A Framework for Automatic Burn Image Segmentation and Burn Depth Diagnosis Using Deep Learning

Burn is a common traumatic disease with high morbidity and mortality. The treatment of burns requires accurate and reliable diagnosis of burn wounds and burn depth, which can save lives in some cases. However, due to the complexity of burn wounds, the early diagnosis of burns lacks accuracy and difference. Therefore, we use deep learning technology to automate and standardize burn diagnosis to reduce human errors and improve burn diagnosis. First, the burn dataset with detailed burn area segmentation and burn depth labelling is created. Then, an end-to-end framework based on deep learning method for advanced burn area segmentation and burn depth diagnosis is proposed. The framework is firstly used to segment the burn area in the burn images. On this basis, the calculation of the percentage of the burn area in the total body surface area (TBSA) can be realized by extending the network output structure and the labels of the burn dataset. Then, the framework is used to segment multiple burn depth areas. Finally, the network achieves the best result with IOU of 0.8467 for the segmentation of burn and no burn area. And for multiple burn depth areas segmentation, the best average IOU is 0.5144.

Download Full-text

Deep Learning Assisted Burn Wound Diagnosis (Preprint)

10.2196/preprints.22798 ◽

2020 ◽

Author(s):

Che Wei Chang ◽

Feipei Lai ◽

Mesakh Christian ◽

Yu Chun Chen ◽

Ching Hsu ◽

...

Keyword(s):

Deep Learning ◽

Total Body Surface Area ◽

Ground Truth ◽

Burn Patients ◽

Burn Wound ◽

Burn Wounds ◽

Burn Victims ◽

Step Procedure ◽

Resuscitation Fluid ◽

Total Body

BACKGROUND Accurate assessment of the percentage of total body surface area (%TBSA) of burn wounds is crucial in the management of burn patients. The resuscitation fluid and nutritional needs of burn patients, their need for intensive unit care, and probability of mortality are all directly related to %TBSA. It is difficult to estimate a burn area of irregular shape by inspection. Many articles have reported discrepancy in estimating %TBSA by different doctors. OBJECTIVE We propose a method, based on deep learning, for burn wound detection, segmentation and calculation of % TBSA on a pixel-to-pixel basis. METHODS A two-step procedure was used to convert burn wound diagnosis into %TBSA. In the first step, images of burn wounds were collected and labeled by burn surgeons and the dataset was then input into two deep learning architectures, U-Net and Mask R-CNN, each configured with two different backbones, to segment the burn wounds. In the second step, we collected and labeled images of hands to create another dataset, which was also input into U-Net and Mask R-CNN to segment the hands. The percentage of TBSA of the burn wounds was then calculated by comparing the pixels of mask areas on the images of the burn wound and hand of the same patient according to the rule of hand, which says that one’s hand accounts for 0.8% of TBSA. RESULTS A total of 2591 images of burn wounds were collected and labeled to form the burn-wound dataset. The dataset was randomly split into a ratio of 8:1:1 to form the training, validation, and testing sets. Four hundred images of volar hands were collected and labeled to form the hand dataset, which was also split into three sets using the same method. For the images of burn wounds, Mask R-CNN with ResNet101 had the best segmentation result with a Dice coefficient (DC) of 0.9496, while U-Net with ResNet101 had a DC of 0.8545. For the hand images, U-Net and Mask R-CNN had similar performance with a DC of 0.9920 and 0.9910, respectively. Lastly, we conducted a test diagnosis in a burn patient. Mask R-CNN with ResNet-101 had on average less deviation (0.115% TBSA) from the ground truth than burn surgeons. CONCLUSIONS This is one of the first studies to diagnose all depths of burn wounds and convert the segmentation results into %TBSA using different deep learning models. We aimed to assist medical staff in estimating burn size more accurately and thereby helping to provide precise care to burn victims.

Download Full-text

Effect of oral olive oil on healing of 10–20% total body surface area burn wounds in hospitalized patients

Burns ◽

10.1016/j.burns.2014.08.010 ◽

2015 ◽

Vol 41 (3) ◽

pp. 493-496 ◽

Cited By ~ 13

Author(s):

Mahtab Najmi ◽

Zahra Vahdat Shariatpanahi ◽

Mohammad Tolouei ◽

Zohreh Amiri

Keyword(s):

Olive Oil ◽

Surface Area ◽

Body Surface Area ◽

Body Surface ◽

Total Body Surface Area ◽

Hospitalized Patients ◽

Burn Wounds ◽

Total Body

Download Full-text

720 Scalds and Contact Burns: Are They Really Different?

Journal of Burn Care & Research ◽

10.1093/jbcr/iraa024.304 ◽

2020 ◽

Vol 41 (Supplement_1) ◽

pp. S191-S191

Author(s):

