Enhancing Safety and Outcomes in Brazilian Butt Lift: Mobile Ultrasound With Artificial Intelligence

Abstract

Background:

The Brazilian butt lift (BBL) is increasingly guided by ultrasound to avoid deep or intramuscular fat placement, which has been associated with fatal pulmonary fat embolism. Real-time visualization of the superficial and deep gluteal fasciae improves safety, but consistent plane identification can remain challenging. Artificial intelligence (AI) may augment ultrasound by assisting with tissue-layer recognition, cannula tracking, and standardized interpretation.

Methods:

A narrative review of gluteal anatomy, historical safety guidelines, and contemporary evidence on ultrasound-guided BBL was performed. We additionally outlined an AI-assisted ultrasound workflow and described a detailed surgical technique using mobile, handheld ultrasound with integrated AI-based optimization features. A preliminary retrospective clinical experience using this method was evaluated to assess feasibility and early safety signals.

Results:

Ultrasound reliably enables identification of the gluteal fascial layers and supports subcutaneous-only fat placement. AI-assisted visualization further enhances anatomical consistency, facilitates plane confirmation during injection, and may reduce operator-dependent variability. In our early clinical experience, the workflow was feasible, with no major complications observed and satisfactory short-term outcomes.

Conclusions:

Ultrasound guidance has significantly improved the safety profile of BBL by allowing continuous confirmation of subcutaneous fat placement. The integration of AI offers a promising adjunct that may enhance precision, standardize technique, and expand accessibility for surgeons with varying levels of experience. Preliminary findings support the practicality of AI-assisted ultrasound; however, larger datasets are needed to validate its impact on outcomes and long-term safety.

Takeaways

Question: Can mobile ultrasound coupled with artificial intelligence (AI) improve safety and outcomes in Brazilian butt lift?

Findings: The deep subcutaneous fat layer offers improved graft survival and aesthetics. Ultrasound guidance enhances safety, and AI supports image interpretation and standardization.

Meaning: AI-assisted ultrasound may standardize BBL techniques, reduce complications, and enhance aesthetic outcomes.

INTRODUCTION

The Brazilian butt lift (BBL) gained worldwide popularity given its successful results and rapid recovery.^1–4 Using autologous fat grafting, this technique demonstrates its effectiveness in enhancing buttock size and shape naturally.^1,4 However, complications including fat embolisms—often fatal and linked to deep injections—have arisen following BBLs.^1,3–6 Hence, international surgical societies now recommend subcutaneous fat placement rather than intramuscular injection.^1,4,7 Still, implementing this guideline remains difficult, as discerning tissue layers solely on tactile feedback remains challenging.² This issue is addressed by mobile ultrasound machines, offering real-time intraoperative visualization of tissue layers.² Although artificial intelligence (AI) has been widely incorporated in medicine, its use in ultrasound-assisted BBL procedures remains uninvestigated. This review investigated the anatomy, technology, and clinical implications of ultrasound-guided, AI-enhanced BBL. We also provided an overview of our AI-assisted, ultrasound-guided injection technique and included a preliminary retrospective assessment of our early clinical experience using this method, aiming to offer practical insights regarding feasibility and safety.

ANATOMY REVISITED: WHY THE DEEP SUBCUTANEOUS LAYER MATTERS

Gluteal anatomy must be mastered to perform BBLs safely and effectively. Two gluteal fasciae exist beneath the dermis: the deep gluteal fascia (DGF), below which fat grafting is prohibited, and the superficial gluteal fascia, which separates the deep and superficial subcutaneous layers.^3,7 Although both subcutaneous layers can accommodate fat grafts, the deep layer is preferred for its advantageous properties, the first of which is that it features less fibrous septa than the superficial layer, promoting an even fat integration and distribution.³

Second, according to a number of fat grafting studies, fat extracted from and injected into the deep subcutaneous layer (DCL) is biologically advantageous, with higher concentrations of mature adipocytes and stromal vascular components and less nonadipose debris. As a result, it is regarded as an optimal plane for grafting in many modern protocols.^8,9 Another advantage would be its proximity to muscles, directly improving circulatory supply for graft longevity while maintaining a safe distance from intramuscular structures, thereby reducing the risk of venous penetration and embolic events.^3,6,10 Indeed, minor DGF perforations are unlikely to allow intramuscular fat migration, making the procedure safe when subcutaneous placement is confirmed.^3,10 Hence, reliable access to this space is essential for maximizing BBL safety and aesthetic results.

