Aesthetic Surgery Journal最新文献

Background: It remains unclear how long a plastic surgeon should wait for the optimal vasoconstrictive effect of epinephrine as local anesthetic before incision for aesthetic facial surgery.

Objectives: In this study we investigate the optimal timing for epinephrine-induced vasoconstriction in facelift procedures by measuring cheek skin temperature changes with forward-looking infrared thermography (FLIR).

Methods: A retrospective chart review was conducted on all patients who underwent facelift surgery by J.E.Z. between July 2023 and June 2024. Skin surface temperature was recorded at baseline and at predetermined time points up to 15 minutes following injection of the standardized epinephrine-containing local anesthetic solution. Additional patient data were obtained from electronic medical record review.

Results: Twenty-seven patients were included in the study. The median time for each patient to reach the lowest recorded cheek temperature was 5 minutes postinjection (mean 5.1 ± 2.9 minutes). Injected cheeks exhibited the largest median temperature decrease of 2.3°C at 7 minutes, followed by gradual rewarming to baseline. Uninjected nasal skin warmed steadily throughout the observation period. Of the 27 patients, 15 (55.6%) reached their minimum cheek temperature by 5 minutes, 23 (85.2%) by 7 minutes, and all 27 (100%) by 11 minutes.

Conclusions: These findings demonstrate that the maximal vasoconstrictive effect of epinephrine in facelift surgery occurs approximately 5 to 7 minutes after injection. We recommend that facelift surgeons wait 5 to 7 minutes before initiating incisions and dissection to balance optimal hemostasis with procedural efficiency.

Level of evidence: 4 (therapeutic):

Background: The integration of AI chatbots into plastic surgery websites is now standard, providing asynchronous, real-time engagement for patients. Although promoted as scheduling and medical guidance tools, their contribution to clinical workflow improvement and patient satisfaction remains unclear.

Objectives: The aim of this study was to evaluate the accuracy of AI chatbot performance in clinical triage of plastic surgery patients, focusing on triage accuracy and quality of patient interactions.

Methods: The responses of chatbots on top-ranking plastic surgery websites, identified by search engine optimization (SEO) rankings, were analyzed with standardized clinical scenarios representing emergent, urgent, and elective patient inquiries. Responses were analyzed by the chatbot's triage sensitivity and specificity, classification accuracy, escalation metrics, and content quality. Patient experience was quantified with a chatbot usability questionnaire and a visual analog scale. Subgroup analysis by chatbot platform and thematic analysis was performed to identify tonal patterns in chatbot language.

Results: Performance varied significantly across 60 clinical scenarios, particularly in urgency classification. Emergent classifications were most mislabeled as urgent, with a low sensitivity (20%), negative predictive value (0.71), and high false negative rate (80.0%). Agreement with physician-determined classifications was moderate (Cohen's kappa = 0.47), and over half of conversations required human-provider escalation. Misclassified interactions were associated with lower patient usability scores compared to correct classifications (49.1 vs 60.8, P < .05). Thematic analysis revealed reliance on templated, administrative language.

Conclusions: Chatbots are practical and useful tools for managing elective plastic surgery inquiries but are ill-equipped to handle urgent and emergent patient needs. To move beyond utilization as basic administrative assistants, deployment of more clinically adept chatbots is needed.

The global demand for cosmetic procedures is accelerating, with over 1.6 million aesthetic surgical procedures performed in the US in 2023. Concurrently, AI is transforming surgical practice through advanced analytics, predictive modeling, and computer vision. Cosmetic surgery, characterized by subjective outcomes and limited standardized metrics, presents a unique opportunity for AI integration to enhance precision, objectivity, and patient communication. Following PRISMA 2020 guidelines, we systematically searched MEDLINE/PubMed, Embase, and the Cochrane Library (January 2020-July 2025) for studies applying AI, machine learning, deep learning, computer vision, or large language models to cosmetic or aesthetic procedures. Eligible designs included randomized controlled trials, observational studies, diagnostic accuracy studies, feasibility studies, and prediction model development. Two reviewers independently screened titles/abstracts, assessed full texts, extracted data, and evaluated risk of bias using ROBINS-I for non-randomized studies. Of 3941 records, 38 met the inclusion criteria. AI applications spanned preoperative planning (predictive risk modeling, 3D outcome simulation), intraoperative guidance (augmented reality overlays), and postoperative monitoring (smartphone-based complication detection, objective aesthetic scoring). Benefits included improved patient-surgeon communication, enhanced risk stratification, and standardized outcome measurement. However, most studies were early-phase, with limited external validation, heterogeneous datasets, and inconsistent outcome metrics. Risk of bias was moderate to serious in most studies. AI in cosmetic surgery shows significant potential but remains in early clinical adoption. Progress requires multicenter validation, standardized datasets, explainable algorithms, and clear regulatory frameworks. Large language model-driven tools may accelerate development and integration, provided ethical, equitable, and patient-centered principles guide implementation.

