Using an Ultrasound Tissue Phantom Model for Hybrid Training of Deep Learning Models for Shrapnel Detection.

Tissue phantoms are important for medical research to reduce the use of animal or human tissue when testing or troubleshooting new devices or technology. Development of machine-learning detection tools that rely on large ultrasound imaging data sets can potentially be streamlined with high quality phantoms that closely mimic important features of biological tissue. Here, we demonstrate how an ultrasound-compliant tissue phantom comprised of multiple layers of gelatin to mimic bone, fat, and muscle tissue types can be used for machine-learning training. This tissue phantom has a heterogeneous composition to introduce tissue level complexity and subject variability in the tissue phantom. Various shrapnel types were inserted into the phantom for ultrasound imaging to supplement swine shrapnel image sets captured for applications such as deep learning algorithms. With a previously developed shrapnel detection algorithm, blind swine test image accuracy reached more than 95% accuracy when training was comprised of 75% tissue phantom images, with the rest being swine images. For comparison, a conventional MobileNetv2 deep learning model was trained with the same training image set and achieved over 90% accuracy in swine predictions. Overall, the tissue phantom demonstrated high performance for developing deep learning models for ultrasound image classification.

Hardware-in-Loop Comparison of Physiological Closed-Loop Controllers for the Autonomous Management of Hypotension.

Trauma and hemorrhage are leading causes of death and disability worldwide in both civilian and military contexts. The delivery of life-saving goal-directed fluid resuscitation can be difficult to provide in resource-constrained settings, such as in forward military positions or mass-casualty scenarios. Automated solutions for fluid resuscitation could bridge resource gaps in these austere settings. While multiple physiological closed-loop controllers for the management of hypotension have been proposed, to date there is no consensus on controller design. Here, we compare the performance of four controller types-decision table, single-input fuzzy logic, dual-input fuzzy logic, and proportional-integral-derivative using a previously developed hardware-in-loop test platform where a range of hemorrhage scenarios can be programmed. Controllers were compared using traditional controller performance metrics, but conclusions were difficult to draw due to inconsistencies across the metrics. Instead, we propose three aggregate metrics that reflect the target intensity, stability, and resource efficiency of a controller, with the goal of selecting controllers for further development. These aggregate metrics identify a dual-input, fuzzy-logic-based controller as the preferred combination of intensity, stability, and resource efficiency within this use case. Based on these results, the aggressively tuned dual-input fuzzy logic controller should be considered a priority for further development.

Training Ultrasound Image Classification Deep-Learning Algorithms for Pneumothorax Detection Using a Synthetic Tissue Phantom Apparatus.

Ultrasound (US) imaging is a critical tool in emergency and military medicine because of its portability and immediate nature. However, proper image interpretation requires skill, limiting its utility in remote applications for conditions such as pneumothorax (PTX) which requires rapid intervention. Artificial intelligence has the potential to automate ultrasound image analysis for various pathophysiological conditions. Training models require large data sets and a means of troubleshooting in real-time for ultrasound integration deployment, and they also require large animal models or clinical testing. Here, we detail the development of a dynamic synthetic tissue phantom model for PTX and its use in training image classification algorithms. The model comprises a synthetic gelatin phantom cast in a custom 3D-printed rib mold and a lung mimicking phantom. When compared to PTX images acquired in swine, images from the phantom were similar in both PTX negative and positive mimicking scenarios. We then used a deep learning image classification algorithm, which we previously developed for shrapnel detection, to accurately predict the presence of PTX in swine images by only training on phantom image sets, highlighting the utility for a tissue phantom for AI applications.

Evaluation of an Object Detection Algorithm for Shrapnel and Development of a Triage Tool to Determine Injury Severity.

