skip to main content

Title: Autonomous Multi-modality Burn Wound Characterization using Artificial Intelligence
ABSTRACT Introduction

Between 5% and 20% of all combat-related casualties are attributed to burn wounds. A decrease in the mortality rate of burns by about 36% can be achieved with early treatment, but this is contingent upon accurate characterization of the burn. Precise burn injury classification is recognized as a crucial aspect of the medical artificial intelligence (AI) field. An autonomous AI system designed to analyze multiple characteristics of burns using modalities including ultrasound and RGB images is described.

Materials and Methods

A two-part dataset is created for the training and validation of the AI: in vivo B-mode ultrasound scans collected from porcine subjects (10,085 frames), and RGB images manually collected from web sources (338 images). The framework in use leverages an explanation system to corroborate and integrate burn expert’s knowledge, suggesting new features and ensuring the validity of the model. Through the utilization of this framework, it is discovered that B-mode ultrasound classifiers can be enhanced by supplying textural features. More specifically, it is confirmed that statistical texture features extracted from ultrasound frames can increase the accuracy of the burn depth classifier.


The system, with all included features selected using explainable AI, is capable of classifying burn depth with accuracy and F1 average above 80%. Additionally, the segmentation module has been found capable of segmenting with a mean global accuracy greater than 84%, and a mean intersection-over-union score over 0.74.


This work demonstrates the feasibility of accurate and automated burn characterization for AI and indicates that these systems can be improved with additional features when a human expert is combined with explainable AI. This is demonstrated on real data (human for segmentation and porcine for depth classification) and establishes the groundwork for further deep-learning thrusts in the area of burn analysis.

more » « less
Author(s) / Creator(s):
; ; ; ; ; ; ; ; ;
Publisher / Repository:
Oxford University Press
Date Published:
Journal Name:
Military Medicine
Medium: X Size: p. 674-681
["p. 674-681"]
Sponsoring Org:
National Science Foundation
More Like this
  1. Current forest monitoring technologies including satellite remote sensing, manned/piloted aircraft, and observation towers leave uncertainties about a wildfire’s extent, behavior, and conditions in the fire’s near environment, particularly during its early growth. Rapid mapping and real-time fire monitoring can inform in-time intervention or management solutions to maximize beneficial fire outcomes. Drone systems’ unique features of 3D mobility, low flight altitude, and fast and easy deployment make them a valuable tool for early detection and assessment of wildland fires, especially in remote forests that are not easily accessible by ground vehicles. In addition, the lack of abundant, well-annotated aerial datasets – in part due to unmanned aerial vehicles’ (UAVs’) flight restrictions during prescribed burns and wildfires – has limited research advances in reliable data-driven fire detection and modeling techniques. While existing wildland fire datasets often include either color or thermal fire images, here we present (1) a multi-modal UAV-collected dataset of dual-feed side-by-side videos including both RGB and thermal images of a prescribed fire in an open canopy pine forest in Northern Arizona and (2) a deep learning-based methodology for detecting fire and smoke pixels at accuracy much higher than the usual single-channel video feeds. The collected images are labeled to “fire” or “no-fire” frames by two human experts using side-by-side RGB and thermal images to determine the label. To provide context to the main dataset’s aerial imagery, the included supplementary dataset provides a georeferenced pre-burn point cloud, an RGB orthomosaic, weather information, a burn plan, and other burn information. By using and expanding on this guide dataset, research can develop new data-driven fire detection, fire segmentation, and fire modeling techniques. 
    more » « less
  2. Abstract Background

    Lung cancer is the deadliest and second most common cancer in the United States due to the lack of symptoms for early diagnosis. Pulmonary nodules are small abnormal regions that can be potentially correlated to the occurrence of lung cancer. Early detection of these nodules is critical because it can significantly improve the patient's survival rates. Thoracic thin‐sliced computed tomography (CT) scanning has emerged as a widely used method for diagnosing and prognosis lung abnormalities.


    The standard clinical workflow of detecting pulmonary nodules relies on radiologists to analyze CT images to assess the risk factors of cancerous nodules. However, this approach can be error‐prone due to the various nodule formation causes, such as pollutants and infections. Deep learning (DL) algorithms have recently demonstrated remarkable success in medical image classification and segmentation. As an ever more important assistant to radiologists in nodule detection, it is imperative ensure the DL algorithm and radiologist to better understand the decisions from each other. This study aims to develop a framework integrating explainable AI methods to achieve accurate pulmonary nodule detection.


    A robust and explainable detection (RXD) framework is proposed, focusing on reducing false positives in pulmonary nodule detection. Its implementation is based on an explanation supervision method, which uses nodule contours of radiologists as supervision signals to force the model to learn nodule morphologies, enabling improved learning ability on small dataset, and enable small dataset learning ability. In addition, two imputation methods are applied to the nodule region annotations to reduce the noise within human annotations and allow the model to have robust attributions that meet human expectations. The 480, 265, and 265 CT image sets from the public Lung Image Database Consortium and Image Database Resource Initiative (LIDC‐IDRI) dataset are used for training, validation, and testing.