Background: Nasolabial-fold (NLF) severity is a key indicator of facial aging and a frequent target in aesthetic treatments. The Wrinkle Severity Rating Scale (WSRS) is widely used for clinical grading but remains inherently subjective and vulnerable to inter-observer variability.

Objectives: The authors of this study aim to develop and validate DeepFold, a deep learning-based ensemble model for automated, objective, and clinically interpretable grading of NLF severity based on the WSRS.

Methods: A dataset of 6718 facial images was constructed, including 1718 images from clinical outpatients and 5000 from the CelebA dataset. All images were split into left and right halves and annotated independently by 3 senior plastic surgeons using the WSRS. ResNet-50 served as the base model architecture, and an ensemble strategy was applied using majority voting over 3 independently trained networks. Model training used focal loss to address class imbalance and was conducted in PyTorch with early stopping based on validation loss. Performance was assessed using accuracy, F1 score, and confusion matrix analysis.

Results: The DeepFold ensemble model achieved a validation accuracy and F1 score of 0.917, outperforming individual baseline models such as ResNet-50 (accuracy: 0.904) and SeResNet-50 (accuracy: 0.882). Ensemble strategies reduced prediction variance and enhanced model robustness under class imbalance.

Conclusions: DeepFold provides a reliable and standardized approach to NLF severity assessment, offering potential clinical value in aesthetic evaluation, treatment planning, and outcome monitoring.

Botulinum toxin type A (BoNT-A) chemodenervation is commonly used for facial synkinesis and aesthetic indications, but landmark-based techniques are limited by anatomical variability and risk of off-target delivery. High-resolution ultrasound (US) can be used to enhance precision and safety. This systematic review explores the role of US-guided BoNT-A facial chemodenervation in evaluating anatomical accuracy, clinical efficacy, complications, and patient satisfaction. A comprehensive search of 6 databases through April 2025 identified studies assessing US-guided BoNT-A chemodenervation for facial indications. Sixteen studies were included, comprising randomized controlled trials, prospective cohorts, cadaveric trials, and anatomical mapping investigations. Data on injection accuracy, clinical outcomes, adverse events, and patient-reported measures were extracted. Risk of bias was assessed using validated tools. US guidance improved injection accuracy, with cadaveric trials demonstrating up to 88% accuracy, whereas landmark-based techniques reported 50%. Clinical studies reported improvements in rhytid reduction, oral commissure elevation, neck relaxation, and facial symmetry. Adverse events were infrequent and mild. Patient satisfaction was consistently higher with US guidance. Anatomical studies identified muscle depth variation and vascular risk zones, supporting real-time sonographic targeting. As a result, the authors found that US-guided BoNT-A chemodenervation improves the safety, precision, and outcomes and should be considered in both therapeutic and aesthetic applications. Level of Evidence: 3 (Therapeutic).