Adam Singer ◽

Evyatar Baer ◽

Henry Thode

Keyword(s):

Total Body Surface Area ◽

Retrospective Chart Review ◽

Stepwise Logistic Regression ◽

Full Thickness ◽

Chi Square ◽

Included Patient ◽

Body Location ◽

Total Body ◽

Fire Flame ◽

Burn Depth

Abstract Introduction Prior studies comparing scalds and contact burns rarely address the affect of burn etiology on burn depth. However, it is commonly believed that scalds tend to heal faster than similarly sized contact and flame burns. As a result, expectant therapy is often preferred after scald injuries. We compared the percentages of full-thickness burns based on burn etiology controlling for burn size, location and patient age. We hypothesized that the percentage of full thickness burns would be lower after scalds compared with contact and flame burns. Methods We performed a retrospective chart review of a prospectively collected burn registry of all patients admitted to a regional burn center between 2000–2010. Data collection included patient and burn characteristics including age, gender, body location, and burn etiology. We compared the percentages of full thickness burns among scald, contact and flame burns using Chi-square tests. Stepwise logistic regression was used to adjust for age, location, and burn size. Results There were 1038 patients in the study with either scald (n=537, 52%), fire/flame (n=434, 42%) or contact (n=67, 6%) burn. Mean (SD) age was 29 (25); 75% were male. Mean (SD) total body surface area (TBSA) was 11 (13)%. Mean (SD) length of stay was 10 (18) days. The percentages of full thickness burns by etiology were scalds 13.1%, fire/flame 34%, and contact burns 45% (P< 0.001). Patients with scalds were younger (22+/-24) than patients with contacts (32+/-28) and fire/flame (38+/-22) burns. Multivariate analysis for predicting full thickness burns found that compared with contact burns, scalds were less likely full thickness (OR 0.23, 95%CI 0.11–0.48) while fire/flame burns were as likely to be full thickness (OR 0.54, 95%CI 0.26–1.15). TBSA and age were also associated with full thickness burns (OR 1.06, 95%CI 1.04–1.09 and OR 1.015, 95%CI 1.007–1.024, respectively). Burns on the head and neck were less likely to be full thickness (OR 0.30, 95%CI, 0.11–0.82). Conclusions Scald burns are significantly less likely to be full thickness than contact or fire/flame burns. Applicability of Research to Practice Based on these results, expectant therapy may be more appropriate for scalds than contact or fire/flame burns.

Download Full-text

Rehabilitation of the Burn Patient

10.2310/7ccsp.2145 ◽

2019 ◽

Author(s):

Barclay T Stewart ◽

Samuel P Mandell ◽

Nicole Gibran

Keyword(s):

Injury Severity ◽

Community Integration ◽

Total Body Surface Area ◽

Rehabilitation Medicine ◽

Rehabilitation Program ◽

Loved One ◽

Total Body ◽

The Moment ◽

Burn Depth ◽

Need For Rehabilitation

Burn rehabilitation begins the moment the patient contacts the healthcare system and, for many patients, it never truly ends. Returning severely burned patients to their pre-injury level of function has become more challenging, as more severely burned patients survive their injury. The need for rehabilitation does not simply correspond with burn depth, total body surface area involved, or injury severity. Other factors also impact need for rehabilitation, such as anatomic location of injury (eg, across finger joints, face), additional psychosocial strain from a house fire or having an injured loved one, or body image concerns related to visible scarring. To meet these and other challenges posed by burn injuries, rehabilitation requires a multidisciplinary team including, but not limited to, nursing, surgery, burn therapy, rehabilitation medicine, vocational counseling, rehabilitation psychology, psychiatry, and nutrition. Accordingly, verification as a burn center by the American Burn Association requires a goal-oriented, multidisciplinary rehabilitation program. This chapter broadly discusses the spectrum and integral components of burn rehabilitation. This review contains 13 figures, 8 tables, and 71 references. Key Words: burn, community integration, function, mental health, multidisciplinary, outcome, pain, quality of life, reconstruction, rehabilitation, therapy.

