674 Machine Learning and Automation in Burn Care: A Systematic Review

2021 ◽  
Vol 42 (Supplement_1) ◽  
pp. S193-S193
Author(s):  
Samantha Huang ◽  
Justin Dang ◽  
Clifford C Sheckter ◽  
Haig A Yenikomshian ◽  
Justin Gillenwater
Keyword(s):  
Machine Learning ◽  
Systematic Review ◽  
Total Body Surface Area ◽  
Burn Wound ◽  
Burn Care ◽  
Surgeon Experience ◽  
Automated Software ◽  
Burn Depth ◽  
Automated Methods ◽  
Equal Weighting

Abstract Introduction Current methods of burn evaluation and treatment are subjective and dependent on surgeon experience, with high rates of inter-rater variability leading to inaccurate diagnoses and treatment. Machine learning (ML) and automated methods are being used to develop more objective and accurate methods for burn diagnosis and triage. Defined as a subfield of artificial intelligence that applies algorithms capable of knowledge acquisition, machine learning draws patterns from data, which it can then apply to clinically relevant tasks. This technology has the potential to improve burn management by quantitating diagnoses, improving diagnostic accuracy, and increasing access to burn care. The aim of this systematic review is to summarize the literature regarding machine learning and automated methods for burn wound evaluation and treatment. Methods A systematic review of articles available on PubMed and MEDLINE (OVID) was performed. Keywords used in the search process included burns, machine learning, deep learning, burn classification technology, and mobile applications. Reviews, case reports, and opinion papers were excluded. Data were extracted on study design, study objectives, study models, devices used to capture data, machine learning, or automated software used, expertise level and number of evaluators, and ML accuracy of burn wound evaluation. Results The search identified 592 unique titles. After screening, 35 relevant articles were identified for systematic review. Nine studies used machine learning and automated software to estimate percent total body surface area (%TBSA) burned, 4 calculated fluid requirements, 18 estimated burn depth, 5 estimated need for surgery, 6 predicted mortality, and 2 evaluated scarring in burn patients. Devices used to estimate %TBSA burned showed an accuracy comparable to or better than traditional methods. Burn depth estimation sensitivities resulted in unweighted means >81%, which increased to >83% with equal weighting applied. Mortality prediction sensitivity had an unweighted mean of 96.75%, which increased to 99.35% with equal weighting. Conclusions Machine learning and automated technology are promising tools that provide objective and accurate measures of evaluating burn wounds. Existing methods address the key steps in burn care management; however, existing data reporting on their robustness remain in the early stages. Further resources should be dedicated to leveraging this technology to improve outcomes in burn care.

Use of Infrared Thermography for Assessment of Burn Depth and Healing Potential: A Systematic Review

2021 ◽  
Author(s):  
Justin Dang ◽  
Matthew Lin ◽  
Calvin Tan ◽  
Christopher H Pham ◽  
Samantha Huang ◽  
...  
Keyword(s):  
Systematic Review ◽  
Sensitivity And Specificity ◽  
Clinical Assessment ◽  
Meta Analysis ◽  
Ir Thermography ◽  
Burn Wound ◽  
Burn Care ◽  
Wound Assessment ◽  
Burn Depth ◽  
Mean Sensitivity

