Identifying pediatric heart murmurs and distinguishing innocent from pathologic using deep learning.

OBJECTIVE: To develop a deep learning algorithm to perform multi-class classification of normal pediatric heart sounds, innocent murmurs, and pathologic murmurs. METHODS: We prospectively enrolled children under age 18 being evaluated by the Division of Pediatric Cardiology. Parents provided consent for a deidentified recording of their child's heart sounds with a digital stethoscope. Innocent murmurs were validated by a pediatric cardiologist and pathologic murmurs were validated by echocardiogram. To augment our collection of normal heart sounds, we utilized a public database of pediatric heart sound recordings (Oliveira, 2022). We propose two novel approaches for this audio classification task. We train a vision transformer on either Markov transition field or Gramian angular field image representations of the frequency spectrum. We benchmark our results against a ResNet-50 CNN trained on spectrogram images. RESULTS: Our final dataset consisted of 366 normal heart sounds, 175 innocent murmurs, and 216 pathologic murmurs. Innocent murmurs collected include Still's murmur, venous hum, and flow murmurs. Pathologic murmurs included ventricular septal defect, tetralogy of Fallot, aortic regurgitation, aortic stenosis, pulmonary stenosis, mitral regurgitation and stenosis, and tricuspid regurgitation. We find that the Vision Transformer consistently outperforms the ResNet-50 on all three image representations, and that the Gramian angular field is the superior image representation for pediatric heart sounds. We calculated a one-vs-rest multi-class ROC curve for each of the three classes. Our best model achieves an area under the curve (AUC) value of 0.92 ± 0.05, 0.83 ± 0.04, and 0.88 ± 0.04 for identifying normal heart sounds, innocent murmurs, and pathologic murmurs, respectively. CONCLUSION: We present two novel methods for pediatric heart sound classification, which outperforms the current standard of using a convolutional neural network trained on spectrogram images. To our knowledge, we are the first to demonstrate multi-class classification of pediatric murmurs. Multiclass output affords a more explainable and interpretable model, which can facilitate further model improvement in the downstream model development cycle and enhance clinician trust and therefore adoption.

Infection After Lower-Limb Osseointegration: A Single-Center Retrospective Evaluation of Pathogens, Management, and Outcomes.

PURPOSE: Osseointegration (OI) is a novel alternative to traditional socket-suspended prostheses for lower-limb amputees, eliminating the socket-skin interface and allowing for weight bearing directly on the skeletal system. However, the stoma through which the implant attaches to the external prosthesis creates an ingress route for bacteria, and infection rates as high as 66% have been reported. The aims of this study are to classify infection management and long-term outcomes in this patient population to maximize implant salvage. METHODS: An institutional review board-approved retrospective analysis was performed on all patients who underwent lower-limb OI at our institution between 2017 and 2022. Demographic, operative, and outcome data were collected for all patients. Patients were stratified by the presence and severity of infection. Chi-square and t tests were performed on categorical and continuous data, respectively, using an alpha of 0.05. RESULTS: One hundred two patients met our study criteria; 62 had transfemoral OI and 40 had transtibial OI. Patients were followed for 23.8 months on average (range, 3.5-63.7). Osteomyelitis was more likely than soft tissue infection to be polymicrobial in nature (71% vs 23%, P < 0.05). Infections at the stoma were mostly (96%) managed with oral antibiotics alone, whereas deeper soft tissue infections also required intravenous antibiotics (75%) or operative washout (19%). Osteomyelitis was managed with intravenous antibiotics and required operative attention; 5 (71%) underwent washout and 2 (29%) underwent explantation. Both implants were replaced an average of 3.5 months after explantation. There was no correlation between history of soft tissue infection and development of osteomyelitis (P > 0.05). The overall implant salvage rate after infection was 96%. CONCLUSIONS: This study describes our institution's experience managing infection after OI and soft tissue reconstruction. Although infections do occur, they are easily treatable and rarely require operative intervention. Explantation due to infection is rare and can be followed up with reimplantation, reaffirming that OI is a safe and effective treatment modality.

Comparative Analysis of Recurrence-Free Survival and Mortality in Postmastectomy Patients: Insights From Multidecade, Multiinstitutional Propensity Score Matching.