Artificial intelligence (AI) has become pervasive in and beyond plastic surgery. Myriad applications exist, and patients and plastic surgeons are increasingly turning to AI for information. This narrative review examines the current scope of AI applications in plastic surgery and highlights challenges and limitations based on current literature. A PubMed search for articles about or using AI in plastic surgery published before September 22, 2025 identified 1866 articles. Letters, commentaries, review articles, surveys, and articles not in the English language were excluded. Titles and abstracts were reviewed and studies classified according to AI modality, plastic surgery application, and subspecialty. Studies were classified under multiple categories, if applicable. This narrowed the results search to 460 qualifying articles, of which 54 involved patient education, 35 plastic surgeon education, 79 clinical decision-making, 62 outcome prediction or risk assessment, 46 clinical outcome assessment, 133 diagnosis, 46 practice management, and 17 research. Study methodologies and AI models varied widely. In terms of the types of AI used, 155 articles utilized large language models, 6 natural language processing, 9 text-to-imaging models, and 299 other machine-learning or deep-learning systems. Large language models were most often used in patient education studies, while machine learning predominated in diagnostic studies. AI spans the breadth of plastic surgery, although the literature is limited by heterogeneity. Plastic surgeons must know the advantages and opportunities provided by AI, while recognizing its limitations, pitfalls, and areas needing improvement. Ethical, safe, and forward-thinking AI in plastic surgery requires a multidisciplinary approach involving plastic surgeons, data scientists, ethicists, legal experts, and policymakers.

Background: Addressing postoperative scarring remains a significant challenge across all surgical procedures, often leading to long-term aesthetic deficits. Traditional methods for scar management have shown limited efficacy, prompting exploration into innovative approaches such as the application of stromal vascular fraction (SVF).

Objectives: The efficacy of intraoperative SVF injection in modulating scar formation was assessed by evaluating postoperative scar quality in patients undergoing free tissue transfer.

Methods: This multicenter, double-blinded, placebo-controlled, randomized study assessed the efficacy of intraoperative SVF injections in mitigating scar formation in patients undergoing free tissue transfer. A total of 45 patients were enrolled from three institutions, undergoing either breast reconstruction with abdominal flaps or soft tissue reconstruction with anterolateral thigh flaps. The donor site incision was split in half, with one side receiving SVF injections and the other saline injections, assigned randomly. Scar quality was evaluated using the Vancouver Scar Scale (VSS) and Patient Scar Assessment Questionnaire (PSAQ), along with objective measures of pigmentation using an analyzing device.

Results: The primary analysis endpoint at 6 months after surgery showed a statistically significant improvement in VSS scores in the SVF group compared to the saline group. However, these differences were not maintained at the 9-month follow-up. PSAQ results indicated improved satisfaction with scar appearance at 6 and 9 months in the SVF group, despite no significant changes in erythema and melanin levels.

Conclusions: Intraoperative SVF infection can improve postoperative scar appearance in patients undergoing free tissue transfer. This study supports further exploration of SVF as a potential tool for enhancing aesthetic outcomes in scar management.

Background: Concerns regarding information inaccuracy when using general-purpose large language models have prompted the quest for alternative tools. OpenEvidence has emerged as a healthcare-focused large language model trained exclusively on data from peer-reviewed medical literature.

Objectives: This study compared the quality, accuracy, and readability of aesthetic surgery patient education materials generated by OpenEvidence and ChatGPT.

Methods: A standardized prompt requesting comprehensive postoperative discharge instructions for 20 of the most common aesthetic surgery procedures was entered into OpenEvidence and ChatGPT-5. Outputs were evaluated using 4 validated assessment tools: the DISCERN instrument for information quality (1-5), the Patient Education Materials Assessment Tool for Printable Materials (PEMAT-P) for information understandability and actionability (0-100), the Flesch-Kincaid scale for estimated grade level (fifth grade to professional level) and reading ease (0-100), and a Likert scale for citation accuracy (1-4).

Results: OpenEvidence scored significantly higher than ChatGPT-5 in DISCERN (3.3 ± 0.4 vs 1.7 ± 0.4, P < .001) and the citation accuracy scale (2.4 ± 1.3 vs 1.5 ± 0.7, P = .007). Scores were comparable among both tools in PEMAT-P understandability (71 ± 5 vs 69 ± 0, P = .3) and actionability (52 ± 12 vs 54 ± 5, P = .6), as well as on the Flesch Kincaid Grade Level (9.3 ± 1.0 vs 9.2 ± 0.6, P = .8) and the Flesch Reading Ease Score (40.0 ± 6.6 vs 41.0 ± 5.5, P = .6).

Conclusions: OpenEvidence generated materials of significantly higher quality and reliability than ChatGPT, suggesting it may serve as a more reliable alternative for patient education in aesthetic surgery practice.