Download Full-text

Epidemiology of Scars and Their Consequences: Burn Scars

Textbook on Scar Management ◽

10.1007/978-3-030-44766-3_5 ◽

2020 ◽

pp. 37-43

Author(s):

Margriet E. van Baar

Keyword(s):

Risk Factors ◽

Burn Injury ◽

Healing Process ◽

Total Body Surface Area ◽

Female Gender ◽

Burn Injuries ◽

Post Injury ◽

The People ◽

Total Body ◽

Burn Depth

AbstractPathological scarring in burn wounds can result in hypertrophic scars and/or contractures. Prevalences of hypertrophic scarring after burn injuries between 8% and 67% are reported. A recent prospective study revealed a prevalence of 8%. Data on prevalence of burn scar contractures are limited; reported prevalence at discharge varied between 38 and 54% and decreased with an increasing time post burn. About 5–20% of the people who suffered from burn injuries received reconstructive surgery after burns, up to 10 years post injury.Factors predicting pathological scar formation after burn injuries include patient, injury and treatment characteristics. Injury- and treatment-related characteristics are the main predictors of scar outcomes after burn injury. These characteristics are related to burn size (total body surface area burned) and burn depth (number or type of surgery) or the overall healing process in general (length of stay, wound healing complications). Intrinsic patient-related risk factors seem to play a role as well but are less consistent predictors of scar outcome. This includes the risk factors like the female gender and also a younger age and darker skin.Knowledge on risk factors for poor scar outcome can be used to tailor treatment, aftercare and scar prevention to these patients with a high-risk profile.

Download Full-text

Automatic Detection of Airway Invasion from Videofluoroscopy via Deep Learning Technology

Applied Sciences ◽

10.3390/app10186179 ◽

2020 ◽

Vol 10 (18) ◽

pp. 6179

Author(s):

Seong Jae Lee ◽

Joo Young Ko ◽

Hyun Il Kim ◽

Sang-Il Choi

Keyword(s):

Deep Learning ◽

Present Model ◽

Spinal Column ◽

Region Of Interest ◽

Learning Technology ◽

Accurate Analysis ◽

Human Errors ◽

Airway Invasion ◽

Human Effort ◽

Deep Learning Model

In dysphagia, food materials frequently invade the laryngeal airway, potentially resulting in serious consequences, such as asphyxia or pneumonia. The VFSS (videofluoroscopic swallowing study) procedure can be used to visualize the occurrence of airway invasion, but its reliability is limited by human errors and fatigue. Deep learning technology may improve the efficiency and reliability of VFSS analysis by reducing the human effort required. A deep learning model has been developed that can detect airway invasion from VFSS images in a fully automated manner. The model consists of three phases: (1) image normalization, (2) dynamic ROI (region of interest) determination, and (3) airway invasion detection. Noise induced by movement and learning from unintended areas is minimized by defining a “dynamic” ROI with respect to the center of the cervical spinal column as segmented using U-Net. An Xception module, trained on a dataset consisting of 267,748 image frames obtained from 319 VFSS video files, is used for the detection of airway invasion. The present model shows an overall accuracy of 97.2% in classifying image frames and 93.2% in classifying video files. It is anticipated that the present model will enable more accurate analysis of VFSS data.

Download Full-text

Evaluation of Demographic and Clinical Characteristics of Patients who Attempted Suicide by Self-Inflicted Burn Using Catalyzer

International Surgery ◽

10.9738/intsurg-d-13-00189.1 ◽

2015 ◽

Vol 100 (2) ◽

pp. 304-308 ◽

Cited By ~ 3

Author(s):

Hakan Yabanoglu ◽

Huseyin Ozgur Aytac ◽

Emin Turk ◽

Erdal Karagulle ◽

Sedat Belli ◽

...

Keyword(s):

Mortality Rate ◽

Clinical Characteristics ◽

Attempted Suicide ◽

Total Body Surface Area ◽

Inhalation Injury ◽

Demographic Characteristics ◽

High Morbidity ◽

Average Percentage ◽

Study Population ◽

Total Body

Our aim was to assess demographic and clinical characteristics of patients treated at our units who attempted suicide by self-incineration, and to compare the results of burns with or without catalyzer use. Twenty patients who attempted suicide by self-incineration were examined in terms of clinical and demographic characteristics. Average age of the study population was 35 years (range 13−85 years). Average percentage of total body surface area burn was 53% (9%−100%). Six (30%) patients used gasoline and 5 (25%) used paint thinner in order to catalyze burning. Of these 11 patients who used a catalyzer, 5 (45.4%) had inhalation injury and 7 (63.6%) died. Among 9 patients who did not use any catalyzer, 1 (11.1%) had inhalation injury and 4 (44.4%) died. In general, inhalation injury was diagnosed in 6 patients (30%) while 11 (55%) patients died. A high morbidity and mortality rate was found in patients who used a catalyzer.