HISTORICAL PERSPECTIVES AND SAFETY CHALLENGES

Originally, intramuscular injection was preferred for gluteal fat grafting, as muscle vascularity was thought to enhance fat integration and survival.⁶ However, cadaver studies revealed that fat injected intramuscularly may migrate toward gluteal veins, risking venous penetration and potentially fatal consequences such as pulmonary fat embolism (PFE).⁷ Historic survey data from the Aesthetic Surgery Education and Research Foundation Gluteal Fat Grafting Task Force highlighted BBL surgery as having the highest mortality rate of any aesthetic procedure, primarily due to PFE associated with deep or intramuscular injections.¹¹ In their 2017 report, Mofid et al¹¹ estimated a relatively elevated mortality risk and recommended avoiding intramuscular injection, using larger diameter cannulas, maintaining a superficial trajectory, and limiting fat placement to the subcutaneous compartment. Since then, complementary epidemiological data have been obtained from expert-opinion surveys of high-volume gluteal surgeons.^12,13 Again highlighting fat embolism as the most life-threatening complication, Cansancao et al¹² found nonfatal fat embolism in approximately 1 in 9530 BBL patients and predicted a mortality ratio of about 1 in 20,000. As further evidence, in a multidisciplinary series of gluteal fat grafting patients carried out by skilled surgeons, Agullo et al¹³ found that seroma was the most frequent adverse event, and PFE remained the most serious complication, albeit rare, with a death rate of about 1:23878.

Multisociety safety actions were spurred by these findings, in a follow-up survey of American Society for Aesthetic Plastic Surgery and International Society of Aesthetic Plastic Surgery members.¹⁴ Thus, global assessments by surgical societies, including the Multi-Society Task Force for Safety in Gluteal Fat Grafting (American Society for Aesthetic Plastic Surgery, American Society of Plastic Surgeons, International Society of Aesthetic Plastic Surgery, International Federation for Adipose Therapeutics and Science, and International Society of Plastic Regenerative Surgeons), established a worldwide consensus that suggested injecting fat exclusively into the subcutaneous plane and prohibiting intramuscular injections to reduce the risk of embolism.^1,4,5,14 Despite this decision being widely accepted, technique variation and patient anatomy made it difficult to reliably achieve subcutaneous-only injection.^2,3,6 This highlighted the need for real-time visualization, leading to ultrasound integration in BBL procedures.^2,14 On the other hand, the AI tools covered in this article are designed as interpretative and decision-support adjuncts built on top of ultrasound guiding, with the goal of automating tissue-layer detection, cannula tracking, and documentation.

REVOLUTIONIZING SAFETY AND AESTHETIC OUTCOMES: ULTRASOUND AND THE DEEP SUBCUTANEOUS LAYER

Ultrasound has completely revolutionized BBLs by facilitating real-time visualization, guaranteeing accurate subcutaneous implantation, and enhancing safety and aesthetic precision.² Indeed, through real-time observation, surgeons can accurately identify the DGF and safely guide the cannula inside the subcutaneous plane, minimizing overcorrection and muscle penetration risk.^3,5 Moreover, because the deep subcutaneous fat layer provides an optimal environment for fat survival, ultrasound-assisted fat grafting in this layer partakes in achieving even fat placement and volume retention, thus contributing to long-term patient satisfaction, while reducing the need for revisions.^5,6 Modern handheld ultrasound devices demonstrate clinical practicability by being portable and easily integrated into the operative field.^2,3,6 Additionally, intraoperative ultrasound offers a documentation advantage by enabling the surgeon to capture plane confirmation images or short videos during injection, which may serve both medicolegal and educational purposes. Nonetheless, despite its benefits, classical ultrasound requires extensive training and thorough anatomical knowledge.^6,15,16

