Thermal Imaging for Burn Wound Depth Assessment: A Mixed-Methods Implementation Study

Background: Implementing innovations emerging from clinical research can be challenging. Thermal imagers provide an accessible diagnostic tool to increase the accuracy of burn wound depth assessment. This mixed-methods implementation study aimed to assess the barriers and facilitators, design implementation strategies, and guide the implementation process of thermal imaging in the outpatient clinic of a burn centre. Methods: This study was conducted between September 2022 and February 2023 in Beverwijk, The Netherlands. Semi-structured interviews with burn physicians guided by the Consolidated Framework for Implementation Research (CFIR) were conducted to identify barriers and facilitators. Based on the barriers, implementation strategies were developed with the CFIR-ERIC Matching Tool, and disseminated to support the uptake of the thermal imager. Subsequently, thermal imaging was implemented in daily practice, and an iterative RE-AIM approach was used to evaluate the implementation process. Results: Common facilitators for the implementation of the thermal imager were the low complexity, the relative advantage above other diagnostic tools, and benefits for patients. Common barriers were physicians' attitude towards and perceived value of the intervention, the low compatibility with the current workflow, and a lack of knowledge about existing evidence. Six implementation strategies were developed: creating a formal implementation blueprint, promoting adaptability, developing educational materials, facilitation, conducting ongoing training, and identifying early adopters. These strategies resulted in the effective implementation of the thermal imager, reflected by a >70% reach among eligible patients, and >80% effectiveness and adoption. Throughout the implementation process, compatibility, and available resources remained barriers, resulting in low ratings on RE-AIM dimensions. Conclusions: This study developed implementation strategies based on the identified CFIR constructs that impacted the implementation of a thermal imager for burn wound assessment in our outpatient clinic. The experiences and findings of this study could be leveraged to guide the implementation of thermal imaging and other innovations in burn care.

Review of machine learning for optical imaging of burn wound severity assessment

Significance

Over the past decade, machine learning (ML) algorithms have rapidly become much more widespread for numerous biomedical applications, including the diagnosis and categorization of disease and injury.

Aim

Here, we seek to characterize the recent growth of ML techniques that use imaging data to classify burn wound severity and report on the accuracies of different approaches.

Approach

To this end, we present a comprehensive literature review of preclinical and clinical studies using ML techniques to classify the severity of burn wounds.

Results

The majority of these reports used digital color photographs as input data to the classification algorithms, but recently there has been an increasing prevalence of the use of ML approaches using input data from more advanced optical imaging modalities (e.g., multispectral and hyperspectral imaging, optical coherence tomography), in addition to multimodal techniques. The classification accuracy of the different methods is reported; it typically ranges from ∼ 70 % to 90% relative to the current gold standard of clinical judgment.

Conclusions

The field would benefit from systematic analysis of the effects of different input data modalities, training/testing sets, and ML classifiers on the reported accuracy. Despite this current limitation, ML-based algorithms show significant promise for assisting in objectively classifying burn wound severity.

Accurate and early prediction of the wound healing outcome of burn injuries using the wavelet Shannon entropy of terahertz time-domain waveforms

Significance

Severe burn injuries cause significant hypermetabolic alterations that are highly dynamic, hard to predict, and require acute and critical care. The clinical assessments of the severity of burn injuries are highly subjective and have consistently been reported to be inaccurate. Therefore, the utilization of other imaging modalities is crucial to reaching an objective and accurate burn assessment modality.

Aim

We describe a non-invasive technique using terahertz time-domain spectroscopy (THz-TDS) and the wavelet packet Shannon entropy to automatically estimate the burn depth and predict the wound healing outcome of thermal burn injuries.

Approach

We created 40 burn injuries of different severity grades in two porcine models using scald and contact methods of infliction. We used our THz portable handheld spectral reflection (PHASR) scanner to obtain the in vivo THz-TDS images. We used the energy to Shannon entropy ratio of the wavelet packet coefficients of the THz-TDS waveforms on day 0 to create supervised support vector machine (SVM) classification models. Histological assessments of the burn biopsies serve as the ground truth.

Results

We achieved an accuracy rate of 94.7% in predicting the wound healing outcome, as determined by histological measurement of the re-epithelialization rate on day 28 post-burn induction, using the THz-TDS measurements obtained on day 0. Furthermore, we report the accuracy rates of 89%, 87.1%, and 87.6% in automatic diagnosis of the superficial partial-thickness, deep partial-thickness, and full-thickness burns, respectively, using a multiclass SVM model.