INTRODUCTION: Disparities in postmastectomy reconstructive care are widely acknowledged. However, there is limited understanding regarding the impact of reconstructive services on cancer recurrence and breast cancer-related mortality. Therefore, this study aims to examine how patient-specific factors and breast reconstruction status influence recurrence-free survival and mortality rates in breast cancer patients. METHODS: Retrospective chart review was performed to collect data on patients who underwent mastectomy at 2 institutions within the New York-Presbyterian system from 1979 to 2019. Sociodemographic information, medical history, and the treatment approach were recorded. Propensity score matching, logistic regression, unpaired t test, and chi-square test were used for statistical analysis. RESULTS: Overall, cancer recurrence occurred in 6.62% (317) of patients, with 16.8% (803) overall mortality rate. For patients who had relapsed disease, completion of the reconstruction sequence was correlated with an earlier detection of cancer recurrence and improved survival odds (P < 0.05). Stratified analysis of the reconstruction group alone showed mortality benefit among patients who underwent free flap procedures (P < 0.05). CONCLUSION: Patients undergoing breast reconstruction after mastectomy are likely to have better access to follow-up care and improved interfacing with the healthcare system. This may increase the speed at which cancer recurrence is detected. This study highlights the need for consistent plastic surgery referral and continued monitoring by all members of the breast cancer care team for cancer recurrence among patients.

Assessing Long-Term Volume Retention in Breast Fat Grafting: A Comparative Study of Lipoaspirate Processing Techniques.

INTRODUCTION: Autologous fat grafting is a method of improving aesthetic outcomes after both breast reconstruction and aesthetic surgery through volume enhancement and tissue contouring. Long-lasting effects are linked to greater patient satisfaction and more optimal augmentation results. Harvesting, processing, and injection techniques may all affect the longevity of deformity filling. Our objective is to evaluate the effect of lipoaspirate processing modality on longitudinal volume retention after surgery. METHODS: A prospective, single-institution, randomized control trial placed consented postmastectomy fat grafting patients into 1 of 3 treatment arms (active filtration, low-pressure decantation, and standard decantation) in a 1:1:1 ratio. A preoperative 3-dimensional scan of the upper torso was taken as baseline. At the 3-month postoperative visit, another 3D scan was taken. Audodesk Meshmixer was used to evaluate the volume change. RESULTS: The volume of fat injected during the initial procedure did not differ significantly between the treatment arms (P > 0.05). Both active filtration and low-pressure decantation resulted in higher percentage volume retention than traditional decantation (P < 0.05). Active filtration and low-pressure decantation exhibited comparable degrees of fat maintenance at 3 months (P > 0.05). DISCUSSION: Compared with using traditional decantation as the lipoaspirate purification technique, active filtration and low-pressure decantation may have led to higher levels of cell viability by way of reduced cellular debris and other inflammatory components that may contribute to tissue resorption and necrosis. Further immunohistochemistry studies are needed to examine whether active filtration and low-pressure decantation lead to lipoaspirates with more concentrated viable adipocytes, progenitor cells, and factors for angiogenesis.

An Evolving Landscape: Return of Breast Sensation After Mastectomy Varies by Anatomic Region and Reconstructive Method.

PURPOSE: Loss of breast sensation after mastectomy has been well documented. Postoperative reinnervation of the breast is influenced by factors including reconstructive technique, patient comorbidities, and adjuvant treatment. However, little attention has been paid to the differences in sensation across regions of the breast and the impact of reconstructive method on these regional differences over time. METHODS: Patients undergoing nipple-sparing mastectomy with immediate autologous or alloplastic reconstruction were prospectively followed. Neurosensory testing was performed in 9 breast regions using a pressure-specified sensory device. Patients were stratified by reconstructive technique, and regional sensation was compared at different preoperative and postoperative time points using Student t tests. RESULTS: One hundred ninety-two patients were included; 106 underwent autologous reconstruction via neurotized deep inferior epigastric artery perforator flap, and 86 underwent 2-stage alloplastic reconstruction. Preoperative sensation thresholds did not differ between reconstructive cohorts in any region and averaged 18.1 g/mm2. In the first year after mastectomy, decreased sensation was most pronounced in the inner breast regions and at the nipple areolar complex (NAC) in both reconstructive cohorts. At 4 years postoperatively, sensation increased the most at the NAC in the alloplastic cohort (34.0 g/mm2 decrease) and at the outer lateral region in the autologous cohort (30.4 g/mm2 threshold decrease). The autologous cohort experienced improved sensation compared with the alloplastic cohort in 5 of 9 regions at 1 year postoperatively, and in 7 of 9 regions at 4 years postoperatively; notably, only sensation at the outer superior and outer medial regions did not differ significantly between cohorts at 4 years postoperatively. CONCLUSIONS: Although patients undergoing breast reconstruction experience increased breast sensation over time, the return of sensation is influenced by type of reconstruction and anatomic region. Regions closer to and at the NAC experience the greatest loss of sensation after mastectomy, although the NAC itself undergoes the most sensation recovery of any breast region in those with alloplastic reconstruction.Autologous reconstruction via a neurotized deep inferior epigastric artery perforator flap results in increased return of sensation compared with alloplastic reconstruction, particularly in the inferior and lateral quadrants of the breast.

