Sofia I. Hernandez Torres’s research while affiliated with United States Army Institute of Surgical Research and other places

What is this page?


This page lists works of an author who doesn't have a ResearchGate profile or hasn't added the works to their profile yet. It is automatically generated from public (personal) data to further our legitimate goal of comprehensive and accurate scientific recordkeeping. If you are this author and want this page removed, please let us know.

Publications (15)


Real-Time Deployment of Ultrasound Image Interpretation AI Models for Emergency Medicine Triage Using a Swine Model
  • Article
  • Full-text available

January 2025

·

15 Reads

Technologies

Sofia I. Hernandez Torres

·

Lawrence Holland

·

Theodore Winter

·

[...]

·

Eric J. Snider

Ultrasound imaging is commonly used for medical triage in both civilian and military emergency medicine sectors. One specific application is the eFAST, or the extended focused assessment with sonography in trauma exam, where pneumothorax, hemothorax, or abdominal hemorrhage injuries are identified. However, the diagnostic accuracy of an eFAST exam depends on obtaining proper scans and making quick interpretation decisions to evacuate casualties or administer necessary interventions. To improve ultrasound interpretation, we developed AI models to identify key anatomical structures at eFAST scan sites, simplifying image acquisition by assisting with proper probe placement. These models plus image interpretation diagnostic models were paired with two real-time eFAST implementations. The first implementation was a manual AI-driven ultrasound eFAST tool that used guidance models to select correct frames prior to making any diagnostic predictions. The second implementation was a robotic imaging platform capable of providing semi-autonomous image acquisition combined with diagnostic image interpretation. We highlight the use of both real-time approaches in a swine injury model and compare their performance of this emergency medicine application. In conclusion, AI can be deployed in real time to provide rapid triage decisions, lowering the skill threshold for ultrasound imaging at or near the point of injury.

Download

Figure 2. Summary of Wearable Healthcare Devices (WHDs) for Electroencephalogram (EEG) Measurement. Potential uses in the prehospital setting, clinical decision support applications for the sensor technology, and current technology limitations are summarized.
Figure 3. Summary of Wearable Healthcare Devices (WHDs) for Photoplethysmography (PPG) Measurement. Potential uses in the prehospital setting, clinical decision support applications for the sensor technology, and current technology limitations are summarized.
Figure 4. Summary of Wearable Healthcare Devices (WHDs) for Medical Imaging. Potential uses in the prehospital setting, CDS applications for the sensor technology, and current technology limitations are summarized.
Figure 5. Summary of Wearable Healthcare Devices (WHDs) for Chemical Sensing. Potential uses in the prehospital setting, clinical decision support applications for the sensor technology, and current technology limitations are summarized.
Figure 6. Summary of Wearable Healthcare Devices (WHDs) for Electrocardiogram (ECG) Measurement. Potential uses in the prehospital setting, clinical decision support applications for the sensor technology, and current technology limitations are summarized.

+2

Overview of Wearable Healthcare Devices for Clinical Decision Support in the Prehospital Setting

December 2024

·

31 Reads

·

1 Citation

Sensors

Prehospital medical care is a major challenge for both civilian and military situations as resources are limited, yet critical triage and treatment decisions must be rapidly made. Prehospital medicine is further complicated during mass casualty situations or remote applications that require more extensive medical treatments to be monitored. It is anticipated on the future battlefield where air superiority will be contested that prolonged field care will extend to as much 72 h in a prehospital environment. Traditional medical monitoring is not practical in these situations and, as such, wearable sensor technology may help support prehospital medicine. However, sensors alone are not sufficient in the prehospital setting where limited personnel without specialized medical training must make critical decisions based on physiological signals. Machine learning-based clinical decision support systems can instead be utilized to interpret these signals for diagnosing injuries, making triage decisions, or driving treatments. Here, we summarize the challenges of the prehospital medical setting and review wearable sensor technology suitability for this environment, including their use with medical decision support triage or treatment guidance options. Further, we discuss recommendations for wearable healthcare device development and medical decision support technology to better support the prehospital medical setting. With further design improvement and integration with decision support tools, wearable healthcare devices have the potential to simplify and improve medical care in the challenging prehospital environment.