Abstract Introduction Burn wound depth assessments are an important component of determining patient prognosis and making appropriate management decisions. Clinical appraisal of the burn wound by an experienced burn surgeon is standard of care but has limitations. IR thermography is a technology in burn care that can provide a non-invasive, quantitative method of evaluating burn wound depth. IR thermography utilizes a specialized camera that can capture the infrared emissivity of the skin, and the resulting images can be analyzed to determine burn depth and healing potential of a burn wound. Though IR thermography has great potential for burn wound assessment, its use for this has not been well documented. Thus, we have conducted a systematic review of the current use of IR thermography to assess burn depth and healing potential. Methods A systematic review and meta-analysis of the literature was performed on PubMed and Google Scholar between June 2020-December 2020 using the following keywords: FLIR, FLIR ONE, thermography, forward looking infrared, thermal imaging + burn*, burn wound assessment, burn depth, burn wound depth, burn depth assessment, healing potential, burn healing potential. A meta-analysis was performed on the mean sensitivity and specificity of the ability of IR thermography for predicting healing potential. Inclusion criteria were articles investigating the use of IR thermography for burn wound assessments in adults and pediatric patients. Reviews and non-English articles were excluded. Results A total of 19 articles were included in the final review. Statistically significant correlations were found between IR thermography and laser doppler imaging (LDI) in 4/4 clinical studies. A case report of a single patient found that IR thermography was more accurate than LDI for assessing burn depth. Five articles investigated the ability of IR thermography to predict healing time, with four reporting statistically significant results. Temperature differences between burnt and unburnt skin were found in 2/2 articles. IR thermography was compared to clinical assessment in five articles, with varying results regarding accuracy of clinical assessment compared to thermography. Mean sensitivity and specificity of the ability of IR thermography to determine healing potential <15 days was 44.5 and 98.8 respectively. Mean sensitivity and specificity of the ability of FLIR to determine healing potential <21 days was 51.2 and 77.9 respectively. Conclusion IR thermography is an accurate, simple, and cost-effective method of burn wound assessment. FLIR has been demonstrated to have significant correlations with other methods of assessing burns such as LDI and can be utilized to accurately assess burn depth and healing potential.

558 Use of Forward-Looking Infrared (FLIR) Technology to Assess Burn Depth and Healing Potential: A Systematic Review and Meta-Analysis

2021 ◽  
Vol 42 (Supplement_1) ◽  
pp. S128-S129
Author(s):  
Justin Dang ◽  
Matthew E Lin ◽  
Samantha Huang ◽  
Ian F Hulsebos ◽  
Haig A Yenikomshian ◽  
...  
Keyword(s):  
Systematic Review ◽  
Sensitivity And Specificity ◽  
Clinical Assessment ◽  
Meta Analysis ◽  
Burn Wound ◽  
Burn Care ◽  
Wound Assessment ◽  
Burn Depth ◽  
Mean Sensitivity ◽  
Forward Looking

Abstract Introduction Burn wound depth assessments are an important component of determining patient prognosis and making appropriate management decisions. Clinical appraisal of the burn wound by an experienced burn surgeon is standard of care but has limitations. Forward-looking Infrared (FLIR) is a new technology in burn care that can provide a non-invasive, quantitative method of evaluating burn wound depth. FLIR utilizes a specialized camera that can capture the infrared emissivity of the skin, and the resulting images can be analyzed to determine burn depth and healing potential of a burn wound. Though FLIR has great potential for burn wound assessment, its use for this has not been well documented. Thus, we have conducted a systematic review and meta-analysis of the current use of FLIR technology to assess burn depth and healing potential. Methods A systematic review of the literature was performed on PubMed and Google Scholar between June 2020-August 2020 using the following keywords: thermal imaging, FLIR, forward looking infrared, burn, burn depth. Meta-analysis was performed on the mean sensitivity and specificity of the ability of FLIR to predict healing potential. Inclusion criteria were articles investigating the use of FLIR for burn wound assessments in adults, pediatric patients and animal models. Reviews and non-English articles were excluded. Results A total of 11 articles were included in the final review. Statistically significant correlations were found between FLIR and laser doppler imaging (LDI) in 3/3 clinical studies. A case report of a single patient found that FLIR was more accurate than LDI for assessing burn depth. Three articles investigated the ability of FLIR to predict healing potential, with all three reporting statistically significant results. Significant temperature differences between burnt and unburnt skin were found in 2/2 articles. FLIR was compared to clinical assessment by burn surgeons in two articles; one article found that FLIR was more accurate for assessing burn depth, while the other article found that clinical assessment was more accurate for predicting healing potential < 21 days. Mean sensitivity and specificity of the ability of FLIR to determine healing potential < 15 days was 44.5 and 98.8 respectively. Mean sensitivity and specificity of the ability of FLIR to determine healing potential < 21 days was 44.0 and 77.4 respectively. Conclusions FLIR is an accurate, simple, and cost-effective method of burn wound assessment. FLIR has been demonstrated to have significant correlations with other methods of assessing burns such as LDI and can be utilized to accurately assess burn depth and healing potential.