Impact of Comorbidities on Sensory Return After Breast Reconstruction.

PURPOSE: Breast anesthesia after mastectomy and reconstruction has been an ongoing concern with few improvements made in recent years. At present, there is a lack of studies evaluating the impact of comorbidities on sensation restoration. Identifying risk factors (RF) will be helpful with preoperative counseling. METHODS: This was a prospective study on patients who underwent mastectomy and immediate implant-based or neurotized deep inferior epigastric perforator (DIEP) flap-based reconstruction. Neurosensory testing was performed at predefined time points using a pressure specified device. Patients were stratified based on reconstruction type and comorbidities, including obesity (≥30 kg/m2), age (>55 years), hypertension, alcohol use, and smoking status. Sensory comparisons among the comorbidity groups were conducted using unpaired 2-sample t tests. RESULTS: A total of 239 patients were included in this study with 109 patients in the implant cohort and 131 patients in the DIEP cohort. One patient underwent bilateral reconstruction using both reconstructive modalities. Preoperatively, age older than 55 years was identified as an RF for reduced breast sensation in the implant cohort (difference in threshold, 10.7 g/mm2), whereas obesity was identified as an RF in the DIEP cohort (difference in threshold, 8 g/mm2). During the first 2 years postreconstruction, age older than 55 years and tobacco use history were found to be negatively correlated with breast sensation for both cohorts. With DIEP reconstruction specifically, obesity was identified as an additional RF during the early postoperative period. Of note, none of the comorbidities were found to be long-term RFs for reduced breast sensitivity. All breast sensation levels returned to comparable levels across all comorbidities by 4 years postreconstruction. CONCLUSIONS: Currently, various comorbidities have been recognized as RFs for several postoperative complications including extended postoperative stay, necrosis, infection, and reoperation. However, our findings suggest that, although age, smoking history, and obesity showed transient associations with reduced breast sensation during the initial years postreconstruction, they play no role in the long-term potential of sensory nerve regeneration.

Proliferation Patterns of MCF-7 Breast Cancer Cells in Lipoaspirate Conditioned Media.

INTRODUCTION: Autologous fat grafting (AFG) is a common technique used to enhance aesthetic outcomes in postmastectomy breast reconstruction patients. Adipokines are hormones secreted by adipose tissue that play a critical role in regulating metabolic processes and the immune system. However, dysregulated adipokine secretion and signaling can contribute to the development and progression of cancer by promoting angiogenesis, altering the immune response, and inducing the epithelial mesenchymal transition. We aimed to assess how breast cancer cells behave in conditioned media derived from fat grafting lipoaspirates and gain a better understanding of the potential interactions that may occur within the tumor microenvironment. METHODS: Patients who were undergoing AFG as a part of breast reconstruction at NY-Presbyterian/Weill Cornell Medical Center between March 2021 and July 2023 were consented and enrolled in the study. This study was approved by the Weill Cornell Medicine Institutional Review Board (#20-10022850-14). Conditioned media is created using 20% of patient lipoaspirate secretome and 80% starving media. The growth of MCF-7, a human ER/PR+ breast cancer cell line, in conditioned media is assessed using CyQUANT. RESULTS: The breast cancer cells incubated in conditioned media displayed similar growth trends as those in complete media, which is enriched for cell growth (P > 0.05). MCF-7 cell behavior in conditioned media differed significantly from their proliferation patterns when serum starved in 100% starving media (P < 0.05). DISCUSSION: Our results suggest that there may be inherent factors within the lipoaspirate that may promote MCF-7 proliferation. One potential implication is that AFG used for breast reconstruction should be delayed until local-regional disease control has been established. In addition, based on the in vitro proliferation patterns of breast cancer cells in conditioned media, the safety profile of AFG may be enhanced if the procedure is performed after attaining negative margins and the completion breast cancer treatment.

Using a Machine Learning Approach to Predict the Need for Elective Revision and Unplanned Surgery after Implant-based Breast Reconstruction.

