Comparing artificial intelligence guided image assessment to current methods of burn assessment.

Appropriate identification of burn depth and size is paramount. Despite the development of burn depth assessment aids [e.g., laser doppler imaging (LDI)], clinical assessment, which assesses partial thickness burn depth with 67% accuracy, currently remains the most consistent standard of practice. We sought to develop an image-based artificial intelligence system that predicts burn severity and wound margins for use as a triaging tool in thermal injury management. Modified EfficientNet architecture trained by 1684 mobile-device-captured images of different burn depths were previously utilized to create a convoluted neural network (CNN). The CNN was modified to a novel Boundary-Attention Mapping (BAM) algorithm using elements of saliency mapping, which was utilized to recognize the boundaries of burns. For validation, 144 patient charts that included clinical assessment, burn location, total body surface area, and LDI assessment were retrieved for a retrospective study. The clinical images underwent CNN-BAM assessment and were directly compared with the LDI assessment. CNN using a four-level burn severity classification achieved an accuracy of 85% (micro/macro-averaged ROC scores). The CNN-BAM system can successfully highlight burns from surrounding tissue with high confidence. CNN-BAM burn area segmentations attained a 91.6% accuracy, 78.2% sensitivity, and 93.4% specificity, when compared to LDI methodology. Results comparing the CNN-BAM outputs to clinical and LDI assessments have shown a high degree of correlation between the CNN-BAM burn severity predictions to those extrapolated from LDI healing potential (66% agreement). CNN-BAM algorithm gives equivalent burn-depth detection accuracy as LDI with a more economical and accessible application when embedded in a mobile device.

Using Computer Vision and Artificial Intelligence to Track the Healing of Severe Burns.

Burn care management includes assessing the severity of burns accurately, especially distinguishing superficial partial-thickness burns from deep partial-thickness burns, in the context of providing definitive, downstream treatment. Moreover, the healing of the wound in the subacute care setting requires continuous tracking to avoid complications. Artificial intelligence (AI) and computer vision (CV) provide a unique opportunity to build low-cost and accessible tools to classify burn severity and track changes in wound parameters, both in the clinic by physicians and nurses and asynchronously in the remote setting by the patient themselves. Wound assessments can be achieved by AI-CV using the principles of image-guided therapy using high-quality 2D color images. Wound parameters can include wound 2D spatial dimension and the characterization of wound color changes, which demonstrates physiological changes such as the presentation of eschar/necrotic tissue, pustulence, granulation tissue, and scabbing. Here we present the development of AI-CV-based Skin Abnormality Tracking Algorithm pipeline. Additionally, we provide the results on a single localized burn tracked for a 6-week period in the clinic and an additional 2-week period of home monitoring.

Trends 10 years after burn injury: A Burn Model System National Database study.

BACKGROUND: The Burn Specific Health Scale-Brief (BSHS-B) evaluates 9 aspects of health and has been validated globally. Existing reports typically focus on outcomes shortly after injury. The purpose of this study is to determine whether quality of life remains a concern for burn survivors ten years after-injury. METHODS: Cross sectional data of survivors admitted from 1994 to 2006 to four US burn centers were collected in the Burn Model System National Database 10 years after injury. Responses to the items in the nine BSHS-B domains range from 0 to 4. Lower scores indicating poorer quality of life. Median scores are reported and differences were compared using Wilcoxon-Mann-Whitney test. RESULTS: Ten-year survivor injury characteristics suggest a moderate severity of injury. Survivors scored lower in heat sensitivity, affect, body image, and work (median=3.2, 3.6, 2.8, and 3.6, respectively). Affect, body image, and interpersonal scores were significantly lower for females (median=3.1, 2.8, 3.8, respectively) than males [median=3.6, 3.3, 4, respectively (p=0.008, 0.004, 0.022, respectively)]. CONCLUSIONS: Our results suggest certain domains of burn specific health benefit from support at 10 years after injury, and select populations such as females may necessitate additional treatment to restore burn-specific health. These results support that burn injuries represent a chronic condition and long-term medical and psychosocial support may benefit burn survivor recovery.

Feasibility of a Kinect®-based rehabilitation strategy after burn injury.