Conclusions

The THz PHASR scanner promises a robust, high-speed, and accurate diagnostic modality to improve the clinical triage of burns and their management.

Optical coherence tomography correlates multiple measures of tissue damage following acute burn injury

Background

The visual assessment of burned skin is inherently subjective, and whilst a number of imaging modalities have identified quantifiable parameters to characterize vascular and structural changes following burn damage, none have become common place in the assessment protocol. Here, we use optical coherence tomography (OCT)-based angiography (OCTA) to introduce novel correlations between vessel depth, i.e., the depth of functional blood vessels beneath the tissue surface, edema depth, i.e., the depth of interstitial fluid buildup beneath the tissue surface, and tissue injury depth, i.e., the depth of collagen denaturation beneath the tissue surface, following burn injury.

Methods

A clinical prototype OCT system was used to collect OCT images from various sites of burned skin in patients. Optical microangiography (OMAG) algorithm was used to derive OCTA information from the acquired OCT images, from which the presence of blood vessels and edema were detected. The optical attenuation mapping of structural OCT information was used to detect tissue injury depth. The depths of vessel, edema and tissue injury were measured using a semi-automatic segmentation algorithm. Correlation analysis was performed using a Pearson correlation coefficient using one-tailed analysis with significance being established by a P value ≤0.05.

Results

Four burn patients were recruited and scanned at multiple sites using the prototype system within 3-6 days of injury. Approximate measurements include a vessel depth range of 320-1,360 µm, an edema depth range of 0-400 µm, and a tissue injury depth range of 130-420 µm. Correlations were subsequently observed between vessel depth and edema depth (r=0.8521, P=0.0001), and vessel depth and tissue injury depth (r=0.6296, P=0.0106).

Conclusions

OCT is feasible to provide the critical information of vessel depth, edema depth, and tissue injury depth of skin burns, which may represent viable assessment criteria for the characterization of cutaneous burns in future.

The effect of ambient lighting on Laser Doppler Imaging of a standardized cutaneous injury model

OBJECTIVE

The aim of this study was to investigate the potential confounding effects of four different types of ambient lighting on the results of Laser Doppler Imaging (LDI) of a standardized cutaneous injury model.

METHODS

After applying a mechanical stimulus to the anterior forearm of a healthy volunteer and inducing a wheal and arteriolar flare (the Triple response), we used a Laser Doppler Line Scanner (LDLS) to image the forearm under four different types of ambient lighting: light-emitting-diode (LED), compact fluorescent lighting (CFL), halogen, daylight, and darkness as a control. A spectrometer was used to measure the intensity of light energy at 785 nm, the wavelength used by the scanner for measurement under each type of ambient lighting.

RESULTS

Neither the LED nor CFL bulbs emitted detectable light energy at a wavelength of 785 nm. The color-based representation of arbitrary perfusion unit (APU) values of the Triple response measured by the scanner was similar between darkness, LED, and CFL light. Daylight emitted 2 mW at 785 nm, with a slight variation tending more towards lower APU values compared to darkness. Halogen lighting emitted 6 mW of light energy at 785 nm rendering the color-based representation impossible to interpret.

CONCLUSIONS

Halogen lighting and daylight have the potential to confound results of LDI of cutaneous injuries whereas LED and CFL lighting did not. Any potential sources of daylight should be reduced and halogen lighting completely covered or turned off prior to wound imaging.

Burn injury: what's in a name? Labels used for burn injury classification: a review of the data from 2000-2012

Through the years, the burn injury has been described using a variety of labels. These labels have ranged from one word terms to phrases including degrees of injury or more descriptive terms. A search was conducted relying on a common general internet search engine. After multiple searches varying the keywords, the top 100 searches identified the most prevalent terms or phrases, ranging from the common to the more obscure. The search was repeated using the most prevalent terms or phrases identified in the common internet search engine, focusing on either the title or abstract for all papers indexed in PubMed. This process narrowed the attention to the most common terms or phrases used by the academics in their published work. This work therefore focused on measuring the specific terms being used today and their frequency of use in the peer reviewed papers indexed in the PubMed system. It is difficult to focus on the unique aspects of any given profession when there is confusion surrounding a common vocabulary. By identifying and noting in the academic literature the most commonly used labels, a point of reference can be created for future work. Furthermore, having a common and accurate set of labels that are uniformly applied across the profession is critical for academia to include in training and education programs for physicians, nurses, and paramedical staff.