Inhalation Injury Grading Using Transfer Learning Based on Bronchoscopy Images and Mechanical Ventilation Period.

The abbreviated injury score (AIS) is commonly used as a grading system for inhalation injuries. While inhalation injury grades have inconsistently been shown to correlate positively with the time mechanical ventilation is needed, grading is subjective and relies heavily on the clinicians' experience and expertise. Additionally, no correlation has been shown between these patients' inhalation injury grades and outcomes. In this paper, we propose a novel inhalation injury grading method which uses deep learning algorithms in bronchoscopy images to determine the injury grade from the carbonaceous deposits, blistering, and fibrin casts in the bronchoscopy images. The proposed method adopts transfer learning and data augmentation concepts to enhance the accuracy performance to avoid overfitting. We tested our proposed model on the bronchoscopy images acquired from eighteen patients who had suffered inhalation injuries, with the degree of severity 1, 2, 3, 4, 5, or 6. As performance metrics, we consider accuracy, sensitivity, specificity, F-1 score, and precision. Experimental results show that our proposed method, with both transfer learning and data augmentation components, provides an overall 86.11% accuracy. Moreover, the experimental results also show that the performance of the proposed method outperforms the method without transfer learning or data augmentation.

Racial and geographic disparities in reconstructive procedures following melanoma resection.

INTRODUCTION: Melanoma occurs most commonly in non-Hispanic White patients; however, Black and Hispanic patients experience greater morbidity and mortality. This study assesses how race and socioeconomic factors influence rates of reconstructive procedures and hospital-based outcomes in melanoma patients. METHODS: Data were extracted from the National Inpatient Sample database from the years 2010-2015. Patients with melanoma who underwent a reconstructive procedure were identified. Univariate and multivariate logistic regression analysis was used to identify the relationship between dependent variables and various patient/hospital components for patients undergoing reconstructive procedures. RESULTS: Black and Hispanic patients had a greater length of stay (LOS) than non-Hispanic White patients (OR: 2.252, p = 0.0307, and OR: 2.592, p = 0.0014), and Hispanic patients were less likely to receive more complex reconstructive procedures (OR: 0.449, p = 0.0487). Patients living in rural areas were less likely to receive complex reconstructive procedures than those in both urban teaching and non-teaching hospitals (OR: 3.313, p = 0.0135, and OR: 3.505, p = 0.0074). Pedicled or rotational flaps were less likely to be performed at medium- or large-sized hospitals (OR: 0.610, p = 0.0296, and OR: 0.496, p = 0.0002). CONCLUSION: Race and socioeconomic factors are important predictors of access to complex reconstructive procedures and hospital-based outcomes following extirpation in melanoma patients.

Limited Access to Facial Feminization Geographically Despite Nationwide Expansion of Other Gender-Affirming Surgeries.

Facial feminization surgery (FFS) is an integral aspect of gender-affirming surgery (GAS) for individuals seeking to align secondary sex characteristics and gender identity. Despite the importance of FFS in treating gender dysphoria, current trends and prevalence remain unknown. We sought to examine trends in GAS and FFS and investigate the payer status of facial feminization procedures in the United States. Methods: Data was extracted from the National Inpatient Sample from 2008 to 2017 by using International Classification of Diseases Ninth or Tenth diagnosis codes for gender identity disorder and procedure codes for FFS. Results: From 2008 to 2017, 3015 patients underwent GAS. The yearly number of cases increased as did the average cost of GAS, which rose from $13,657 in 2008 to $50,789 in 2017. From 2015 to 2017, when FFS data was available, 110 of 1215 (9.1%) GAS patients had FFS. Most were non-Hispanic White (66.7%) or Black (23.8%). Fifty percent of FFS cases occurred in the West, followed by the Northeast (31.8%), South (13.6%), and Midwest (4.8%) (P = 0.015). By payer the cases were, 36.4% self-pay, 31.8% Medicaid, and 27.3% private insurance (P < 0.0001). Approximately, 18% of patients undergoing male-to-female transition received FFS. Conclusions: From 2008 to 2017, GAS cases increased nationwide while the average cost of surgery rose steeply. FFS cases were primarily in the Western and Northeast United States. Despite high cost, roughly 18% of transgender women in our sample received FFS, highlighting the importance of FFS in gender transition.

An Online Calculator for Estimating Breast Implant Volume from Imaging.

Breast implant surgery remains one of the most common surgical procedures performed in the United States. Implant exchange can be complicated by unavailability of medical records or implant identification cards. Using chest imaging of 154 breast implants, an algorithm for estimating breast implant volume was generated. Based on four simple measurements and patient body mass index, a free, online calculator was created with a mean error of volume estimate of less than 1 cm3 and a SD of 44 cm3. In instances where a surgeon does not have implant records available but has chest imaging, this online tool can be used to obtain a relatively accurate estimate of implant volume.