Scientific Abstract Presentations: Research & Technology Session 5 Saturday, October 11 Sat, Oct 11 10:30 a.m.-12:30 p.m. CDT

Purpose: Visiting student externships are essential to the plastic surgery residency application process. By providing opportunities for students to learn about programs through direct experience and for programs to evaluate applicants' clinical performances, externships allow a bidirectional assessment of fit. Multiple studies suggest that strong performances on sub-internships can improve one's chances of matching at a given institution1-2. This study aims to characterize plastic surgery residency program website content for visiting student externship information. This study aims to characterize information on externships posted by plastic surgery residency program websites and the visiting student learning opportunities (VSLO) platform.

Methods: All integrated and independent plastic surgery externship listings on residency program websites and VSLO were assessed for information regarding the rotation application process, logistics, and details. Chi-Squared and Fisher's Exact testing were used to compare comprehensibility by program characteristics including the residency model, program size by resident and faculty number, age in years by ACGME accreditation, amount of NIH/VA funding, Doximity ranking by prestige, and social media presence by number of Instagram posts and followers. McNemar testing was used to compare program information between websites and VSLO.

Results: A total of 104 residency programs were identified, with 85.6% listing information on their websites and 76.0% on VSLO. Information regarding rotation applications, logistics, and details, is listed on 87.6%, 85.4%, and 71.9% of websites and 100.0%, 100.0%, and 68.4% of VSLO pages. VSLO had significantly more information on the application process and rotation logistics than program websites (p≤0.004), while program websites had more rotation descriptions (p=0.003), information on call schedules (p = 0.011), presentations (p = 0.036), and simulation opportunities (p=0.008). Independent programs were significantly less likely to provide online information on their externships (p<0.001).

Conclusions: Program websites and VSLO are valuable tools for recruiting medical students to their externships, and eventually, to their residency programs. While most describe the application process and logistics of their externships, fewer provide detailed descriptions of what the rotation experience is like, making informed decisions on where to apply and rotate challenging. In addition, the content and degree of online information provided on these two platforms is variable. Improving the comprehensibility, accessibility, and consistency, of online plastic surgery externship listings will ultimately improve the residency recruitment process for all, especially orphaned medical students who otherwise lack resources and connections to the field.

1. Drolet BC, Brower JP, Lifchez SD, Janis JE, Liu PY. Away Rotations and Matching in Integrated Plastic Surgery Residency: Applicant and Program Director Perspectives. Plast Reconstr Surg 2016;137:1337-43.
2. Sergesketter AR, Glener AD, Shammas RL, et al. The Association between Away Rotations and Rank Order in the Integrated Plastic Surgery Match. Plast Reconstr Surg 2021;147:1050e-6e.

Background 3D printing has revolutionized plastic surgery by enabling detailed preoperative planning with lifelike anatomical models, enhancing surgical precision and outcomes. This study focuses on understanding the indications, characteristics, and outcomes of 3D-printed models produced through an institutional in-house, streamlined printing process at a radiology anatomic modeling unit (AMU). Methods This retrospective review included all the Plastic Surgery Division AMU orders at our institution. Orders from other departments were excluded. Data was categorized into two patient age groups (≤25 years, N=43; >25 years, N=52). Indication, segmentation times, CAD processing times, anatomical structures modeled, intended use, technologies used, and delivery time. Statistical analysis was performed with ANOVA and Chi-squared tests in BlueSky Statistics© 2024. Results In total 1400 AMU orders were requested at Rochester, of those, 95 were from our division placed by seven plastic surgeons from 2019 to 2024. Craniofacial structures were the most frequently modeled (61.7%), with specific models including skull reconstructions (34.6%) and combined structures such as the mandible and maxilla (21.0%). Chest models (6.2%) included the newly added indication for external breast prostheses customization (6.1%). Indications varied, with congenital conditions (27.7%) (primarily in patients ≤25 years) and trauma/reconstruction (31.9%) (more common in patients >25 years). Gender dysphoria cases (21.3%) were predominant in patients >25 years. The highest number of models were produced in 2020 (30.5%), followed by 2023 (22.1%) and 2019 (14.7%). Technological Insights: "Vat Photopolymerization" was the most used technology (64.2%), followed by "Material Jetting" (13.6%). Formlabs was the primary 3D printer brand (66.2%), particularly the "Form 3B" model (53.2%). The mean calculated final printing time was 28.16 ± 19.60 hours, and post-processing cleaning averaged 1.68 ± 1.36 hours. Multi-color models (58.3%) were produced more frequently than mono-color. Beyond clinical cases, the in-house printing process also facilitated the creation of medical simulation models, such as chest micro anastomosis models for breast reconstruction and comprehensive head and neck models, enhancing surgical training. Conclusion In-house 3D-printed anatomical models are vital for plastic surgery, especially in craniofacial and reconstructive procedures. The streamlined, institution-based process, predominantly using Formlabs technology and "Vat Photopolymerization," demonstrates high efficiency and quality. The addition of models for external breast prostheses highlights the expanding scope of 3D printing for non-surgical applications aimed at improving patient quality of life through enhanced body contour and fitting solutions. Future studies should investigate the correlation between 3D model use and patient outcomes and the benefits of medical training.

