represented the darkest and deepest pigmentation, while Types III and IV represented lighter and more superficial pigment patterns. The greatest improvements occurred in the darkest pigment category. 

• Right Facial Side: Type V pigmentation decreased significantly from baseline through the fifth treatment session (P = .036). Combined Type III and Type IV pigmentation also increased significantly on the right side, reflecting a shift toward lighter pigmentation patterns (P = .024).
• Left Facial Side: Although similar trends were observed on the left side of the face and in overall facial analysis, these changes did not consistently reach statistical significance. Investigators suggested that asymmetric ultraviolet exposure may partially explain the difference between facial sides.

Beyond pigment intensity, the AI-based texture analysis identified improvements in pigment fragmentation and skin uniformity that investigators noted would likely be difficult to appreciate through routine visual assessment alone. The authors described this capability as providing a type of "digital biopsy," allowing subtle treatment-induced pigment changes to be tracked objectively over time.

Favorable Safety Profile and Patient Satisfaction

Safety findings were particularly notable given the patient population. Throughout at least 6 months after the final treatment session, investigators observed no rebound pigmentation, expansion of melasma lesions, sensitive skin reactions, hypopigmentation, or PIH. Patient satisfaction was also reported to correlate closely with the quantitative pigment improvements measured by the imaging system.

“By reducing the number of passes, we decreased the cumulative thermal burden, effectively lowering the likelihood of PIH and shortening post-treatment erythema,” the authors noted.

Conclusions and Limitations

The investigators concluded that low-fluence fractional 694-nm QSRL offers a safe and effective treatment option for melasma in Asian patients with Fitzpatrick skin types III through V. They also suggested that AI-assisted imaging may provide a more reproducible and sensitive method for monitoring treatment response than traditional subjective scoring systems. The authors acknowledged several limitations, including the retrospective design, small sample size, and relatively short follow-up period, noting that additional studies will be needed to determine long-term durability and recurrence rates.

References

1. Tetin I, Wang CY, Zevini A, et al. Fractional 694-Nm Ruby Laser Therapy for Melasma: A Clinical Study in East Asian Patients. Journal of Cosmetic Dermatology 25, no. 7 (2026): e71071, doi:10.1111/jocd.71071

2. Hilton S, Heise H, Buhren BA, Schrumpf H, Bölke E, Gerber PA. Treatment of melasma in Caucasian patients using a novel 694-nm Q-switched ruby fractional laser. Eur J Med Res. 2013;18(1):43. Published 2013 Nov 14. doi:10.1186/2047-783X-18-43