Emergency medicine in austere environments rely on ultrasound imaging as an essential diagnostic tool. Without extensive training, identifying abnormalities such as shrapnel embedded in tissue, is challenging. Medical professionals with appropriate expertise are limited in resource-constrained environments. Incorporating artificial intelligence models to aid the interpretation can reduce the skill gap, enabling identification of shrapnel, and its proximity to important anatomical features for improved medical treatment. Here, we apply a deep learning object detection framework, YOLOv3, for shrapnel detection in various sizes and locations with respect to a neurovascular bundle. Ultrasound images were collected in a tissue phantom containing shrapnel, vein, artery, and nerve features. The YOLOv3 framework, classifies the object types and identifies the location. In the testing dataset, the model was successful at identifying each object class, with a mean Intersection over Union and average precision of 0.73 and 0.94, respectively. Furthermore, a triage tool was developed to quantify shrapnel distance from neurovascular features that could notify the end user when a proximity threshold is surpassed, and, thus, may warrant evacuation or surgical intervention. Overall, object detection models such as this will be vital to compensate for lack of expertise in ultrasound interpretation, increasing its availability for emergency and military medicine.

Development and Characterization of an Ex Vivo Testing Platform for Evaluating Automated Central Vascular Access Device Performance.

Access to the central vasculature is critical for hemodynamic monitoring and for delivery of life-saving therapeutics during emergency medicine and battlefield trauma situations but requires skill often unavailable in austere environments. Automated central vascular access devices (ACVADs) using ultrasound and robotics are being developed. Here, we present an ex vivo lower-body porcine model as a testing platform for evaluation of vascular devices and compare its features to commercially available platforms. While the commercially available trainers were simpler to set-up and use, the scope of their utility was limited as they were unable to provide realistic anatomic, physiologic, and sonographic properties that were provided by the ex vivo model. However, the ex vivo model was more cumbersome to set-up and use. Overall, both have a place in the development and evaluation pipeline for ACVADs before testing on live animals, thus accelerating product development and translation.

An Automated Hardware-in-Loop Testbed for Evaluating Hemorrhagic Shock Resuscitation Controllers.

Hemorrhage remains a leading cause of death, with early goal-directed fluid resuscitation being a pillar of mortality prevention. While closed-loop resuscitation can potentially benefit this effort, development of these systems is resource-intensive, making it a challenge to compare infusion controllers and respective hardware within a range of physiologically relevant hemorrhage scenarios. Here, we present a hardware-in-loop automated testbed for resuscitation controllers (HATRC) that provides a simple yet robust methodology to evaluate controllers. HATRC is a flow-loop benchtop system comprised of multiple PhysioVessels which mimic pressure-volume responsiveness for different resuscitation infusates. Subject variability and infusate switching were integrated for more complex testing. Further, HATRC can modulate fluidic resistance to mimic arterial resistance changes after vasopressor administration. Finally, all outflow rates are computer-controlled, with rules to dictate hemorrhage, clotting, and urine rates. Using HATRC, we evaluated a decision-table controller at two sampling rates with different hemorrhage scenarios. HATRC allows quantification of twelve performance metrics for each controller configuration and scenario, producing heterogeneous results and highlighting the need for controller evaluation with multiple hemorrhage scenarios. In conclusion, HATRC can be used to evaluate closed-loop controllers through user-defined hemorrhage scenarios while rating their performance. Extensive controller troubleshooting using HATRC can accelerate product development and subsequent translation.

Closed-Loop Controlled Fluid Administration Systems: A Comprehensive Scoping Review.

Physiological Closed-Loop Controlled systems continue to take a growing part in clinical practice, offering possibilities of providing more accurate, goal-directed care while reducing clinicians' cognitive and task load. These systems also provide a standardized approach for the clinical management of the patient, leading to a reduction in care variability across multiple dimensions. For fluid management and administration, the advantages of closed-loop technology are clear, especially in conditions that require precise care to improve outcomes, such as peri-operative care, trauma, and acute burn care. Controller design varies from simplistic to complex designs, based on detailed physiological models and adaptive properties that account for inter-patient and intra-patient variability; their maturity level ranges from theoretical models tested in silico to commercially available, FDA-approved products. This comprehensive scoping review was conducted in order to assess the current technological landscape of this field, describe the systems currently available or under development, and suggest further advancements that may unfold in the coming years. Ten distinct systems were identified and discussed.

Development of a Modular Tissue Phantom for Evaluating Vascular Access Devices.