Introduction: Preoperative anxiety is a common experience prior to surgery and has been increasingly recognized for its impact on surgical outcomes. Traditional preoperative education largely focuses on discussing procedural information and expectations regarding the perioperative pathway and postoperative recovery, however, it varies widely based on surgeon preferences, institutional practices, and time constraints. Multimedia-based education, delivered through videos, websites, and mobile apps, has shown promise in reducing anxiety and opioid use while improving satisfaction and knowledge. Given the evolving landscape of digital technologies, their role in surgical education warrants further exploration. In plastic and reconstructive surgery, where patient expectations, aesthetic concerns, and psychological readiness plays a significant role in surgical outcomes, these multimodal modalities may further enhance perioperative care. This systematic review aims to evaluate the impact of multimedia preoperative educational interventions and explores its potential integration in standard surgical care.

Methods: A systematic review and meta-analysis of randomized controlled trials (RCTs) from 2004-2024 were conducted per PRISMA guidelines to evaluate the effect of preoperative surgical patient educational interventions on surgical outcomes. RCTs were retrieved from MEDLINE, CENTRAL, and EMBASE databases (December 2024) utilizing terms such as "surgery," "preop," "patient education," and "preoperative information." RCTs were included if they had a sample size of over 45 adult patients, assessed a preoperative education intervention beyond standard care, and measured relevant postoperative outcomes such as pain, anxiety, LOS, knowledge, and quality of life. Random-effects meta-analysis was utilized to estimate pooled effect sizes.

Results: 850 articles were initially identified. Upon review, 47 articles met inclusion criteria, comprising 4,580 patients across 13 surgical subspecialties. Various educational formats were used with videos being the most common (27.4%), followed by multimedia materials combining text, video, and animations (21.3%), written booklets (21.3%), and in-person sessions (12.8%). As for outcomes, anxiety was the most frequently assessed (48.9%), with studies also examining pain (36.2%), satisfaction (36.2%), knowledge (29.8%), quality of life (17.0%), and length of stay (8.5%). Pooled analysis showed that for studies assessing anxiety, the preoperative educational intervention significantly reduced anxiety postoperatively compared to the control group (standard mean difference (SMD) = 1.14, 95% confidence interval (CI): [0.02,2.26], p < 0.045, I2 = 95.0%). Similarly, pooled results from studies measuring postoperative pain demonstrated a significant reduction in postoperative pain among the intervention group (SMD = 0.56, 95% CI: [0.04, 1.06], p = 0.03, I2 = 74.9%). Lastly, analysis of studies on satisfaction found that the intervention significantly increased satisfaction compared to the control group (Odds Ratio = 1.98, CI: 1.44-2.73, p<0.0001, I2 = 0.00%).

Conclusion: Our systematic review found that multimedia preoperative educational interventions significantly reduce postoperative anxiety and pain while improving patient satisfaction. Given the impact of mental health on postoperative recovery, implementing such interventions may further improve surgical outcomes and overall patient care. In plastic and reconstructive surgery, where mental preparedness readiness is crucial, integrating such interventions into preoperative care can further strengthen patient confidence, enhance patient engagement, and improve adherence to postoperative care protocols.

Purpose: Metformin (MTF) has demonstrated pleiotropic effects, including the promotion of mesenchymal stem cell (MSC) proliferation and differentiation to enhance angiogenesis, particularly in diabetic wound healing. Adipose-derived stem cells (ASCs) are of great interest for wound healing due to their ability to differentiate into various cell types, including endothelial cells (EC), which are essential for neovascularization and tissue repair. Small intestinal submucosa (SIS), a biological scaffold commonly used in surgical tissue repair, has shown promise in regenerative medicine. In this study, we present a model using SIS as a scaffold for human ASCs to explore the role of metformin in enhancing ASC attachment, proliferation, and EC differentiation on the SIS scaffold.

Methods: Human ASCs were treated with metformin at varying concentrations (0–10 mM). Cell viability and proliferation were evaluated using the MTT assay and growth curve analysis. Growth factor secretion by ASCs was quantified via ELISA. Tracker Green-labeled ASCs were cultured on SIS scaffolds for three days, and cell adherence was assessed through live-cell staining and confocal microscopy. To assess endothelial differentiation and angiogenic potential, ASCs were cultured on SIS in EC induction media with or without MTF, followed by qPCR and immunofluorescence staining. RNA sequencing was conducted to characterize gene expression differences between MTF-treated and untreated groups.

Results: Metformin (0.1 mM) significantly increased ASC proliferation (1.8-fold, p = 0.05) and reduced doubling time compared to controls (p < 0.05). MTF-treated ASCs also exhibited enhanced secretion of FGF (1.8-fold, p < 0.01) and VEGF (1.2-fold, p < 0.05). Laser confocal microscopy demonstrated that MTF significantly improved ASC attachment and retention on the SIS scaffold membrane compared to untreated groups (46.7% vs. 37.2% in controls, p = 0.05) under normal and high glucose conditions. After 8 days of culture in endothelial growth media, MTF-treated ASCs showed increased mRNA expression of the EC-specific marker CD31 (2.44 ± 0.17 vs. 1.55 ± 0.21, p = 0.05) compared to controls. Immunofluorescence analysis further confirmed a higher distribution of CD31-expressing cells throughout the scaffold in the MTF-treated group. RNA-seq analysis revealed that metformin modulates ASC gene expression by regulating key transcription factors and signaling pathways associated with pro-angiogenesis and wound healing. An ongoing in-vivo study is evaluating the wound-healing effects of MTF-treated ASCs on scaffolds in a murine model.

Conclusion: These results indicate that: (1) Metformin enhances ASC proliferation, adhesion, and delivery while promoting endothelial differentiation on SIS scaffolds; and (2) coating a biological scaffold with metformin-treated ASCs may serve as an effective strategy to enhance the patency and functionality of stem cell-based therapies, particularly for wound healing, surgical reconstruction, and fat grafting in patients.