    Using only 10, 30, 50, and 100 training samples sequentially, our method constantly improves the classification performance and explanation quality of baseline in terms of Area Under the Curve (AUC) and Intersection over Union (IoU). In particular, our framework with a learnable imputation kernel improves IoU from baseline by 24.0% to 80.0%. A pre‐defined Gaussian imputation kernel achieves an even greater improvement, from 38.4% to 118.8% from baseline. Compared to the baseline trained on 100 samples, our method shows less drop in AUC when trained on fewer samples. A comprehensive comparison of interpretability shows that our method aligns better with expert opinions.


    A pulmonary nodule detection framework was demonstrated using public thoracic CT image datasets. The framework integrates the robust explanation supervision (RES) technique to ensure the performance of nodule classification and morphology. The method can reduce the workload of radiologists and enable them to focus on the diagnosis and prognosis of the potential cancerous pulmonary nodules at the early stage to improve the outcomes for lung cancer patients.

    more » « less
  3. Abstract

    Though consistently shown to detect mammographically occult cancers, breast ultrasound has been noted to have high false-positive rates. In this work, we present an AI system that achieves radiologist-level accuracy in identifying breast cancer in ultrasound images. Developed on 288,767 exams, consisting of 5,442,907 B-mode and Color Doppler images, the AI achieves an area under the receiver operating characteristic curve (AUROC) of 0.976 on a test set consisting of 44,755 exams. In a retrospective reader study, the AI achieves a higher AUROC than the average of ten board-certified breast radiologists (AUROC: 0.962 AI, 0.924 ± 0.02 radiologists). With the help of the AI, radiologists decrease their false positive rates by 37.3% and reduce requested biopsies by 27.8%, while maintaining the same level of sensitivity. This highlights the potential of AI in improving the accuracy, consistency, and efficiency of breast ultrasound diagnosis.

    more » « less
  4. Simultaneous visualization of the teeth and periodontium is of significant clinical interest for image-based monitoring of periodontal health. We recently reported the application of a dual-modality photoacoustic-ultrasound (PA-US) imaging system for resolving periodontal anatomy and periodontal pocket depths in humans. This work utilized a linear array transducer attached to a stepper motor to generate 3D images via maximum intensity projection. This prior work also used a medical head immobilizer to reduce artifacts during volume rendering caused by motion from the subject (e.g., breathing, minor head movements). However, this solution does not completely eliminate motion artifacts while also complicating the imaging procedure and causing patient discomfort. To address this issue, we report the implementation of an image registration technique to correctly align B-mode PA-US images and generate artifact-free 2D cross-sections. Application of the deshaking technique to PA phantoms revealed 80% similarity to the ground truth when shaking was intentionally applied during stepper motor scans. Images from handheld sweeps could also be deshaken using an LED PA-US scanner. Inex vivoporcine mandibles, pigmentation of the enamel was well-estimated within 0.1 mm error. The pocket depth measured in a healthy human subject was also in good agreement with our prior study. This report demonstrates that a modality-independent registration technique can be applied to clinically relevant PA-US scans of the periodontium to reduce operator burden of skill and subject discomfort while showing potential for handheld clinical periodontal imaging.

    more » « less
  5. Abstract Background

    Improving the prediction ability of a human-machine interface (HMI) is critical to accomplish a bio-inspired or model-based control strategy for rehabilitation interventions, which are of increased interest to assist limb function post neurological injuries. A fundamental role of the HMI is to accurately predict human intent by mapping signals from a mechanical sensor or surface electromyography (sEMG) sensor. These sensors are limited to measuring the resulting limb force or movement or the neural signal evoking the force. As the intermediate mapping in the HMI also depends on muscle contractility, a motivation exists to include architectural features of the muscle as surrogates of dynamic muscle movement, thus further improving the HMI’s prediction accuracy.


    The purpose of this study is to investigate a non-invasive sEMG and ultrasound (US) imaging-driven Hill-type neuromuscular model (HNM) for net ankle joint plantarflexion moment prediction. We hypothesize that the fusion of signals from sEMG and US imaging results in a more accurate net plantarflexion moment prediction than sole sEMG or US imaging.


    Ten young non-disabled participants walked on a treadmill at speeds of 0.50, 0.75, 1.00, 1.25, and 1.50 m/s. The proposed HNM consists of two muscle-tendon units. The muscle activation for each unit was calculated as a weighted summation of the normalized sEMG signal and normalized muscle thickness signal from US imaging. The HNM calibration was performed under both single-speed mode and inter-speed mode, and then the calibrated HNM was validated across all walking speeds.


    On average, the normalized moment prediction root mean square error was reduced by 14.58 % ($$p=0.012$$p=0.012) and 36.79 % ($$p<0.001$$p<0.001) with the proposed HNM when compared to sEMG-driven and US imaging-driven HNMs, respectively. Also, the calibrated models with data from the inter-speed mode were more robust than those from single-speed modes for the moment prediction.


    The proposed sEMG-US imaging-driven HNM can significantly improve the net plantarflexion moment prediction accuracy across multiple walking speeds. The findings imply that the proposed HNM can be potentially used in bio-inspired control strategies for rehabilitative devices due to its superior prediction.

    more » « less