Two-stage implant-based reconstruction after mastectomy may require secondary revision procedures to treat complications, correct defects, and improve aesthetic outcomes. Patients should be counseled on the possibility of additional procedures during the initial visit, but the likelihood of requiring another procedure is dependent on many patient- and surgeon-specific factors. This study aims to identify patient-specific factors and surgical techniques associated with higher rates of secondary procedures and offer a machine learning model to compute individualized assessments for preoperative counseling. A training set of 209 patients (406 breasts) who underwent two-stage alloplastic reconstruction was created, with 45.57% of breasts (185 of 406) requiring revisional or unplanned surgery. On multivariate analysis, hypertension, no tobacco use, and textured expander use corresponded to lower odds of additional surgery. In contrast, higher initial tissue expander volume, vertical radial incision, and larger nipple-inframammary fold distance conferred higher odds of additional surgery. The neural network model trained on clinically significant variables achieved the highest collective performance metrics, with ROC AUC of 0.74, sensitivity of 84.2, specificity of 63.6, and accuracy of 62.1. The proposed machine learning model trained on a single surgeon's data offers a precise and reliable tool to assess an individual patient's risk of secondary procedures. Machine learning models enable physicians to tailor surgical planning and empower patients to make informed decisions aligned with their lifestyle and preferences. The utilization of this technology is especially applicable to plastic surgery, where outcomes are subject to a variety of patient-specific factors and surgeon practices, including threshold to perform secondary procedures.

A Machine Learning Approach to Predicting Donor Site Complications Following DIEP Flap Harvest.

BACKGROUND: The additional donor site incisions in autologous breast reconstruction can predispose to abdominal complications. The purpose of this study is to delineate predictors of donor site morbidity following deep inferior epigastric perforator (DIEP) flap harvest and use those predictors to develop a machine learning model that can identify high-risk patients. METHODS: This is a retrospective study of women who underwent DIEP flap reconstruction from 2011 to 2020. Donor site complications included abdominal wound dehiscence, necrosis, infection, seroma, hematoma, and hernia within 90 days postoperatively. Multivariate regression analysis was used to identify predictors for donor site complications. Variables found significant were used to construct machine learning models to predict donor site complications. RESULTS: Of 258 patients, 39 patients (15%) developed abdominal donor site complications, which included 19 cases of dehiscence, 12 cases of partial necrosis, 27 cases of infection, and 6 cases of seroma. On univariate regression analysis, age (p = 0.026), body mass index (p = 0.003), mean flap weight (p = 0.006), and surgery time (p = 0.035) were predictors of donor site complications. On multivariate regression analysis, age (p = 0.025), body mass index (p = 0.010), and surgery duration (p = 0.048) remained significant. Radiographic features of obesity, such as abdominal wall thickness and total fascial diastasis, were not significant predictors of complications (p > 0.05). In our machine learning algorithm, the logistic regression model was the most accurate at predicting donor site complications with the accuracy of 82%, specificity of 0.93, and negative predictive value of 0.87. CONCLUSION: This study demonstrates that body mass index is superior to radiographic features of obesity in predicting donor site complications following DIEP flap harvest. Other predictors include older age and longer surgery duration. Our logistic regression machine learning model has the potential to quantify the risk of donor site complications.

Targeted Nipple Reinnervation in Gender-affirming Mastectomy Using Autologous Nerve Graft.

Our team recently described targeted nipple reinnervation (TNR) during female-to-male gender-affirming mastectomy with free nipple grafting using either direct nerve coaptation or nerve allograft. The goals of TNR are to improve sensation (including erogenous sensation) and prevent numbness, paresthesias, chronic pain, and phantom sensation. Here, we describe our modified technique, which has evolved to use autologous intercostal nerve branches as donor nerves for reinnervation if direct nerve coaptation cannot be achieved. During TNR, the T3-T5 sensory branches are preserved and coapted to the repositioned nipple-areolar complex (NAC). In patients with donor nerves that were not adequate in length to allow for direct coaptation, autologous intercostal nerve branches were not used for coaptation (branches present along the chest wall that would otherwise be lost) or one of the T3-T5 branches were harvested. An end-to-end nerve repair between the autograft and donor nerves was done, and the donor nerve/autograft complex was coapted to the NAC. Targeted muscle reinnervation was performed after autograft harvest to prevent neuroma formation. TNR with intercostal nerve autograft is technically feasible in female-to-male gender-affirming mastectomy with free nipple grafting when direct coaptation is not possible. Chest reinnervation using autologous intercostal nerve branches as donor nerves is another option for reinnervation when the nerves are too short for direct coaptation. Because the collection of long-term data is ongoing, the effectiveness of NAC reinnervation using our technique will be described in a future publication.