Background The perceived preparedness of graduating integrated plastic surgery (PRS) residents varies widely, potentially influenced by disparities in subspecialty exposure. Although the ACGME and ABPS provide general guidelines, residency curricula lack standardization. This study quantifies the variability in subspecialty exposure across integrated PRS programs, highlighting potential implications for competency development and the need for curricular reform.

Methods Rotation schedules, fellowship programs, and graduate career trends from 36 integrated PRS residency programs were analyzed. Subspecialties evaluated included hand, craniofacial, microsurgery, aesthetics, burn, gender affirmation, pediatrics, and wound care. Data were extracted to determine the number of dedicated weeks allocated to each subspecialty and identify statistical differences between programs. Variability was assessed using descriptive statistics and correlation analysis to examine the relationship between subspecialty exposure and fellowship pursuit.

Results Subspecialty exposure varied significantly across programs. The mean duration of exposure ranged from 4.1 weeks (SD: 1.8) in microsurgery to 2.8 weeks (SD: 1.8) in gender affirmation surgery, with only one program offering a dedicated gender affirmation rotation. While 50% of programs had dedicated rotations in hand, craniofacial, burn, and aesthetics, less than half provided structured exposure to microsurgery. Programs with dedicated microsurgery, craniofacial, or aesthetics fellowships allocated significantly fewer weeks to these subspecialties during residency (p < 0.05), suggesting a reliance on post-residency training to supplement exposure. Conversely, programs with burn and pediatric fellowships provided significantly increased exposure to these fields. Correlation analysis revealed key trends. Microsurgery and gender affirmation surgery exhibited a negative correlation (-0.33 to -0.44) with fellowship matches, suggesting that residents with limited residency exposure in these fields are more likely to pursue specialized fellowships to supplement their training. In contrast, craniofacial surgery showed a slight positive correlation (+0.35), indicating that greater residency exposure may encourage further specialization but is not a strong predictor of fellowship pursuit. Aesthetics and hand surgery exhibited no significant correlation (p > 0.3) with fellowship choices, implying that personal career goals, mentorship, and market demand may be stronger determinants than exposure duration.

Conclusions This study provides a comprehensive quantitative analysis of subspecialty exposure across integrated PRS residency programs, revealing substantial inconsistencies. The findings suggest that reliance on fellowship training to supplement gaps in residency education may not be sustainable, particularly for niche subspecialties like gender affirmation and microsurgery. Correlation analysis indicates that limited residency exposure may be driving fellowship pursuit in certain fields, emphasizing the need for curricula to provide baseline competency in high-demand subspecialties. To best address and obviate such inconsistencies, programs should clarify whether subspecialty exposure is explicitly structured or incorporated within general plastics rotations, as some training experiences may not be separately advertised but remain embedded in residency training. Standardized minimum exposure durations could enhance consistency in training and optimize resident preparedness for independent practice. Further discussions within the PRS education community are warranted to establish evidence-based curricular reforms.

Background: The advent of general-purpose large language models (LLMs) like ChatGPT has revolutionized natural language processing, but their applicability in specialized medical fields like plastic surgery remains limited due to a lack of domain-specific knowledge. This study aims to develop and evaluate PlasticSurgeryGPT, a dedicated LLM fine-tuned on plastic surgery literature, to enhance performance in clinical decision support, surgical education, and research within the field.

Methods: A comprehensive dataset of 25,389 plastic surgery research abstracts published between January 1, 2010, and January 1, 2024, was retrieved from PubMed. The abstracts underwent rigorous preprocessing, including text cleaning and tokenization. We fine-tuned the pre-trained GPT-2 model on this dataset using the PyTorch and HuggingFace frameworks. The performance of PlasticSurgeryGPT was evaluated against the default GPT-2 model using BLEU, METEOR, and ROUGE-1 metrics.

Results: The fine-tuned model, named PlasticSurgeryGPT, demonstrated substantial improvements over the generic GPT-2 model in capturing the semantic nuances of plastic surgery text. PlasticSurgeryGPT outperformed GPT-2 across BLEU, METEOR, and ROUGE-1 metrics, with scores of 0.135519, 0.583554, and 0.216813, respectively, compared to GPT-2's scores of 0.130179, 0.550498, and 0.215494.

Conclusion: PlasticSurgeryGPT represents the first plastic surgery-specific LLM, demonstrating enhanced performance in generating relevant and accurate content compared to a general-purpose model. This work underscores the potential of domain-specific LLMs in improving clinical practice, surgical education, and research in plastic surgery. Future studies should focus on incorporating full-text articles, multimodal data, and larger models to further enhance performance and applicability.

Purpose:
Artificial intelligence (AI) is growing increasingly useful in medicine to help guide practitioners and recognize patterns within patients. While previous research has explored AI's role in staging indocyanine green (ICG) lymphography patterns, none have utilized ChatGPT. This study evaluates ChatGPT's ability to differentiate between varying severities of dermal backflow using ICG lymphography, offering an easily accessible diagnostic tool for lymphatic surgeons and therapists.

Methods:
We utilized a dataset comprised of 40 upper extremity ICG lymphographs diagnosed as either diffuse, linear, splash, or stardust by a lymphatic surgeon. Three reviewers independently used ChatGPT to stage the ICG images provided. Cohen's Kappa statistics were used to compare inter-rater reliability between the various ChatGPT responses and in comparison, to the professional diagnoses.