Central vascular access (CVA) may be critical for trauma care and stabilizing the casualty. However, it requires skilled personnel, often unavailable during remote medical situations and combat casualty care scenarios. Automated CVA medical devices have the potential to make life-saving therapeutics available in these resource-limited scenarios, but they must be properly designed. Unfortunately, currently available tissue phantoms are inadequate for this use, resulting in delayed product development. Here, we present a tissue phantom that is modular in design, allowing for adjustable flow rate, circulating fluid pressure, vessel diameter, and vessel positions. The phantom consists of a gelatin cast using a 3D-printed mold with inserts representing vessels and bone locations. These removable inserts allow for tubing insertion which can mimic normal and hypovolemic flow, as well as pressure and vessel diameters. Trauma to the vessel wall is assessed using quantification of leak rates from the tubing after removal from the model. Lastly, the phantom can be adjusted to swine or human anatomy, including modeling the entire neurovascular bundle. Overall, this model can better recreate severe hypovolemic trauma cases and subject variability than commercial CVA trainers and may potentially accelerate automated CVA device development.

Evaluation of a Proportional-Integral-Derivative Controller for Hemorrhage Resuscitation Using a Hardware-in-Loop Test Platform.

Hemorrhage is a leading cause of preventable death in trauma, which can often be avoided with proper fluid resuscitation. Fluid administration can be cognitive-demanding for medical personnel as the rates and volumes must be personalized to the trauma due to variations in injury severity and overall fluid responsiveness. Thus, automated fluid administration systems are ideal to simplify hemorrhagic shock resuscitation if properly designed for a wide range of hemorrhage scenarios. Here, we highlight the development of a proportional-integral-derivative (PID) controller using a hardware-in-loop test platform. The controller relies only on an input data stream of arterial pressure and a target pressure; the PID controller then outputs infusion rates to stabilize the subject. To evaluate PID controller performance with more than 10 controller metrics, the hardware-in-loop platform allowed for 11 different trauma-relevant hemorrhage scenarios for the controller to resuscitate against. Overall, the two controller configurations performed uniquely for the scenarios, with one reaching the target quicker but often overshooting, while the other rarely overshot the target but failed to reach the target during severe hemorrhage. In conclusion, PID controllers have the potential to simplify hemorrhage resuscitation if properly designed and evaluated, which can be accomplished with the test platform shown here.

An image classification deep-learning algorithm for shrapnel detection from ultrasound images.

Ultrasound imaging is essential for non-invasively diagnosing injuries where advanced diagnostics may not be possible. However, image interpretation remains a challenge as proper expertise may not be available. In response, artificial intelligence algorithms are being investigated to automate image analysis and diagnosis. Here, we highlight an image classification convolutional neural network for detecting shrapnel in ultrasound images. As an initial application, different shrapnel types and sizes were embedded first in a tissue mimicking phantom and then in swine thigh tissue. The algorithm architecture was optimized stepwise by minimizing validation loss and maximizing F1 score. The final algorithm design trained on tissue phantom image sets had an F1 score of 0.95 and an area under the ROC curve of 0.95. It maintained higher than a 90% accuracy for each of 8 shrapnel types. When trained only on swine image sets, the optimized algorithm format had even higher metrics: F1 and area under the ROC curve of 0.99. Overall, the algorithm developed resulted in strong classification accuracy for both the tissue phantom and animal tissue. This framework can be applied to other trauma relevant imaging applications such as internal bleeding to further simplify trauma medicine when resources and image interpretation are scarce.

Comparison of Ultrasound Image Classifier Deep Learning Algorithms for Shrapnel Detection.