Machine-Learning Prediction of Capsular Contraction after Two-Stage Breast Reconstruction.

Background: Two-stage breast reconstruction is a common technique used to restore preoperative appearance in patients undergoing mastectomy. However, capsular contracture may develop and lead to implant failure and significant morbidity. The objective of this study is to build a machine-learning model that can determine the risk of developing contracture formation after two-stage breast reconstruction. Methods: A total of 209 women (406 samples) were included in the study cohort. Patient characteristics that were readily accessible at the preoperative visit and details pertaining to the surgical approach were used as input data for the machine-learning model. Supervised learning models were assessed using 5-fold cross validation. A neural network model is also evaluated using a 0.8/0.1/0.1 train/validate/test split. Results: Among the subjects, 144 (35.47%) developed capsular contracture. Older age, smaller nipple-inframammary fold distance, retropectoral implant placement, synthetic mesh usage, and postoperative radiation increased the odds of capsular contracture (p < 0.05). The neural network achieved the best performance metrics among the models tested, with a test accuracy of 0.82 and area under receiver operative curve of 0.79. Conclusion: To our knowledge, this is the first study that uses a neural network to predict the development of capsular contraction after two-stage implant-based reconstruction. At the preoperative visit, surgeons may counsel high-risk patients on the potential need for further revisions or guide them toward autologous reconstruction.

Deep learning analysis of blood flow sounds to detect arteriovenous fistula stenosis.

For hemodialysis patients, arteriovenous fistula (AVF) patency determines whether adequate hemofiltration can be achieved, and directly influences clinical outcomes. Here, we report the development and performance of a deep learning model for automated AVF stenosis screening based on the sound of AVF blood flow using supervised learning with data validated by ultrasound. We demonstrate the importance of contextualizing the sound with location metadata as the characteristics of the blood flow sound varies significantly along the AVF. We found the best model to be a vision transformer trained on spectrogram images. Our model can screen for stenosis at a performance level comparable to that of a nephrologist performing a physical exam, but with the advantage of being automated and scalable. In a high-volume, resource-limited clinical setting, automated AVF stenosis screening can help ensure patient safety via early detection of at-risk vascular access, streamline the dialysis workflow, and serve as a patient-facing tool to allow for at-home, self-screening.

Diagnostic Scrutiny and Patterns of Elevated Cancer Risk: Uncovering Overdiagnosis Through Standardized Incidence Ratios.

Certain medical diagnoses and environmental or occupational exposures may be associated with elevated risk of cancer diagnosis, either through causal mechanisms or via increased detection of a subclinical reservoir through increased diagnostic scrutiny (overdiagnosis). The present study aimed to investigate the distribution of elevated cancer risks associated with different diagnoses and exposures. A systematic literature search was conducted to identify studies published in the last 30 years that examined the standardized incidence ratio (SIR) associated with exposures and risk factors. Meta-SIRs for each cancer type were calculated. The distribution of elevated cancer risks was then compared between cancer types previously reported to be susceptible to overdiagnosis and those that have not been associated with overdiagnosis. The review of 108 studies identified four patterns: SIR generally elevated for 1) only overdiagnosis-susceptible cancer types, 2) both overdiagnosed and non-overdiagnosed cancer types, 3) select cancers in accordance with risk factor or exposure, and 4) SIRs that did not exhibit a distinct increase in any cancer type. The distribution of elevated cancer risks may serve as a signature of whether the underlying risk factor or exposure is a carcinogenic process or a mechanism of increased diagnostic scrutiny uncovering clinically occult diseases. The identification of increased cancer risk should be viewed with caution, and analyzing the pattern of elevated cancer risk distribution can potentially reveal conditions that appear to be cancer risk factors but are in fact the result of exposure to medical surveillance or other healthcare activities that lead to the detection of indolent tumors.

Return of the Flap; The Empire State Mandate: A Multi-decade Multi-institutional Analysis of the 2010 New York State Legislature's Impact on Postmastectomy Immediate Flap-Based Reconstruction.

