5 ways AI stands to advance the state of burn care

AI has “remarkable potential” to improve diagnostic accuracy, care efficiency and workflow optimization in the surgical subspecialty of burn care.

So conclude researchers in the U.K who systematically selected and analyzed 46 relevant studies published in English.

The team concentrated on research testing the use of AI and machine learning specifically in burn care, breaking out such utilization aspects as clinical applications, algorithms, outcomes and validation methods.

The literature review was led by plastic and reconstructive surgeon Francisco Serra Moura, MD, of Norfolk and Norwich University Hospital and is running in Burns & Trauma.

The authors describe improvements that AI stands to help make in numerous areas of care, as documented in the reviewed studies. Betterments could include:  

1. Prevention. AI can assist clinical teams in identifying and risk-stratifying individuals before inflammatory accidents even happen, Moura and colleagues suggest. Predictive components might include occupation, substance abuse, self-harm or other socioeconomic factors that “place patients at an increased risk of sustaining a burn injury,” they write. “An algorithm able to identify this higher-risk group may prompt more regular community review or workplace inspection to mitigate any given risk.”

2. Pre-hospital care. In many regions, specialized burn-care personnel and equipment are in short supply. An automated system could help identify patients whose recovery will require dedicated burn-care expertise. In cases, AI technology “can be useful to improve the standard of burn care for patients where burn experts may not be readily available,” the authors point out.

3. Post-operative care. Machine learning can help care teams better predict surgical-site infections in immunosuppressed burn patients. Moura and co-authors cite a recent study showing AI outperforming conventional means of infection prediction by “building non-linear models that incorporate multiple data sources, including diagnoses, treatments and laboratory values. Consequently, this can influence subsequent treatments including antibiotic therapy and dressing changes.”

4. Rehabilitation. Digital platforms for monitoring patient-reported outcomes could generate unique algorithms to screen symptoms. This step would not only alert the care team but also automatically generate a customized patient care plan. Two studies have shown how merging AI with manual patient input “offers the potential to improve outcomes and quality of care as well as evaluating the efficacy of treatments.”

5. Medical education. Virtual reality combined with simulation based in real-world data will soon be “fundamental” to burn-care training, the authors predict. “Many adverse events in burns emergency are a consequence of nontechnical skills, such as communication, leadership and teamwork,” they write. “[E]nhanced data from significant adverse events has the potential to make an impact on the acquisition of nontechnical skills.”

For such potential to be reached, extensive education and training must be aimed at the clinical workforce and, on some scale, offered to the general public, the authors suggest.

Also needed is “the cultivation of a cross-disciplinary approach that includes data scientists, computer scientists and engineers, in addition to pharmacists, nurses, physiotherapists, psychologists and doctors, to generate meaningful interpretation of data.”


With the large volume of [available] burn data, AI can assist clinicians in evaluating burn surface, diagnose burn depth, the need for surgery or other therapies, guide fluid resuscitation, and predict complications and prognosis with a high degree of accuracy.

Education and encouragement of AI technologies are key to delivering burn care on a far more rational, efficient and tailored basis. However, it cannot replace the art of caring.”

The study is available in full for free.

Around the web

The Palo Alto giant used exams from nearly 250,000 patients to upgrade its already robust algorithm.

Exams performed using the deep learning-based reconstruction tool also maintained high image quality, experts reported Wednesday.

Stratifying exams according to risk can reduce unnecessary imaging and downstream costs of care, Hawaiian researchers reported in Radiology.

Trimed Popup
Trimed Popup