Results:
Overall, we found that ChatGPT was 93% accurate in recognizing the presence of dermal backflow. Inter-rater reliability between the reviewer responses indicated fair agreement between the three groups (κ=0.24 (p<0.001) and κ=0.32 (p<0.001)). After combining each reviewer's ratings into a composite ChatGPT group, we compared ChatGPT responses to the lymphatic surgeon diagnoses. We found that ChatGPT exhibited a fair agreement with the reference staging of lymphatic patterns with a Cohen's kappa of 0.4 (p<0.001). ChatGPT's accuracy in identifying lymphatic patterns was 80% for linear, 50% for splash and stardust, and 40% for diffuse.

Conclusion:
ChatGPT demonstrated fair levels of agreement with a lymphatic surgeon's staging of various dermal backflow patterns. It was most accurate in identifying linear flow patterns but showed promise in identifying more severe forms of lymphatic disruption. With increased utility, this learning model may become a useful tool to aid in diagnostic imaging.

Purpose: The United States Medical Licensing Exam Step 1 has historically been an important criteria used to assess the competency of medical students applying for plastic surgery residency. However, in 2022, Step 1 transitioned from a numerical score to a pass/fail grading system, redirecting emphasis towards other components of students' portfolios, particularly research. This study aims to understand if the change to pass/fail has had a measurable impact on student authorship trends in plastic surgery journals.

Methods: Publications from 2017 to 2024 in Plastic and Reconstructive Surgery and Aesthetic Surgery Journal were reviewed. Authors were categorized based on their credentials. "Student authors" were defined as authors having only the following degrees: BA, BS, MA, MS, MSE, or MPH. "Doctor authors" were defined as authors with any of the following degrees: MD, DO, PhD. Articles were divided into "pre-Pass/Fail" (2017–2021) and "post-Pass/Fail" (2022–2024) cohorts. Independent two sample t-tests were used to compare means.

Results: 9,557 articles were identified: 6,560 (64%) in the pre-Pass/Fail cohort and 3,267 (36%) in the post-Pass/Fail cohort. There were 41,581 authors across all publications, including 3,453 (8%) students and 34,101 (82%) doctors. 2,086 (22%) manuscripts included a student author, while 7,471 (78%) did not. Publications with a student author averaged 6.13 total authors (SD=2.72), compared to 3.85 (SD=2.68) for publications without a student (p<0.001). In the pre-Pass/Fail cohort, there were, on average, 0.31 student authors and 3.37 doctor authors per publication. This rose to 0.46 students (p<0.001) and 3.96 doctors (p<0.001) in the post-Pass/Fail cohort, representing 48% and 18% increases, respectively. 2022, the year Step 1 changed to Pass/Fail, saw the greatest increase (30%) in students per publication. In six of the seven years analyzed (86%), the average number of student authors per publication either increased or remained unchanged.

Conclusions: The transition to a pass/fail grading system for Step 1 has coincided with an increase in the average number of student authors per publication in both Plastic and Reconstructive Surgery and Aesthetic Surgery Journal. However, our data also reveals a consistent yearly rise in average authorships for both students and physicians. This finding may reflect a broader trend in biomedical research, as prior research has shown that the mean authors per publication increased by 57% over the last two decades. However, the greatest increase in student authors occurred in 2022, the year that Step 1 was changed, suggesting that the switch may have amplified the existing trend. Our analysis also indicated that publications featuring students had a greater number of authors. Moreover, the 3,453 students in our sample were concentrated into 2,086 manuscripts. Together, these findings suggest a shift towards collaborative, multi-student projects as students aim to strengthen their academic portfolios. The growing number of authors in academic publications raises questions about the value of individual contributions in publications. Further research is warranted to better understand the impact of the increasing authorships and to determine if these trends persist in other plastic surgery journals.

Objective:**
Breast-related plastic surgeries, such as implant placement, breast reduction, and breast reconstruction, require detailed preoperative planning. Parameters such as breast area, volume, and bilateral symmetry play a crucial role in selecting implant size, determining the amount of glandular tissue and fat to be removed, and preserving the thickness of the nipple-areola vascular pedicle. In clinical practice, these assessments are typically based on manual markings and subjective evaluations by surgeons using their experience. Artificial intelligence and neural networks can automatically and rapidly identify breast features, enabling fast, accurate, and three-dimensional objective assessments of breast-related metrics.

Methods:
This study prospectively included 300 patients scheduled for breast plastic surgery, collecting multi-angle macroscopic photographs preoperatively. Based on deep learning, preoperative breast feature calculation models were constructed using these photographs as input. The main models developed include: an automated breast segmentation model, a feature point detection model (identifying the suprasternal notch and bilateral nipple-areola complexes), a 3D information estimation model, and a symmetry analysis model. These models enabled the automatic calculation of preoperative breast-related features and provided preoperative and intraoperative visualizations to guide surgical decision-making and intraoperative adjustments.

Results:
In this study, the automated breast segmentation model achieved a correlation of 0.93 with measured values in identifying breast area size. The feature point detection model achieved a 100% detection rate for the suprasternal notch and bilateral nipple-areola complexes. For the 3D information estimation and symmetry analysis models, the visualized images received a 90% satisfaction rate for preoperative utility, 95% for intraoperative utility, and a 96% long-term patient satisfaction rate during follow-up.