Ultrasound imaging is essential in emergency medicine and combat casualty care, oftentimes used as a critical triage tool. However, identifying injuries, such as shrapnel embedded in tissue or a pneumothorax, can be challenging without extensive ultrasonography training, which may not be available in prolonged field care or emergency medicine scenarios. Artificial intelligence can simplify this by automating image interpretation but only if it can be deployed for use in real time. We previously developed a deep learning neural network model specifically designed to identify shrapnel in ultrasound images, termed ShrapML. Here, we expand on that work to further optimize the model and compare its performance to that of conventional models trained on the ImageNet database, such as ResNet50. Through Bayesian optimization, the model's parameters were further refined, resulting in an F1 score of 0.98. We compared the proposed model to four conventional models: DarkNet-19, GoogleNet, MobileNetv2, and SqueezeNet which were down-selected based on speed and testing accuracy. Although MobileNetv2 achieved a higher accuracy than ShrapML, there was a tradeoff between accuracy and speed, with ShrapML being 10× faster than MobileNetv2. In conclusion, real-time deployment of algorithms such as ShrapML can reduce the cognitive load for medical providers in high-stress emergency or miliary medicine scenarios.

Development of the PhysioVessel: a customizable platform for simulating physiological fluid resuscitation.

Uncontrolled hemorrhage is a leading cause of death in trauma situations. Developing solutions to automate hemorrhagic shock resuscitation may improve the outcomes for trauma patients. However, testing and development of automated solutions to address critical care interventions, oftentimes require extensive large animal studies for even initial troubleshooting. The use of accurate laboratory or in-silico models may provide a way to reduce the need for large animal datasets. Here, a tabletop model, for use in the development of fluid resuscitation with physiologically relevant pressure-volume responsiveness for high throughput testing, is presented. The design approach shown can be applied to any pressure-volume dataset through a process of curve-fitting, 3D modeling, and fabrication of a fluid reservoir shaped to the precise curve fit. Two case studies are presented here based on different resuscitation fluids: whole blood and crystalloid resuscitation. Both scenarios were derived from data acquired during porcine hemorrhage studies, used a pressure-volume curve to design and fabricate a 3D model, and evaluated to show that the test platform mimics the physiological data. The vessels produced based on data collected from pigs infused with whole blood and crystalloid were able to reproduce normalized pressure-volume curves within one standard deviation of the porcine data with mean residual differences of 0.018 and 0.016, respectively. This design process is useful for developing closed-loop algorithms for resuscitation and can simplify initial testing of technologies for this life-saving medical intervention.

Development and Characterization of a Self-Tightening Tourniquet System.

Uncontrolled hemorrhage remains a leading cause of death in both emergency and military medicine. Tourniquets are essential to stopping hemorrhage in these scenarios, but they suffer from subjective, inconsistent application. Here, we demonstrate how tourniquet application can be automated using sensors and computer algorithms. The auto-tourniquet self-tightens until blood pressure oscillations are no longer registered by the pressure sensor connected to the pneumatic pressure cuff. The auto-tourniquet's performance in stopping the bleed was comparable to manual tourniquet application, but the time required to fully occlude the bleed was longer. Application of the tourniquet was significantly smoother, and less variable, for the automatic tourniquet compared to manual tourniquet application. This proof-of-concept study highlights how automated tourniquets can be integrated with sensors to provide a much more consistent application and use compared to manual application, even in controlled, low stress testing conditions. Future work will investigate different sensors and tourniquets to improve the application time and repeatability.

Supervisory Algorithm for Autonomous Hemodynamic Management Systems.

Future military conflicts will require new solutions to manage combat casualties. The use of automated medical systems can potentially address this need by streamlining and augmenting the delivery of medical care in both emergency and combat trauma environments. However, in many situations, these systems may need to operate in conjunction with other autonomous and semi-autonomous devices. Management of complex patients may require multiple automated systems operating simultaneously and potentially competing with each other. Supervisory controllers capable of harmonizing multiple closed-loop systems are thus essential before multiple automated medical systems can be deployed in managing complex medical situations. The objective for this study was to develop a Supervisory Algorithm for Casualty Management (SACM) that manages decisions and interplay between two automated systems designed for management of hemorrhage control and resuscitation: an automatic extremity tourniquet system and an adaptive resuscitation controller. SACM monitors the required physiological inputs for both systems and synchronizes each respective system as needed. We present a series of trauma experiments carried out in a physiologically relevant benchtop circulatory system in which SACM must recognize extremity or internal hemorrhage, activate the corresponding algorithm to apply a tourniquet, and then resuscitate back to the target pressure setpoint. SACM continues monitoring after the initial stabilization so that additional medical changes can be quickly identified and addressed, essential to extending automation algorithms past initial trauma resuscitation into extended monitoring. Overall, SACM is an important step in transitioning automated medical systems into emergency and combat trauma situations. Future work will address further interplay between these systems and integrate additional medical systems.