Timing of Flap Surgery in Acute Burn Patients Does Not Affect Complications

2020 ◽  
Vol 41 (5) ◽  
pp. 967-970
Author(s):  
David Perrault ◽  
Danielle Rochlin ◽  
Christopher Pham ◽  
Arash Momeni ◽  
Yvonne Karanas ◽  
...  
Keyword(s):  
Independent Predictor ◽  
Multivariable Analysis ◽  
Total Body Surface Area ◽  
Free Flaps ◽  
Stepwise Logistic Regression ◽  
Electrical Injury ◽  
Burn Patients ◽  
Burn Wound ◽  
Burn Care ◽  
Flap Coverage

Abstract Pedicled and free flaps are occasionally necessary to reconstruct complex wounds in acute burn patients. Flap coverage has classically been delayed for concern of progressive tissue necrosis and flap failure. We aim to investigate flap complications in primary burn care leveraging national U.S. data. Acute burn patients with known % total body surface area(TBSA) were extracted from the Nationwide/National Inpatient Sample from 2002 to 2014 based on the International Classification of Disease (ICD) codes, ninth edition. Variables included age, sex, race, Elixhauser index, %TBSA, mechanism, inhalation injury, and location of burn. Flap complication was defined by ICD-9 procedure code 86.75, return to the operating room for flap revision. Multivariable analysis evaluated predictors of flap compromise using stepwise logistic regression with backward elimination. The weighted sample included 306,924 encounters of which 526 received a flap (0.17%). About 7.8% of flap encounters sustained electric injury compared to 2.7% of non-flap encounters (odds ratio [OR] 3.76, 95% confidence interval [CI] 1.95–7.24, P < .001). The mean hospital day of the flap procedure was 10.1 (SD 10.7) days. Flap complications occurred in 6.4% of cases. The timing of flap coverage was not associated with complications. The only independent predictor of flap complication was electrical injury (OR 40.49, 95% CI 2.98–550.64, P = .005). Electrical injury was an independent predictor of flap complications compared to other mechanisms. Flap timing was not associated with return to surgery for complications. This suggests that the use of flaps is safe in acute burn care to achieve burn wound closure with an understanding that electrical injuries may warrant particular consideration to avoid failure.

scholarly journals Impact of Virtual Reality Technology on Pain and Anxiety in Pediatric Burn Patients: A Systematic Review and Meta-Analysis

2022 ◽  
Vol 2 ◽  
Author(s):  
Kathryn L. Smith ◽  
Yang Wang ◽  
Luana Colloca
Keyword(s):  
Systematic Review ◽  
Virtual Reality ◽  
Effect Size ◽  
Pediatric Patients ◽  
Wound Care ◽  
Meta Analysis ◽  
Combined Effect ◽  
Burn Wound ◽  
Virtual Reality Technology ◽  
Burn Care

Introduction: Virtual reality (VR) has the potential to lessen pain and anxiety experienced by pediatric patients undergoing burn wound care procedures. Population-specific variables require novel technological application and thus, a systematic review among studies on its impact is warranted.Objective: The objective of this review was to evaluate the effectiveness of VR on pain in children with burn injuries undergoing wound care procedures.Methods: A systematic literature review was performed using PubMed and CINAHL databases from January 2010 to July 2021 with the keywords “pediatric,” “burn,” “virtual reality,” and “pain.” We included experimental studies of between- and within-subjects designs in which pediatric patients’ exposure to virtual reality technology during burn wound care functioned as the intervention of interest. Two researchers independently performed the literature search, made judgements of inclusion/exclusion based on agreed-upon criteria, abstracted data, and assessed quality of evidence using a standardized appraisal tool. A meta-analysis was conducted to evaluate the effectiveness of the VR on burning procedural pain in pediatric population. Standardized mean difference (SMD) was used as an index of combined effect size, and a random effect model was used for meta-analysis.Results: Ten articles published between January 2010 and July 2021 passed the selection criteria: six randomized controlled trials and four randomized repeated-measures studies. Consistent results among the studies provided support for VR as effective in reducing pain and potentially pain related anxiety in children undergoing burn wound care through preprocedural preparation (n = 2) and procedural intervention (n = 8). A random effects meta-analysis model indicated a moderate and significant combined effect size (SMD = 0.60, 95% CI = 0.28–0.93, p = 0.0031) of VR effects on pain intensity ratings with no significant heterogeneity of VR intervention effects between studies. Only one study reported direct influence of VR intervention on pre-procedural situational anxiety with a moderate effect size (Cohen’s d = 0.575, 95%CI = 0.11–1.04).Conclusion: Children’s exposure to VR during burn care procedures was associated with lower levels of pain and pain related anxiety. Moderate to large effect sizes support the integration of VR into traditional pediatric burn pain protocols irrespective of innovative delivery methods and content required for use in burned pediatric patients.