INTRODUCTION: The advent of consoles that deliver both interactive games and therapy may augment rehabilitation options in burn patients. The Jintronix software combines therapy-specific software and interactive gaming as a form of coaching and records patient performance on the Kinect® platform. Our objective was to determine the feasibility of a set of Jintronix games and therapy modules in hospitalized adult burn patients. METHODS: We conducted a prospective single center feasibility study from August through October 2016. The study enrolled subjects to conduct 1 supervised session with 6 Jintronix modules targeting their burned areas of the body, with an acceptability survey and a safety analysis. We also performed qualitative analysis to detect major themes from free-text responses. RESULTS: We enrolled 20 participants. Eleven (55%) completed all the modules; reasons for incompletion included baseline shoulder abduction pain and poor balance. Participants responded that the activity was comfortable (90%), safe (100%), easy to understand (95%), and improved strength/range of motion (100%). Mean module completion time was 43±10min. Mean pain score was 3.8±2.8 (out of 10) and localized to burned areas. The wall climbing module had 4 episodes of temporary imbalance. Eight (40%) participants recorded fatigue at completion and noted "sweating" or "feeling stretched". Qualitative analysis highlighted that the activity was "fun/cool" and a "good challenge". Negative themes included "inaccurate depth" sensing and "too lengthy" on a specific module. CONCLUSIONS: A Jintronix-based therapy demonstrated good acceptability and safety in hospitalized burn patients. Feedback from this study led to software modifications implemented by the Jintronix company. This feasibility study has informed the design of a prospective randomized controlled trial to determine whether a virtual-environment home rehabilitation strategy improves functional outcomes after burn injury.

Electrical stimulation enhances sensory recovery: a randomized controlled trial.

OBJECTIVE: Brief postsurgical electrical stimulation (ES) has been shown to enhance peripheral nerve regeneration in animal models following axotomy and crush injury. However, whether this treatment is beneficial in humans with sensory nerve injury has not been tested. The goal of this study was to test the hypothesis that ES would enhance sensory nerve regeneration following digital nerve transection compared to surgery alone. METHODS: Patients with complete digital nerve transection underwent epineurial nerve repair. After coaptation of the severed nerve ends, fine wire electrodes were implanted before skin closure. Postoperatively, patients were randomized to receiving either 1 hour of 20Hz continuous ES or sham stimulation in a double-blinded manner. Patients were followed monthly for 6 months by a blinded evaluator to monitor physiological recovery of spatial discrimination, pressure threshold, and quantitative small fiber sensory testing. Functional disability was measured using the Disability of Arm, Shoulder, and Hand questionnaire. RESULTS: A total of 36 patients were recruited, with 18 in each group. Those in the ES group showed consistently greater improvements in all sensory modalities by 5 to 6 months postoperatively compared to the controls. Although there was a trend of greater functional improvements in the ES group, it was not statistically significant (p > 0.01). INTERPRETATION: Postsurgical ES enhanced sensory reinnervation in patients who sustained complete digital nerve transection. The conferred benefits apply to a wide range of sensory functions.

Distal nerve transfers are effective in treating patients with upper trunk obstetrical brachial plexus injuries: an early experience.

BACKGROUND: Current surgical management of obstetrical brachial plexus injury is primary reconstruction with sural nerve grafts. Recently, the nerve-to-nerve transfer technique has been used to treat brachial plexus injury in adults, affording the benefit of distal coaptations that minimize regenerative distance. The purpose of this study was to test the hypothesis that nerve transfers are effective in reconstructing isolated upper trunk obstetrical brachial plexus injuries. METHODS: Ten patients aged 10 to 18 months were treated with three nerve transfers: spinal accessory nerve to the suprascapular nerve for shoulder abduction and external rotation; a radial to axillary nerve for shoulder abduction; and ulnar or median nerve transfer to the musculocutaneous nerve for elbow flexion. Patients were assessed preoperatively and postoperatively using the Active Movement Scale. All patients were followed regularly for up to 2 years. RESULTS: Improvement in elbow and shoulder function was observed between 6 and 24 months. By 6 months, all patients passed the cookie test. At 24 months, shoulder abduction improved from 3.7 ± 0.6 to 5.0 ± 0.5, shoulder external rotation from 1.8 ± 0.4 to 4.3 ± 0.6, shoulder flexion from 3.7 ± 0.5 to 5.4 ± 0.5, elbow flexion from 3.7 ± 0.6 to 6.3 ± 0.2, and forearm supination from 2.1 ± 0.4 to 5.9 ± 0.2. There was no clinically appreciable donor-site morbidity. CONCLUSIONS: Nerve transfers reduced operative times compared with traditional nerve grafting procedures. Those patients showed significant gains in Active Movement Scale score by 24 months postoperatively, comparable to results achieved by nerve grafting. These findings support nerve transfers as a potential alternative treatment option for upper trunk obstetrical brachial plexus injuries. CLINICAL QUESTION/LEVEL OF EVIDENCE: Therapeutic, IV.