REPORTED RESULTS IN THE LITERATURE

A 2025 systematic review by Reis et al⁴ found evidence supporting the safety and efficacy of ultrasound-guided BBL. Analyzing 4 studies encompassing 6235 female patients who underwent ultrasound-assisted BBL, the authors reported no cases of death or fat embolism. Minor complications were low (6.32%) and included fat necrosis (0.09%), infection (0.23%), and seroma (2.94%).⁴ Complementing these pooled data, Vidal-Laureano et al¹⁷ published a large single-center retrospective series of 1815 female patients undergoing ultrasound-guided gluteal fat transfer in which fat was confined to the subcutaneous plane. They observed an overall complication rate of 4%, mainly seroma (1.2%), local skin ischemia (1.2%), and surgical site infection (0.8%), with no reported macroscopic fat embolism and no mortality.¹⁷ Taken together, these contemporary findings imply that ultrasound-guided, subcutaneous-only BBL is associated with low rates of mostly minor complications and an absence of reported fat embolism–related deaths in the available series.^4,17 Additionally, patient-reported outcomes in these samples suggest that safety can be attained without compromising aesthetic appeal.^4,17 Indeed, 1 included study reported that more than 90% of patients were satisfied with their gluteal contour, and another documented a median BODY-Q score improvement from 48 preoperatively to 93 postoperatively, with a score of 100 out of 100 for buttock form and profile.^4,17 Additionally, patient satisfaction remained high: although 1 study showed that more than 90% of participants were pleased with their gluteal outline, another reported a median score rise from 48 to 93 postoperatively using BODY-Q, with a 100 out of 100 rating for buttock form and profile.^4,17 These findings contrast with a prior meta-analysis of nonultrasound BBLs, where accidental intramuscular injections caused fat embolisms and fatalities at rates reaching 0.12%.⁴ Although additional research on AI-assisted ultrasound is required, current evidence firmly supports subcutaneous-only, ultrasound-guided grafting for improved BBL results and safety.

AI: THE NEW FRONTIER IN ULTRASOUND-GUIDED BBL

AI is enhancing ultrasound-guided BBL by enabling real-time anatomical recognition, depth measurement, and visual feedback during fat grafting (Fig. 1). AI optimizes accuracy in identifying the deep subcutaneous plane by removing subjective interpretation, reducing the risk of intramuscular injection, and allowing a better aesthetic improvement. AI also aids training, guiding inexperienced surgeons, and standardizing techniques. Early adoption has boosted surgeon confidence, productivity, and minimized errors. With AI-powered mobile ultrasound platforms, the technology will be able to set new standards of safety and improve reproducibility in gluteal augmentation. These advantages support the growing shift toward structured, image-guided approaches and fit naturally within the evolving framework of safer, subcutaneous-only BBL techniques.

Fig. 1.

Real-time ultrasound visualization of gluteal tissue layers using AI analysis “BBL T mode,” highlighting the dermis, superficial subcutaneous layer, DCL, and muscle. The black arrow indicates the cannula precisely positioned within the DCL for safe fat graft placement.

AUTHORS’ PERSPECTIVE AND DETAILED OPERATIVE TECHNIQUE

In our experience, the mobile ultrasound device has been easy to use and allows us to securely save patient data. (See Video [online], which displays a comprehensive demonstration of mobile ultrasound guidance coupled with AI during BBL, including identification of safe subcutaneous fat planes, real-time cannula visualization, and avoidance of intramuscular injection.) When combined with AI, it provides clear, real-time guidance to precisely place fat in the DCL. Because this layer has fewer septa and a better blood supply, the fat survives longer, and our results stay consistent over time, following Coleman fat grafting principles. We have noticed smoother, more even contours with fewer bumps and much less fat necrosis, resulting in less discomfort, skin irritation, and superficial infections for our patients. Thanks to better fat spread in this layer, we also need less fat to achieve a natural shape—especially important for thinner patients with limited donor fat (Fig. 2).

Fig. 2.

Clinical outcomes following BBL using AI-assisted ultrasound-guided deep subcutaneous fat grafting. A, Preoperative view. B, Postoperative view. Postoperative results at 3 months demonstrate improved contour, volume, and symmetry with smooth transitions.

Video 1. This video provides a comprehensive demonstration of mobile ultrasound guidance coupled with artificial intelligence during Brazilian Butt Lift. It illustrates identification of safe subcutaneous fat injection planes, real-time visualization of cannula position, and confirmation of avoidance of intramuscular fat placement. The video highlights how ultrasound and AI integration enhances anatomical precision and procedural safety during gluteal fat grafting.