A systematic review of machine learning and automation in burn wound evaluation: A promising but developing frontier

Burns ◽  
2021 ◽  
Author(s):  
Samantha Huang ◽  
Justin Dang ◽  
Clifford C. Sheckter ◽  
Haig A. Yenikomshian ◽  
Justin Gillenwater
Keyword(s):  
Machine Learning ◽  
Systematic Review ◽  
Burn Wound

539 If Seeing Was Believing - A Retrospective Analysis of Potential Reduced Treatment Delays with a Novel Burn Wound Assessment Device.

2021 ◽  
Vol 42 (Supplement_1) ◽  
pp. S117-S118
Author(s):  
Henry B Huson ◽  
Herbert Phelan ◽  
David G’Sell ◽  
Sydney Smith ◽  
Jeffrey E Carter
Keyword(s):  
Length Of Stay ◽  
Total Body Surface Area ◽  
Machine Learning Algorithms ◽  
Order Entry ◽  
Burn Wound ◽  
Imaging Device ◽  
Treatment Delays ◽  
Burn Care ◽  
Wound Assessment ◽  
Electronic Order

Abstract Introduction Burn care (BC) remains a highly specialized and resource intensive specialty with only 2% of hospitals featuring a burn center and less than 1% of graduating general surgery and plastic surgery residents pursing a burn fellowship each year. Access to specialized care is further complicated by burn wound assessment (BWA) which is commonly performed visually without adjunctive devices. To help clinicians make more accurate assessments and potentially reduce delays in transfer or treatment, a new non-invasive imaging device for BWA is being developed using visible and non-visible wavelengths of light with machine learning algorithms. Our goal was to assess the potential reduced treatment delay (RTD) and associated financial savings by implementing such a device using our burn center’s historical data. Methods The study was an IRB-approved, retrospective review of admissions from 07/01/2018 through 06/30/2019. Inclusion criteria: thermal, chemical, contact, or electrical mechanism of injury, >15 years of age requiring excision, and length of stay >72 hours. Inclusion data included: presence/absence of concomitant trauma, day of surgery, day of admission, day of electronic order entry for case request, and length of stay per percent total body surface area (LOS%TBSA). RTD was defined starting >48 hours after injury daily until electronic order entry for surgical case request. Reduced costs were calculated per day from prior studies ranging $3,000 to $5,100/day. Results A total of 109 patients were included. 29 patients had case requests placed within 48 hours of admission. Of the remaining 80 patients, a potential of 398 days would have been saved had a novel BWA adjunctive imaging devices aided surgeon to requests earlier surgical intervention. Overall savings from reduced length of stay range from $1,194,000 to $2,029,800 dollars. Conclusions Our study demonstrates that should a BWA technology with accuracy 48 hours after injury be developed, even burn centers with 24-hour access to operating rooms can reduce treatment delays. The study does not look at additional cost savings offered by reduced emergent transfers or admissions which offer additional intrigue and promise.