Conclusion:
By integrating multi-stage deep learning models, this study established an automated quantitative analysis and 3D visualization evaluation system for breasts based on macroscopic photographs. The system achieved breast area measurement, volume estimation, and symmetry assessment from 2D images. It demonstrated high segmentation accuracy and volume estimation performance, enabling further evaluation of implant size, refinement of preoperative surgical decisions, and prediction of postoperative outcomes. Intraoperative bilateral symmetry visualization allowed for dynamic quantitative adjustments to the surgical plan, achieving high clinical satisfaction. This system provides objective, rapid, and precise breast assessment, offering robust support for refined and personalized treatment of breast plastic surgery patients.

Purpose: Peripheral compartment syndrome (PCS) occurs in up to 10% of tibial fractures, with mortality rates as high as 15% in patients requiring fasciotomy. Diagnosis relies on clinical evaluation and intracompartmental pressure (ICP) measurements, which identify only 68% of cases. Prior research shows that although invasive, intramuscular (IM) biomarkers can diagnose PCS. We hypothesized that subcutaneous (SC) biomarkers closely correlate with IM metabolic changes and can detect PCS.

Methods: PCS was induced in the peroneus tertius of Yorkshire swine (Male, ~60 kg, n=22 limbs) using a tourniquet on bilateral thighs. ICP was measured every 30 minutes, and PCS was maintained for ~6 hours. Microdialysis was used to collect hourly interstitial fluid (ISF) samples from muscle (n=22) and overlying SC tissue (n=10). IM and SC ISF were analyzed for glucose, pH, and lactate.

Results: ICP increased after tourniquet inflation (pre-inflation: 15.6mmHg ± 3.9 vs. 6 hours: 54.4mmHg ± 37, p<0.0001). There were no significant differences in SC and IM biomarkers at any time point. IM and SC pH decreased significantly at 2 hours (p=0.005, p=0.0007, respectively) and were highly correlated (R² = 0.911). SC pH <7.2 had a 100% positive predictive value for detecting PCS. SC and IM glucose showed the highest correlation (R²=0.98). IM glucose decreased significantly at Hour 2, but this was not seen in SC glucose. SC and IM lactate levels increased non-significantly in parallel (R²=0.61).

Conclusions: SC ischemic biomarkers closely correlate with IM changes in a porcine model of PCS. Positive results require confirmation in a pilot clinical trial.

INTRODUCTION: Effective postoperative guidance is essential for optimizing patient outcomes and satisfaction after any plastic surgery procedure. Using Retreival-Augmented Generation (RAG), Large Language Models (LLMs) become safer tools for providing accurate postoperative support. However, their performance is dependent on how information is organized for retrieval. Thus, an optimal chunking strategy can notably enhance the accuracy and relevance of the models's responses. In this study, we implemented three different advanced chunking techniques to determine whether they could outperform a naïve RAG-based LLM in providing reliable, precise postoperative advice for rhinoplasty patients.

METHODS: Four different RAG architectures were implemented into Gemini 1.0 Pro, leveraging distinct chunking techniques: (1) proposition, (2) semantic, (3) adaptive, and (4) fixed token-dependent (naïve). Each model was presented with 30 postoperative questions addressing common patient concerns after rhinoplasty, and performance was evaluated in terms of medical accuracy, relevance, precision, recall, and F1 score. A 3-point Likert scale (LS) was used for the former two, and metrics were compared using ANOVA and Tukey's post hoc analysis.

RESULTS: The adaptive chunking model demonstrated the highest medical accuracy of 87% (LS 2.37±0.72), which was significantly superior to the naïve model with a score of 50% (LS 1.63±0.72), p-value of 0.001. The proposition (LS 2.07±0.78) and semantic (LS 2.03±0.76) chunking models achieved 73% accuracy, showing no statistical difference compared to the naïve model. The adaptive chunking model provided the most relevant responses, scoring 93% (LS 2.90±0.40), while the semantic chunking model scored 90% (LS 2.87±0.43), the proposition chunking one scored 87% (LS 2.80±0.55), and the naïve model 80% (LS 2.60±0.81). There was no statistical difference between the models (p=0.18). For precision, recall, and F1 score, the model employing adaptive chunking outperformed the rest by scoring 0.50, 0.88, and 0.64, respectively, followed by the proposition (0.39, 0.71, 0.50), semantic (0.33, 0.75, 0.46), and naïve (0.17, 0.40, 0.24) models.

CONCLUSION: This study demonstrates the critical role of optimizing chunking strategies in enhancing the performance of RAG-based LLMs for postoperative care after rhinoplasty. Although adaptive chunking proved to be significantly superior to fixed-sized chunking, both proposition and semantic chunking improved the models' performance scores. These findings highlight the importance of continuing to identify the best methods for segmenting medical knowledge to improve LLMs' understanding and retrieval. Optimal chunking could ultimately lead to safer and more effective postoperative support.