Download video file ^{(21.9MB, mp4)}

Since August 2024, we have incorporated the “Clarius L15 HD3” wireless mobile ultrasound, which includes built-in AI algorithms within its dedicated “BBL T mode,” into our workflow (Fig. 3). This modality automatically applies the AI-driven image processing specific to gluteal fat grafting, enhancing fascial contrast and optimizing the field for safe subcutaneous injection. During the procedure, patients are positioned prone with slight hip abduction, and preoperative markings guide the areas of projection and contour shaping. A high-frequency linear probe is first used to identify the superficial gluteal fascia, DCL, and DGF. AI-assisted real-time anatomy recognition enhances visualization of these structures by automatically adjusting depth, gain, and contrast, helping maintain constant identification of the safe injection plane. Fat is then injected through entry points at the lateral hip, upper intergluteal fold, or posterior superior iliac spine, using a 3- to 4-mm blunt, multiport cannula. Under continuous ultrasound, the cannula tip is advanced until clearly seen within the DCL, and AI-assisted cannula tracking accentuates the hyperechoic cannula tip, supporting accurate trajectory control. Injection is performed using low-pressure, retrograde deposition of small aliquots (1–2 mL per pass), ensuring even distribution and subcutaneous-only placement. Throughout the operation, risk-zone awareness is reinforced by maintaining continuous visualization of the DGF, and if the cannula approaches or indents this fascia, injection is paused and the trajectory corrected. When temporary edema or shadowing reduces clarity, the BBL T mode’s AI-optimized presets restore visualization before proceeding. Beyond safety, integrating AI improves standardization of the technique by reducing subjectivity in layer identification and creating more consistent imaging conditions across patients and operators. Key moments of the procedure—such as initial plane verification and final inspection—are recorded as clips or images, serving both medicolegal documentation and structured training material. This combined workflow of ultrasound guidance supported by an AI-enhanced, BBL-specific modality has allowed us to maintain precise, subcutaneous-only injections and improve operative confidence, especially in patients with challenging anatomy or limited soft-tissue thickness.

Fig. 3.

Clarius L15 HD3 wireless mobile ultrasound system used intraoperatively for real-time visualization of gluteal tissue layers during fat grafting.

AUTHORS’ CASE SERIES

Between August 2024 and July 2025, we evaluated the use of AI-assisted, ultrasound-guided fat grafting in an initial cohort of 34 female patients presenting for primary BBL in our private practice. After applying exclusion criteria—which removed patients with prior gluteal implants, recent trauma, active smoking within 4 weeks of surgery, or inability to complete routine follow-up—a total of 28 patients were included in the final analysis. The treated population reflected typical aesthetic practice demographics, with a fixed age range between 25 and 42 years, representing the most common age group seeking gluteal enhancement in our setting. Body mass index generally ranged between 22 and 29 kg/m², encompassing both leaner individuals with limited donor availability and fuller patients requesting more pronounced augmentation. All procedures were performed as part of routine clinical care, and all data reviewed for this analysis were fully anonymized. Clinical follow-up ranged from 3 to 6 months and included detailed examinations with targeted postoperative ultrasound assessments when indicated. Injected fat volumes varied naturally according to patient anatomy, donor-site availability, and aesthetic goals, rather than a uniform protocol. No major complications—such as fat embolism, deep infection, hematoma requiring surgery, or the need for reoperation—were observed in this series. Minor postoperative issues were infrequent and limited to transient edema or mild early contour unevenness in a small number of cases, all resolving spontaneously with conservative management. Early aesthetic outcomes were favorable, with smooth contours and stable projection throughout follow-up. Patient-reported outcome measures using the BODY-Q “Buttocks” module were collected from patients who completed the 3-month assessment, with most reporting meaningful postoperative improvement. Scores generally increased from preoperative ratings in the 40–55 range to postoperative ratings in the 80–95 range, reflecting increased satisfaction with buttock shape, symmetry, and overall appearance. In addition, subjective satisfaction scores remained high, with most patients rating their results between 8 and 10 out of 10. Overall, this early clinical experience suggests that integrating AI-enhanced mobile ultrasound into routine BBL practice is feasible and may help improve accuracy, consistency, and safety in subcutaneous fat grafting within a standard private-practice setting.

CONCLUSIONS

Mobile ultrasound guidance has significantly advanced the safety of gluteal fat grafting by allowing real-time visualization of tissue planes and ensuring consistent subcutaneous-only fat placement. The addition of AI provides further refinement by enhancing fascial differentiation, improving cannula-tip tracking, and supporting more standardized and reproducible interpretation of intraoperative ultrasound images. These combined technologies facilitate safer injections, particularly in patients with variable or challenging anatomy, while maintaining favorable aesthetic outcomes. Early clinical experience from our case series demonstrates that AI-assisted ultrasound can be readily integrated into routine private-practice BBL workflows, supporting accurate fat placement with a low complication profile and high patient satisfaction. As adoption increases, future research is needed to validate these early observations, quantify long-term outcomes, and determine how AI-enhanced imaging may continue to optimize safety and reliability in gluteal fat grafting.

DISCLOSURE

The authors have no financial interest to declare in relation to the content of this article.