Assessment of Commercial Off-the-Shelf Tissue Adhesives for Sealing Military-Relevant Corneal Perforation Injuries.

INTRODUCTION: Open-globe ocular injuries have increased in frequency in recent combat operations due to increased use of explosive weaponry. Unfortunately, open-globe injuries have one of the worst visual outcomes for the injured warfighter, often resulting in permanent loss of vision. To improve visual recovery, injuries need to be stabilized quickly following trauma, in order to restore intraocular pressure and create a watertight seal. Here, we assess four off-the-shelf (OTS), commercially available tissue adhesives for their ability to seal military-relevant corneal perforation injuries (CPIs). MATERIALS AND METHODS: Adhesives were assessed using an anterior segment inflation platform and a previously developed high-speed benchtop corneal puncture model, to create injuries in porcine eyes. After injury, adhesives were applied and injury stabilization was assessed by measuring outflow rate, ocular compliance, and burst pressure, followed by histological analysis. RESULTS: Tegaderm dressings and Dermabond skin adhesive most successfully sealed injuries in preliminary testing. Across a range of injury sizes and shapes, Tegaderm performed well in smaller injury sizes, less than 2 mm in diameter, but inadequately sealed large or complex injuries. Dermabond created a watertight seal capable of maintaining ocular tissue at physiological intraocular pressure for almost all injury shapes and sizes. However, application of the adhesive was inconsistent. Histologically, after removal of the Dermabond skin adhesive, the corneal epithelium was removed and oftentimes the epithelium surface penetrated into the wound and was adhered to inner stromal tissue. CONCLUSIONS: Dermabond can stabilize a wide range of CPIs; however, application is variable, which may adversely impact the corneal tissue. Without addressing these limitations, no OTS adhesive tested herein can be directly translated to CPIs. This highlights the need for development of a biomaterial product to stabilize these injuries without causing ocular damage upon removal, thus improving the poor vision prognosis for the injured warfighter.

Anterior Segment Organ Culture Platform for Tracking Open Globe Injuries and Therapeutic Performance.

Open globe injuries have poor visual outcomes, often resulting in permanent loss of vision. This is partly due to an extended delay between injury and medical intervention in rural environments and military medicine applications where ophthalmic care is not readily available. Untreated injuries are susceptible to infection after the eye has lost its watertight seal, as well as loss of tissue viability due to intraocular hypotension. Therapeutics to temporarily seal open globe injuries, if properly developed, may be able to restore intraocular pressure and prevent infection until proper ophthalmic care is possible. To facilitate product development, detailed here is the use of an anterior segment organ culture open globe injury platform for tracking therapeutic performance for at least 72 h post-injury. Porcine anterior segment tissue can be maintained in custom-designed organ culture dishes and held at physiological intraocular pressure. Puncture injuries can be created with a pneumatic-powered system capable of generating injury sizes up to 4.5 mm in diameter, similar to military-relevant injury sizes. Loss of intraocular pressure can be observed for 72 h post-injury confirming proper injury induction and loss of the eye's watertight seal. Therapeutic performance can be tracked by application to the eye after injury induction and then tracking intraocular pressure for multiple days. Further, the anterior segment injury model is applicable to widely used methods for functionally and biologically tracking anterior segment physiology, such as assessing transparency, ocular mechanics, corneal epithelium health, and tissue viability. Overall, the method described here is a necessary next step toward developing biomaterial therapeutics for temporarily sealing open globe injuries when ophthalmic care is not readily available.

Characterization of an anterior segment organ culture model for open globe injuries.