Abstract Integrin receptors, heterodimeric transmembrane glycoproteins composed of α and β subunits, play pivotal roles in mediating bidirectional signaling between cells and the extracellular matrix (ECM). Dysregulated integrin expression in localized scleroderma (LoS) fibroblasts drives pathological fibrosis via excessive proliferation, TGF-β pathway activation, and ECM remodeling. Targeting integrins represents a promising therapeutic strategy for LoS. The cyclic iRGD peptide , containing the Arg-Gly-Asp (RGD) motif, exhibits dual functionality: high integrin affinity and cell-penetrating properties. Upon binding to integrins, iRGD undergoes proteolytic cleavage to generate CRGDK/R fragments that interact with neuropilin-1 (NRP-1), facilitating intracellular drug delivery. Here, we engineered iRGD-modified iPSC-derived exosomes (iRGD@iPSC-exo) to enhance targeted delivery to LoS fibroblasts.

iPSC-exo were isolated via ultracentrifugation and conjugated with iRGD using PEG-2000-cholesterol anchors. Conjugation efficiency was quantified fluorometrically (FAM-labeled iRGD: excitation/emission 488/525 nm), revealing a loading efficiency of iRGD@iPSC-exo retained canonical exosome characteristics (50–150 nm diameter, cup-shaped morphology by TEM, CD63/TSG101 positivity) post-modification. To validate targeting specificity, DiD-labeled LoS fibroblasts were co-cultured with unlabeled keratinocytes/endothelial cells and treated with PKH67-labeled iPSC-exo or iRGD@iPSC-exo. Confocal microscopy demonstrated significantly higher co-localization of iRGD@iRGD@iPSC-exo with LoS fibroblasts (p<0.01 vs. unmodified exosomes), with reduced non-specific uptake by bystander cells.

This study establishes a robust method for generating integrin-targeted exosomes. By leveraging iRGD's dual targeting and penetration capabilities, iRGD@iPSC-exo achieves selective delivery to pathological fibroblasts, minimizing off-target effects and amplifying anti-fibrotic efficacy. These findings highlight the translational potential of engineered exosomes for precision therapy in fibrotic skin disorders, offering a novel approach to modulate aberrant integrin signaling in LoS. Further in vivo validation of therapeutic outcomes is warranted to advance clinical applications.

Introduction: Trainee physicians often experience feelings of guilt and fear of judgment surrounding pregnancy and lactation. Plastic surgery residents report a negative stigma attached to pregnancy and frequently delay childbearing until after training. Prior studies show that a barrier to parenthood experienced by many trainees is a lack of support and resources by their program. To further investigate, this study aims to examine and understand faculty knowledge and perceptions surrounding family planning, parental leave, and lactation.

Methods: A 21-question Qualtrics-based survey was created based on previously validated surveys in medical literature. This survey was distributed to the American Council of Academic Plastic Surgeons (ACAPS) membership which includes 503 faculty members. Data on demographics, personal experiences with parenthood, perceptions, and knowledge regarding lactation was collected and qualitatively analyzed.

Results: 35 faculty members completed the survey. The majority of respondents were within the 40 to 49-year age range, 54.3% identified as female, 40% identified as male, and two respondents declined to report their demographics. A large majority (88.7%) of faculty reported having children. Nineteen had a child/children during residency and nineteen had a child/children as an attending. Of those who had children during residency, 31.6% did not take parental leave during residency and 42.1% used vacation time as parental leave. The mean amount of time of parental leave in training was 5.39 weeks, ranging from 0 to 16 weeks. Regarding attitudes towards resident lactation, 100% of faculty members feel comfortable discussing a resident's lactation needs, while all but one strongly support a resident's need to pump. 88.9% of faculty strongly believe their institution has a supportive culture towards lactating residents. 34.3% of faculty agree that there is enough time in the workday for residents to pump. However, over half of faculty (54.3%) agree that the workflow of the operating room is impacted when a resident breaks scrub to pump.

Conclusions: Family planning impacts all levels of plastic surgeons from residents to faculty. Residency programs and faculty are becoming increasingly aware of the importance of supporting lactation in the hospital. However, challenges and stigma remain, especially surrounding the time needed for parental leave and pumping coupled with the intense hours of surgical training. By fostering a more supportive culture within plastic surgery, all plastic surgeons can work towards a healthy balance between surgery and parenthood.

background: Social media is one of the most accessible sources of information for patients. Symptoms, diagnoses, and treatment options are commonly searched topics on these platforms which present an overwhelming amount of information. Previous work has shown low quality and reliability of Youtube videos regarding varying medical issues (1). The primary objective of this study was to measure the quality of information presented to a searcher on YouTube regarding breast reconstruction after mastectomy, one of the most common reconstructive plastic surgery procedures.

Methods: A YouTube account was created to reduce search algorithm confounding from prior search history. Breast reconstruction was searched and videos were filtered by popularity. The first unique 100 videos that comprehensively addressed the search term were included. Each video was evaluated by two surgeons using JAMA and DISCERN scoring systems (2)(3). Videos were categorized as hospital or non-hospital affiliated. Two sample t-test was used for statistical analysis.

Results: 60 percent of the videos were hospital affiliated. When analyzed with JAMA and DISCERN scores, there was no difference in quality between hospital-affiliated and non-hospital affiliated videos. On average, both groups scored fair quality (approximately 3 out of 5 points) for the information presented. When DISCERN variables were compared between hospital and non-hospital affiliated videos, we found that non-hospital affiliated videos were significantly better at providing information for authorship and citing references than Hospital-affiliated videos (2.1 vs 2.7, p <0,04). Notably, we found that there is no statistically significant difference in views between non-hospital and hospital affiliated videos (711600 vs 2049600, p . 0,6) and a significantly higher number of likes for videos that were non-hospital affiliated. (5,511vs 3105, p <0,04).

Conclusion: Youtube has now become an essential tool for patient education. Providers should be aware of the quality of content that patients are consuming. Awareness of the quality of content can be an invaluable tool utilized to aid patients' education about complex medical procedures when discussion at the office comes up as short. This study elucidates that video developers can utilize these scoring tools to identify areas of improvement and increase the quality of information offered.