PURPOSE: Immediate postmastectomy breast reconstruction plays an integral role in patient care because of its psychosocial benefits. New York State (NYS) passed the 2010 Breast Cancer Provider Discussion Law with the aim of increasing patient awareness of reconstructive options through mandating plastic surgery referral at the time of cancer diagnosis. Short-term analysis of the years surrounding implementation suggests the law increased access to reconstruction, especially for certain minority groups. However, given the continued presence of disparities in access to autologous reconstruction, we aimed to investigate the longitudinal effects of the bill on access to autologous reconstruction along various sociodemographic cohorts. METHODS: Retrospective review identified demographic, socioeconomic, and clinical data for patients undergoing mastectomy with immediate reconstruction at Weill Cornell Medicine and Columbia University Irving Medical Center from 2002 to 2019. Primary outcome was receiving implant or autologous-based reconstruction. Subgroup analysis was based on sociodemographic factors. Multivariate logistic regression identified predictors of autologous reconstruction. Interrupted time series modeling analyzed differences in reconstructive trends for subgroups before and after the 2011 implementation of the NYS law. RESULTS: We included 3178 patients; 2418 (76.1%) and 760 (23.9%) patients underwent implant and autologous-based reconstruction, respectively. Multivariate analysis indicated that race, Hispanic status, and income were not predictors of autologous reconstruction. Interrupted time series showed that with each year leading up to 2011 implementation, patients were 19% less likely to receive autologous-based reconstruction. Following implementation, there was a 34% increase in the odds of receiving autologous-based reconstruction with each passing year. Following implementation, Asian American and Pacific Islander patients experienced a 55% greater increase in the rate of flap reconstruction than White patients. Following implementation, the highest-income quartile experienced a 26% greater increase in the rate of autologous-based reconstruction compared with the lowest-income quartile. After implementation, Hispanic patients experienced a 30% greater decrease in the rate of autologous-based reconstruction compared with non-Hispanic patients. CONCLUSIONS: Our data indicate the long-term efficacy of the NYS Breast Cancer Provider Discussion Law in increasing access to autologous-based reconstruction, especially for certain minority groups. These findings underscore the importance of this bill and encourage its adoption into other states.

A Comparative Analysis of Sensory Return in Delayed-Immediate Versus Immediate Neurotized Deep Inferior Epigastric Perforator Flap Breast Reconstruction.

BACKGROUND: Breast reconstruction patients who anticipate adjuvant radiation are not suitable candidates for immediate deep inferior epigastric perforator (DIEP) flap reconstruction due to the risk of flap fibrosis, shrinkage, and fat necrosis. Rather, many of these patients undergo delayed-immediate, or "babysitter," reconstruction, where a tissue expander is placed first as a temporizing measure during adjuvant therapy before definitive flap reconstruction. In this study, we aim to compare sensory changes in delayed-immediate to immediate DIEP flap patients. METHODS: Ninety-one patients, including 26 patients (46 breasts) with "babysitter" procedures and 65 patients (120 breasts) with immediate DIEP flaps, were prospectively identified at their preoperative visit. For both cohorts, baseline level (t = 0) is defined as before mastectomy. RESULTS: "Babysitter" patients underwent final-stage neurotized flap reconstruction on average at 12 months after initial tissue expander placement (range, 3-18 months). At 18 month after mastectomy (6 months after DIEP), delayed-immediate patients had comparable sensitivity measurements as immediate DIEP flap patients in all regions of the breast (P > 0.05). For delayed immediate patients, at 18 months postoperatively, sensitivity measurements were comparable with baseline levels only in the outer superior, outer medial, and outer lateral regions of the breast (P > 0.05). At 24 months postoperatively, cutaneous thresholds were comparable with baseline in all regions of the breast except the inner inferior region (P > 0.05), following a similar sensory recovery trajectory as immediate DIEP flap patients. CONCLUSIONS: In patients who undergo "babysitter" procedures, the combination of sensory return from the native mastectomy skin flap along with the neurotized DIEP flap yields sensory recovery comparable with immediate DIEP flap patients after definitive flap reconstruction. When final-stage flap reconstruction occurs by 12 months after mastectomy, sensation can return beginning 24 months postoperatively, or even sooner in some regions of the breast.

An Evaluation of Native Breast Dimension and Tissue Expander Inflation Rate on the Risk of Capsular Contracture Development in Postmastectomy Reconstruction.