Open-globe injuries have poor visual outcomes and have increased in frequency. The current standard of care is inadequate, and a therapeutic is needed to stabilize the injury until an ophthalmic specialist is reached. Unfortunately, current models or test platforms for open-globe injuries are insufficient. Here, we develop and characterize an open-globe injury model using an anterior segment organ-culture platform that allows therapeutic assessment for up to 72 h post-injury. Anterior segments maintained in organ culture were kept at physiological intraocular pressure throughout, and puncture injuries were created using a novel pneumatic-powered system. This system can create high-speed, military-relevant injuries up to 4.5 mm in diameter through the cornea. From intraocular pressure readings, we confirmed a loss of pressure across the 72 h after open-globe injury. Proof-of-concept studies with a Dermabond tissue adhesive were performed to show how this model system could track therapeutic performance for 72 h. Overall, the organ-culture platform was found to be a suitable next step towards modeling open-globe injuries and assessing wound closure over the critical 72 h post-injury. With improved models such as this, novel biomaterial therapeutics development can be accelerated, improving care, and, thus, improving the prognosis for the patients.

An Open-Globe Porcine Injury Platform for Assessing Therapeutics and Characterizing Biological Effects.

Open-globe injuries can result in permanent vision loss, partly due to extended delays between injury and medical intervention. Even with early intervention, the management of open-globe injuries remains a challenge for ophthalmologists, mostly due to inadequate or suboptimal current therapies. To aid in the development of novel therapeutics and track toxicological and pathophysiological changes, this article details an open-globe injury platform capable of inducing injuries in enucleated porcine eyes. The injury platform relies on a high-speed solenoid device to mimic explosive injury scenarios, allowing for large, complex injury shapes and sizes that are often observed in casualties and are more difficult to treat. The system can be implemented with precise computer control of the injury mechanism to allow for more complex setups. Also, the system can make use of real-time intraocular pressure measurement to track changes during injury induction and to assess therapeutic efficacy for restoring intraocular pressure and the integrity of the eye. These protocols will assist with implementation of the injury model in prospective laboratories seeking to develop therapeutics or studying biological changes that occur from this type of traumatic injury. Published 2020. U.S. Government. Basic Protocol 1: Preparing gelatin molds and porcine eye tissue Basic Protocol 2: Creating an open-globe injury using a solenoid device Alternate Protocol 1: Constructing a computer-controlled system for open-globe injury Alternate Protocol 2: Constructing a pressure measurement system for tracking intraocular pressure Support Protocol 1: Assessing ocular compliance in porcine eyes Support Protocol 2: Assessing outflow rate from the anterior chamber Support Protocol 3: Assessing burst pressure in porcine eyes.

Trojan Horse for Light-Triggered Bifurcated Production of Singlet Oxygen and Fenton-Reactive Iron within Cancer Cells.

Traditional photodynamic therapy for cancer relies on dye-photosensitized generation of singlet oxygen. However, therapeutically effective singlet oxygen generation requires well-oxygenated tissues, whereas many tumor environments tend to be hypoxic. We describe a platform for targeted enhancement of photodynamic therapy that produces singlet oxygen in oxygenated environments and hydroxyl radical, which is typically regarded as the most toxic reactive oxygen species, in hypoxic environments. The 24-subunit iron storage protein bacterioferritin (Bfr) has the unique property of binding 12 heme groups in its protein shell. We inserted the isostructural photosensitizer, zinc(II) protoporphyrin IX (ZnP), in place of the hemes and extended the surface-exposed N-terminal ends of the Bfr subunits with a peptide targeting a receptor that is hyperexpressed on the cell surface of many tumors and tumor vasculature. We then loaded the inner cavity with ∼2500 irons as a ferric oxyhydroxide polymer and finally conjugated 2 kDa polyethylene glycol to the outer surface. We showed that the inserted ZnP photosensitizes generation of both singlet oxygen and the hydroxyl radical, the latter via the reaction of photoreleased ferrous iron with hydrogen peroxide. This targeted iron-loaded ZnP-Bfr construct was endocytosed by C32 melanoma cells and localized to lysosomes. Irradiating the treated cells with light at wavelengths overlapping the ZnP Soret absorption band induced photosensitized intracellular Fe2+ release and substantial lowering of cell viability. This targeted, light-triggered production of intracellular singlet oxygen and Fenton-reactive iron could potentially be developed into a phototherapeutic adjunct for many types of cancers.