References 1. Kara M, Ozduran E, Mercan Kara M, Hanci V, Erkin Y. Assessing the quality and reliability of YouTube videos as a source of information on inflammatory back pain. PeerJ. 2024;12:e17215. Published 2024 Apr 11. doi:10.7717/peerj.17215 2. Charnock D, Shepperd S, Needham G, Gann R. DISCERN: an instrument for judging the quality of written consumer health information on treatment choices. J Epidemiol Community Health. 1999;53(2):105-111. doi:10.1136/jech.53.2.105 3. Silberg WM, Lundberg GD, Musacchio RA. Assessing, Controlling, and Assuring the Quality of Medical Information on the Internet: Caveant Lector et Viewor-Let the Reader and Viewer Beware. JAMA. 1997;277(15):1244–1245. doi:10.1001/jama.1997.03540390074039

Background: Intraoperative bioprinting (IOB) of scaffolds offers a new frontier in reconstructive surgery. Scaffold success relies on rapid vascularization, which can be enhanced through endothelial cell seeding. We have described micropuncture ("MP"), a surgical technique in which the recipient blood vessel wall is precisely disrupted to expedite scaffold vascularization. In this study we examine the combined effects of MP and cell-seeding in a collagen-based bioink at short and long timepoints, hypothesizing that the combination will promote vascularization that continues with time. Methods: In a rat model MP was performed on femoral vessels on one hindlimb while the other served as control. Bioink scaffolds with or without rat aortic endothelial cells ("cell") were printed over the vessels using an extrusion bioprinter. On POD10 or 40, fluorescence angiography was performed and vascular metrics were evaluated with artificial intelligence. Results: All experimental groups demonstrated significantly increased vascular surface area compared to the POD10 bioink only control, and these results were augmented with time. The greatest vascularization was observed in the Bioink+Cell+MP group at POD40. The addition of MP alone also significantly augmented vascular development. Conclusions: This study introduces a new approach that combines MP and IOB to create vascularized tissue. Both MP and cell seeding show potential as effective pro-angiogenic strategies. Future research will focus on utilizing these techniques to develop in-situ flaps for tissue reconstruction.

Purpose Microsurgery is a critical component of plastic and reconstructive surgery. However, the high costs associated with training and equipment pose significant barriers to its adoption, especially in low-resource nations. To address these challenges, the 3D printed microscope was developed by the Microsurgery and Regenerative Medicine Lab within the Division of Plastic Surgery at the University of Wisconsin. This 3D-printed microscope has already been utilized for microsurgery education and training both in the United States and internationally. The primary objective of this study was to validate the use of this novel microscope in microsurgery training and education by surveying surgical residents and attending surgeons about their experiences using this tool relative to traditional magnification methods (e.g., surgical loupes, benchtop microscopes, operating room microscopes).

Methods In this prospective, multi-institutional validation study, surgical residents and attending surgeons with varying levels of experience in microsurgery were invited to perform a single end-to-end microsurgical anastomosis on blue-blood chicken thigh femoral artery vessels using the 3D printed microscopes. Upon completion of the exercise, the participants completed two surveys evaluating and sharing their experience with the microscope. The surveys collected demographic information of the participants and assessed the microscope's adequacy for microsurgical training, ease of use, educational value, and its potential to enhance teaching capacity both locally and globally.

Results Among the participants for the first survey (n=26), 10 (38.5%) were attending surgeons, 2 (7.7%) were fellows, and 14 (53.8%) were residents. Of the respondents, 79.2% (19/24) reported that performing the end-to-end anastomosis was either easier or equivalent under the 3D printed microscope compared to using surgical loupes. All participants expressed a willingness to use the 3D printed microscope for microsurgery teaching and practice, citing its adequate quality and educational value. Additionally, 92.3% (24/26) believed that access to 3D printed microscopes would enhance their capacity to teach microsurgery at their institutions, and 96.0% (24/25) agreed it would improve their ability to teach microsurgery in other settings, such as global surgery trips. Comments provided in the survey reflected overwhelmingly positive feedback, with suggestions for enhancements related to ergonomics and lighting.

Conclusion This validation study of the 3D printed microscopes supports the continued development and refinement of the design based on expert and trainee insight. Implementing these improvements will optimize the microscope's utility in both local and global microsurgery education courses, ultimately expanding access to microsurgery training worldwide.

Introduction Plastic surgery has played a key and visible role in the development of global surgery programs worldwide, which attracts many medical students to the field. This study aimed to identify interest in and perceptions towards global surgery among US medical students interested in plastic surgery, and compare them to students interested in other surgical specialties.

Methods Medical students matriculating at US-based medical schools were distributed an online survey via Qualtrics. Respondents anonymously answered several questions about personal background and future medical interests, and rated agreement with multiple statements about global surgery on a 5-point Likert scale. Descriptive statistics were used characterize survey responses, with logistic regression used to compare survey answers between students interested in plastic surgery and those in other surgical specialties. Results Surveys were completed by 726 students from 37 US-based medical schools. There were 107 (14.7%) students who reported interest in plastic surgery (PS) as a specialty, although the majority of these students (94.4%) also reported interest in at least one other surgical subspecialty. Among all PS students, 58 (54.2%) reported being "very" or "extremely" interested in participating in global surgery programs as a part of their future careers. Students predominantly expressed in participating in surgical trips/missions (91.1%) and research (91.1%), but a majority also expressed a desire to participate in global policy work and/or advocacy (55.4%). Only 26 (44.8%) of PS students interested in global surgery identified as having a mentor in global surgery. Students reported that their mentors predominantly participated in short term trips (80.8%), with less frequent participation in research (57.6%), education (46.1%), or advocacy/policy (26.9%). Only a minority of PS students perceived current global surgery practices to be ethical (44.5%). There were no significant differences in the proportion of students interested in global surgery and mentorship, the perceptions of global surgery, or future goals in global surgery when comparing PS students with students interested in other surgical specialties.