Download Full-text

Divulging the Complexities of Deep Partial- and Full-Thickness Burn Wounds Afflicted by Staphylococcus Aureus Biofilms in a Rat Burn Model

European Journal of Burn Care ◽

10.3390/ebj2030009 ◽

2021 ◽

Vol 2 (3) ◽

pp. 106-124

Author(s):

Alan J. Weaver Jr. ◽

Kenneth S. Brandenburg ◽

S. L. Rajasekhar Karna ◽

Christopher Olverson ◽

Kai P. Leung

Keyword(s):

Staphylococcus Aureus ◽

Total Body Surface Area ◽

Complement C3 ◽

Full Thickness ◽

Sprague Dawley ◽

Burn Wound ◽

Burn Wounds ◽

Gm Csf ◽

Aureus Infection ◽

Total Body

Every year, thousands of soldiers and civilians succumb to burn wound trauma with highly unfavorable outcomes. We previously established a modified Walker-Mason rat scald model exhibiting a P. aeruginosa infection. Here we characterize deep partial- (DPT) and full-thickness (FT) burn wounds inoculated with Staphylococcus aureus. Male Sprague-Dawley rats (350–450 g) inflicted with 10% total body surface area burn inoculated with S. aureus (103–5 CFU/wound) were monitored over an 11-day period. S. aureus rapidly dominated the wound bed, with bacterial loads reaching at least 1 × 109 CFU/g tissue in all wounds. Within 3 days, S. aureus biofilm formation occurred based on genetic transcripts and Giemsa staining of the tissue. S. aureus infection resulted in a slightly faster recruitment of neutrophils in FT wounds, which was related to necrotic neutrophils. The extent of the inflammatory response in S. aureus infected burn wounds correlated with elevated G-CSF, GM-CSF, GRO/KC and/or TNF-α levels, but a majority of pro- and anti-inflammatory cytokines (IL-1β, IL-6, IFN-γ, IL-10, and IL-13) were found to be suppressed, compared to burn-only controls. S. aureus infection resulted in dynamic changes in DAMPs, including elevated HMGB-1 and reduced levels of circulating hyaluronan within FT wounds. S. aureus also reduced complement C3 at all time points in DPT and FT wounds. These changes in DAMPs are believed to be correlated with burn severity and S. aureus specific bioburden. Collectively, this model showcases the evasiveness of S. aureus through dampening the immune response to flourish in the burn wound.

Download Full-text

865 Assessing Opioid Administration in Pediatric Burn Patients with Nonsurgical Management

Journal of Burn Care & Research ◽

10.1093/jbcr/iraa024.431 ◽

2020 ◽

Vol 41 (Supplement_1) ◽

pp. S271-S271

Author(s):

Holden Richards ◽

Junxin Shi ◽

Rajan Thakkar ◽

Sheila Giles ◽

Krista K Wheeler ◽

...

Keyword(s):

Temporal Trend ◽

Total Body Surface Area ◽

Burn Patients ◽

Simple Linear Regression ◽

Nonsurgical Management ◽

Pain Scores ◽

Opioid Administration ◽

Vast Literature ◽

Total Body ◽

Burn Depth

Abstract Introduction Despite the vast literature studying the opioid crisis, there is sparse data in the pediatric burn population. This study sought to assess patient-level characteristics and their potential effects on opioid administration in nonsurgical pediatric burn inpatients. Methods Admitted burn patients from 2013–2018 with nonsurgical management at an American Burn Association (ABA) verified pediatric burn center were retrospectively identified. Morphine milligram equivalents by weight (MME/kg) per admission were evaluated through a multiple loglinear regression with race, sex, age, total body surface area burned (TBSA), and burn depth as predictors. Simple linear regression was used to evaluate the temporal trend of median opioid utilization. Results A total of 806 patients (55% White, 35% Black, 5% Hispanic, 5% Other) were included. In an adjusted analysis, no differences in opioid administration were seen by sex, burn degree, or for Blacks and Hispanics when compared with Whites. Increased MME/kg was associated with older age (10–18 years; p< 0.0001) and larger burns (>5% TBSA burned; p< 0.0001). From 2013 to 2018, median MME/kg per admission declined significantly (2013:0.21, 2018:0.09; p=0.0103). Conclusions Nonsurgical burn patients who were older and presented with larger TBSA experienced marked increases in opioid utilization. Overall, opioid administration decreased over time. Applicability of Research to Practice Future pediatric burn research should evaluate disparities in opioid administration based on injury location and surgical management. Further, the accuracy of pain scores for non-verbal pediatric burn patients should be studied as discrepancies in opioid utilization exist.

Download Full-text