INTRODUCTION: Capsular contracture is a common complication after 2-stage breast reconstruction. The relationships between native breast size, the rate of tissue expander expansion, and capsule formation have not been elucidated. This study aims to evaluate how these factors contribute to capsular contracture and establish cutoff values for increased risk. METHODS: A data set consisting of 229 patients who underwent 2-stage breast reconstruction between 2012 and 2021 was included in the study. The rate of expansion is estimated as the final expanded volume subtracted by the initial filling volume of the tissue expander over time elapsed. The native breast size was estimated using various preoperative breast measurements and the weight of mastectomy specimen (grams). Further stratified analysis evaluated patients separately based on postoperative radiation status. RESULTS: Greater nipple-inframammary fold distance and faster tissue expander enlargement rate conferred decreased odds of developing capsular contracture ( P < 0.05). On stratified analysis, faster tissue expansion rate was not significant in the nonradiated cohort but remained a significant negative predictor in the radiation group (odds ratio, 0.996; P < 0.05). Cut-point analysis showed an expansion rate of <240 mL/mo and a nipple-inframammary fold value of <10.5 cm as conferring a greater risk of capsular contracture. CONCLUSION: Smaller inframammary fold distance may be associated with a higher risk of capsular contracture. Slower expansion rates correlate with increased odds of contracture in patients undergoing adjuvant radiation. Breast geometry should be considered when risk stratifying various reconstruction approaches (implant vs autologous). In addition, longer delays between implant exchange and initial tissue expansion should be avoided if clinically feasible.

Timeline and Incidence of Postoperative Complications in Prepectoral, Dual-Plane, and Total Submuscular Alloplastic Reconstruction With and Without Biosynthetic Scaffold Usage.

INTRODUCTION: Acellular dermal matrices and synthetic meshes are commonly used to improve inframammary-fold definition, minimize muscle excision, and allow for greater control over the surgical technique in implant-based breast reconstruction. The aims of this study are to compare various combinations of placement planes and biosynthetic scaffolds and to further examine the respective incidences of postoperative complications and the timeline of capsular contracture development. METHODS: A data set consisting of 220 patients (393 samples) who underwent 2-stage reconstruction between 2012 and 2021 was used in the study. χ 2 , Fisher exact test, and 1-way analysis of variance were used to identify significant differences between the 4 subgroups. Cox proportional-hazards model and Kaplan-Meier estimator were used for survival analysis. RESULTS: On univariate logistic regression (odds ratio, 0.21; P = 0.005), survival analysis ( P = 0.0082), and Cox-proportional hazard model (hazard ratio, 1.6; P = 0.01), poly-4-hydroxybutyrate mesh usage was linked to an increased risk of capsular contracture development. Prepectoral placement with no mesh and dual-plane placement with acellular dermal matrix showed similar timelines of capsular contracture development. The lowest incidences of capsular contracture occurred in the prepectoral placement and no mesh (49/161, 30.4%) and total submuscular subgroups (3/14, 21.4%). Infection, necrosis, and revision surgery rates did not differ significantly between the 4 groups. CONCLUSIONS: The use of poly-4-hydroxybutyrate mesh in 2-stage breast reconstruction is correlated with a statistically significant increase in capsular contracture. Prepectoral placement with no biosynthetic scaffold had one of the lowest rates of contracture and may provide the most optimal balance between economic and clinical considerations in implant-based reconstruction.

A Pilot Study Comparing Sensation in Buried Versus Nonburied Deep Inferior Epigastric Perforator Flaps.

BACKGROUND: Neurotized deep inferior epigastic perforator (DIEP) flaps have been shown to improve sensory recovery after mastectomy and reconstruction. With the recent trend toward nipple-sparing mastectomies, sensation likely originates within the buried DIEP flap and then innervates the breast skin. In contrast, for patients undergoing skin-sparing mastectomies, the DIEP flap skin is preserved, brought up to the surface, and directly innervated. In this study, we aim to evaluate inner breast region sensation between patients whose DIEP flap is buried and whose DIEP flap skin is brought to the surface. METHODS: Seventy patients who underwent mastectomy with immediate reconstruction using the DIEP flap were prospectively identified. Of these, 60 patients underwent nipple-sparing mastectomy with buried DIEP flap reconstruction while 10 patients underwent skin-sparing mastectomy with nonburied DIEP flap reconstruction. Patients in both cohorts received nerve grafting using the 70 × 1-2-mm Avance Nerve Graft in identical fashion. Sensitivity evaluation was performed in five inner breast regions (corresponding to the nonburied DIEP flap area). RESULTS: In the buried DIEP cohort, at 6 months postoperatively, there was a statistically significant difference in inner breast region sensitivity measurements compared with baseline levels ( P < 0.001). In contrast, in the nonburied DIEP cohort, at 6 months postoperatively, sensation in the inner breast region was comparable with preoperative baseline levels ( P = 0.236). At 24 months postoperatively, inner breast region sensitivity measurements in both cohorts were comparable with preoperative baseline measurements ( P > 0.05). CONCLUSIONS: Neurotized DIEP flap skin raised directly to the surface confers earlier sensory recovery than buried DIEP flaps. In patients who undergo skin-sparing mastectomies with nonburied DIEP flap reconstruction, they can expect significantly better sensation in the inner regions of the breast at 6 months postoperatively. In patients who undergo nipple-sparing mastectomies with buried DIEP flap reconstruction, they can expect sensation in the inner breast to return to preoperative baseline levels at a later time point-beginning as early as 24 months postoperatively.