Conclusion Despite demonstrating a strong desire to incorporate global surgery into their future careers, students interested in plastic surgery lacked mentorship and reported high concerns about the ethics of current global surgery practices. Efforts to develop – and promote – holistic and ethical global surgery practices may help improve the overall perception of plastic surgery and recruit motivated trainees to the specialty.

This study evaluates the effectiveness of using the ARTIS-pheno robotic C-arm system in a Hybrid Operating Room (HOR) for treating vascular malformations. Traditionally, such treatments require separate procedures in an angiography suite followed by surgery in an operating room, increasing risks due to patient transfers. To address this, a HOR was established in 2022, enabling both radiological and surgical interventions in one room, thus enhancing procedural efficiency and patient safety.

Between 2022 and 2024, six cases of vascular malformations were treated using the ARTIS-pheno system, including eyebrow arteriovenous malformations (AVMs), occipital residual AVM, frontal AVM, lymphatic malformations, a large AVM in the right thigh, and an elbow arteriovenous fistula. The procedure involved angiography followed by transarterial embolization using coils or embolization agents, with balloon dilation for flow control by a radiologist. Subsequently, a plastic surgeon performed ligation, excision, and reconstruction in the same setting, eliminating the need for patient transfers.

All vascular malformations were completely resolved, confirmed by follow-up angiography, with no recurrences after six months. Patients reported high satisfaction with their outcomes. This integrated approach significantly reduced the complexity and risks associated with traditional multi-step procedures, leading to improved workflow and enhanced patient safety.

The ARTIS-pheno robotic C-arm system allowed for precise imaging and real-time navigation, facilitating effective embolization and surgical intervention. Additionally, the development of a procedural algorithm improved safety and efficiency by streamlining workflow and reducing patient movement between different rooms.

This study demonstrates that the ARTIS-pheno system in a HOR is an effective and safe approach for treating vascular malformations, offering several advantages, including improved patient safety, better procedural efficiency, and high patient satisfaction. The results suggest that this integrated approach provides high-quality treatment while reducing the risks of traditional methods. However, further studies with larger case numbers are needed to validate these findings and assess the broader applicability of this approach for complex vascular malformations and other intricate surgical procedures.

In conclusion, the use of the ARTIS-pheno robotic C-arm system within a Hybrid Operating Room offers significant advantages in treating vascular malformations. This innovative approach has the potential to improve patient outcomes and procedural efficiency, paving the way for advancements in the management of complex vascular conditions.

Purpose: Burn injuries are a major global health issue, affecting 8.4 million people annually, with most cases occurring in low- and middle-income countries (1, 2). Children are particularly vulnerable, with 24% of cases affecting those under fifteen, primarily from scald injuries (3). Given the high incidence of burns, prevention strategies are essential. Gamified educational models have gained attention, especially among Generation Z, but digital approaches have faced high dropout rates. This study evaluates the effectiveness of an in-person gamified educational intervention in teaching young students about burn prevention and first aid.

Methods: This study was conducted between October 2023 and August 2024 in three high schools (HS) and one vocational school (VS), totaling four Brazilian public schools. Initially, all students completed a 21-question online test (Pre-Test) focused on the classification and prevention of burns, without access to their scores or the correct answers. Following the test, our group presented a 30-minute lecture covering the same topics as the questionnaire to reinforce the understanding of burn classification and prevention. After the lecture, the students were divided into groups to compete in an interactive online trivia game designed to enhance learning retention and engagement (Burn Game). To further motivate participation, the winning team received customized prizes. Finally, all participants individually retook the 21-question online test (Post-Test). The primary outcome of the study was the difference in scores on Pre- and Post-tests.

Results: A total of 134 young students were enrolled in the study. Out of these, 73 were from VS, and 61 were from HS. The global mean age was 18.58 ± 4.86 years, and 98 (73.13%) participants were female. Analyzing the scores, 132 out of the 134 participants had better scores after the intervention. Globally, we found a 73.9% increase comparing the mean scores of the Pre- and Post-Test (12.66 ± 3.57 vs. 19.82 ± 1.97, p < 0.001). In the subgroup analysis, VS students showed a 53.2% increase (14.18 ± 3.01 vs. 20.26 ± 1.28, p < 0.001), while HS students showed a 98.6% increase (10.84 ± 3.37 vs. 19.30 ± 2.48, p < 0.001).

Conclusion: With substantial difference in scores between the Pre- and Post-Test, our study suggests that gamified teaching was indeed able to significantly increase students' knowledge about burn prevention and classification.

(1) Yakupu A, Zhang J, Dong W, et al. The epidemiological characteristic and trends of burns globally. BMC Public Health. 2022;22(1):1596. doi:10.1186/s12889-022-13887-2.

(2) Rybarczyk MM, Schafer JM, Elm CM, et al. A systematic review of burn injuries in low- and middle-income countries: epidemiology in the WHO-defined African region. Afr J Emerg Med. 2017;7(1):30-37. doi:10.1016/j.afjem.2017.06.003.

(3) Haynes HJG. Fire loss in the United States during 2016. National Fire Protection Association. Published 2017. Accessed November 15 2023. https://www.cafsti.org/wp-content/uploads/osFireLoss.pdf.