Diagnostic test accuracy for use of machine learning in diagnosis of autism spectrum disorder: A Systematic Review and Meta-Analysis (Preprint)

2019 ◽  
Author(s):  
Sun Jae Moon ◽  
Jin Seub Hwang ◽  
Rajesh Kana ◽  
John Torous ◽  
Jung Won Kim
Keyword(s):  
Machine Learning ◽  
Systematic Review ◽  
Autism Spectrum Disorder ◽  
Meta Analysis ◽  
Learning Algorithms ◽  
Structural Mri ◽  
Autism Spectrum ◽  
Machine Learning Algorithms ◽  
Spectrum Disorder ◽  
Test Accuracy

BACKGROUND Over the recent years, machine learning algorithms have been more widely and increasingly applied in biomedical fields. In particular, its application has been drawing more attention in the field of psychiatry, for instance, as diagnostic tests/tools for autism spectrum disorder. However, given its complexity and potential clinical implications, there is ongoing need for further research on its accuracy. OBJECTIVE The current study aims to summarize the evidence for the accuracy of use of machine learning algorithms in diagnosing autism spectrum disorder (ASD) through systematic review and meta-analysis. METHODS MEDLINE, Embase, CINAHL Complete (with OpenDissertations), PsyINFO and IEEE Xplore Digital Library databases were searched on November 28th, 2018. Studies, which used a machine learning algorithm partially or fully in classifying ASD from controls and provided accuracy measures, were included in our analysis. Bivariate random effects model was applied to the pooled data in meta-analysis. Subgroup analysis was used to investigate and resolve the source of heterogeneity between studies. True-positive, false-positive, false negative and true-negative values from individual studies were used to calculate the pooled sensitivity and specificity values, draw SROC curves, and obtain area under the curve (AUC) and partial AUC. RESULTS A total of 43 studies were included for the final analysis, of which meta-analysis was performed on 40 studies (53 samples with 12,128 participants). A structural MRI subgroup meta-analysis (12 samples with 1,776 participants) showed the sensitivity at 0.83 (95% CI-0.76 to 0.89), specificity at 0.84 (95% CI -0.74 to 0.91), and AUC/pAUC at 0.90/0.83. An fMRI/deep neural network (DNN) subgroup meta-analysis (five samples with 1,345 participants) showed the sensitivity at 0.69 (95% CI- 0.62 to 0.75), the specificity at 0.66 (95% CI -0.61 to 0.70), and AUC/pAUC at 0.71/0.67. CONCLUSIONS Machine learning algorithms that used structural MRI features in diagnosis of ASD were shown to have accuracy that is similar to currently used diagnostic tools.

scholarly journals Machine Learning for the Classification of Alzheimer’s Disease and Its Prodromal Stage Using Brain Diffusion Tensor Imaging Data: A Systematic Review

Processes ◽  
2020 ◽  
Vol 8 (9) ◽  
pp. 1071
Author(s):  
Lucia Billeci ◽  
Asia Badolato ◽  
Lorenzo Bachi ◽  
Alessandro Tonacci
Keyword(s):  
Machine Learning ◽  
Systematic Review ◽  
Alzheimer’S Disease ◽  
Alzheimer's Disease ◽  
Diffusion Tensor Imaging ◽  
Magnetic Resonance ◽  
Diffusion Tensor ◽  
Classification Problem ◽  
Computer Algorithms ◽  
Imaging Data

Alzheimer’s disease is notoriously the most common cause of dementia in the elderly, affecting an increasing number of people. Although widespread, its causes and progression modalities are complex and still not fully understood. Through neuroimaging techniques, such as diffusion Magnetic Resonance (MR), more sophisticated and specific studies of the disease can be performed, offering a valuable tool for both its diagnosis and early detection. However, processing large quantities of medical images is not an easy task, and researchers have turned their attention towards machine learning, a set of computer algorithms that automatically adapt their output towards the intended goal. In this paper, a systematic review of recent machine learning applications on diffusion tensor imaging studies of Alzheimer’s disease is presented, highlighting the fundamental aspects of each work and reporting their performance score. A few examined studies also include mild cognitive impairment in the classification problem, while others combine diffusion data with other sources, like structural magnetic resonance imaging (MRI) (multimodal analysis). The findings of the retrieved works suggest a promising role for machine learning in evaluating effective classification features, like fractional anisotropy, and in possibly performing on different image modalities with higher accuracy.

Sign in / Sign up

Export Citation Format

Share Document