Osseointegration for Lower Limb Amputation: Understanding the Risk Factors and Time Courses of Soft Tissue Complications.

PURPOSE: Lower-limb osseointegrated prostheses are a novel alternative to traditional socket-suspended prostheses, which are often associated with poor fit, soft tissue damage, and pain. Osseointegration eliminates the socket-skin interface and allows for weight-bearing directly on the skeletal system. However, these prostheses can also be complicated by postoperative issues that can negatively impact mobility and quality of life. Little is known about the incidence of or risk factors for these complications as few centers currently perform the procedure. METHODS: A retrospective analysis was performed on all patients who underwent single-stage lower limb osseointegration at our institution between 2017 and 2021. Patient demographics, medical history, operative data, and outcomes were collected. Fisher exact test and unpaired t tests were performed to identify risk factors for each adverse outcome, and time-to-event survival curves were generated. RESULTS: Sixty patients met our study criteria: 42 males and 18 females with 35 transfemoral and 25 transtibial amputations. The cohort had an average age of 48 years (range, 25-70 years) and follow-up period of 22 months (range, 6-47 months). Indications for amputation were trauma (50), prior surgical complication (5), cancer (4), and infection (1). Postoperatively, 25 patients developed soft tissue infections, 5 developed osteomyelitis, 6 had symptomatic neuromas, and 7 required soft tissue revisions. Soft tissue infections were positively correlated with obesity and female sex. Neuroma development was associated with increased age at osseointegration. Neuromas and osteomyelitis were both associated with decreased center experience. Subgroup analysis by amputation etiology and anatomic location did not show significant differences in outcomes. Notably, hypertension (15), tobacco use (27), and prior site infection (23) did not correlate with worse outcomes. Forty-seven percent of soft tissue infections occurred in the 1 month after implantation, and 76% occurred in the first 4 months. CONCLUSIONS: These data provide preliminary insights into risk factors for postoperative complications arising from lower limb osseointegration. These factors are both modifiable (body mass index, center experience), and unmodifiable (sex, age). As this procedure continues to expand in popularity, such results are necessary to inform best practice guidelines and optimize outcomes. Further prospective studies are needed to confirm the above trends.

On the analysis of data augmentation methods for spectral imaged based heart sound classification using convolutional neural networks.

BACKGROUND: The application of machine learning to cardiac auscultation has the potential to improve the accuracy and efficiency of both routine and point-of-care screenings. The use of convolutional neural networks (CNN) on heart sound spectrograms in particular has defined state-of-the-art performance. However, the relative paucity of patient data remains a significant barrier to creating models that can adapt to a wide range of potential variability. To that end, we examined a CNN model's performance on automated heart sound classification, before and after various forms of data augmentation, and aimed to identify the most optimal augmentation methods for cardiac spectrogram analysis. RESULTS: We built a standard CNN model to classify cardiac sound recordings as either normal or abnormal. The baseline control model achieved a PR AUC of 0.763 ± 0.047. Among the single data augmentation techniques explored, horizontal flipping of the spectrogram image improved the model performance the most, with a PR AUC of 0.819 ± 0.044. Principal component analysis color augmentation (PCA) and perturbations of saturation-value (SV) of the hue-saturation-value (HSV) color scale achieved a PR AUC of 0.779 ± 045 and 0.784 ± 0.037, respectively. Time and frequency masking resulted in a PR AUC of 0.772 ± 0.050. Pitch shifting, time stretching and compressing, noise injection, vertical flipping, and applying random color filters negatively impacted model performance. Concatenating the best performing data augmentation technique (horizontal flip) with PCA and SV perturbations improved model performance. CONCLUSION: Data augmentation can improve classification accuracy by expanding and diversifying the dataset, which protects against overfitting to random variance. However, data augmentation is necessarily domain specific. For example, methods like noise injection have found success in other areas of automated sound classification, but in the context of cardiac sound analysis, noise injection can mimic the presence of murmurs and worsen model performance. Thus, care should be taken to ensure clinically appropriate forms of data augmentation to avoid negatively impacting model performance.