Figure 8. Summary of Vu-Path™ performance and discussion results from usability study. (A) Average time to find vessel and perform successful cannulation (n = 20 participants). (B) Distance and insertion angles for all successful insertion attempts. (C-F) Consensus answers from discussion topics related to (C) most useful feature of the device, (D) most common issue with the device, (E) recommended improvements, and (F) recommended experience level for those using the device. Results for (C-E) highlight opinions from at least 15% of the participants, or three out of twenty of the participants.
Evaluation of a Semi-Automated Ultrasound Guidance System for Central Vascular Access

Bioengineering

Hemorrhage remains a leading cause of death in both military and civilian trauma settings. Oftentimes, the control and treatment of hemorrhage requires central vascular access and well-trained medical personnel. Automated technology is being developed that can lower the skill threshold for life-saving interventions. Here, we conduct independent evaluation testing of one such device, the Vu-Path™ Ultrasound Guidance system, or Vu-Path™. The device was designed to simplify needle insertion using a needle holder that ensures the needle is within the ultrasound field of view during its insertion into tissue, along with guidance lines shown on the user interface. We evaluated the performance of this device in a range of laboratory, animal, and human testing platforms. Overall, the device had a high success rate, achieving an 83% insertion accuracy in live animal testing across both normal and hypotensive blood pressures. Vu-Path™ was faster than manual, ultrasound-guided needle insertion and was nearly 1.5 times quicker for arterial and 2.3 times quicker for venous access. Human usability feedback highlighted that 80% of the participants would use this device for central line placement. Study users noted that the guidance lines and small form factor were useful design features. However, issues were raised regarding the needle insertion angle being too steep, with potential positioning challenges as the needle remains fixed to the ultrasound probe. Regardless, 75% of the participants believed that personnel with any level of clinical background could use the device for central vascular access. Overall, Vu-Path™ performed well across a range of testing situations, and potential design improvements were noted. With adjustments to the device, central vascular access can be made more accessible on battlefields in the future.


In vivo evaluation of an adaptive resuscitation controller using whole blood and crystalloid infusates for hemorrhagic shock

November 2024

·

9 Reads

·

2 Citations

Introduction Hemorrhage remains the leading cause of preventable death on the battlefield. The most effective means to increase survivability is early hemorrhage control and fluid resuscitation. Unfortunately, fluid resuscitation requires constant adjustments to ensure casualty is properly managed, which is often not feasible in the pre-hospital setting. In this study, we showed how an adaptive closed-loop controller for hemorrhage resuscitation can be used to automate hemodynamic management using a swine hemorrhagic shock injury model. Methods The adaptive resuscitation controller (ARC) was previously developed to track pressure–volume responsiveness in real time and adjust its infusion rate to reach the target mean arterial pressure (MAP). Swine while maintained under a surgical plane of anesthesia and analgesia underwent a splenectomy, followed by two hemorrhage and resuscitation events. For the first resuscitation event, hemorrhage was induced to reduce the MAP to 35 mmHg until arterial lactate reached 4 mmol/L. The ARC system then infused whole blood (WB) to reach the target MAP and maintained the subject using crystalloids for 120 min. For the second resuscitation event, the subjects were hemorrhaged again but resuscitated using only crystalloid infusion to reach the target MAP and 120-min maintenance. Results The ARC was effective at WB resuscitation, reaching the target MAP in 2.0 ± 1.0 min. The median performance error was 1.1% ± 4.6%, and target overshoot was 14.4% ± 7.0% of the target MAP. The ARC maintained all animals throughout the 120 min maintenance period. For the second crystalloid-based resuscitation, ARC required a longer time to reach the target MAP, at an average rise time of 4.3 ± 4.0 min. However, target overshoot was reduced to 8.4% ± 7.3% of the target MAP. Much higher flow rates were required to maintain the target MAP during the second resuscitation event than during the first resuscitation event. Discussion The ARC was able to rapidly reach and maintain the target MAP effectively. However, this sometimes required large volumes of fluid as the ARC’s only goal was to reach the target MAP. Further clinical insight is needed regarding the preferred aggression level to achieve the target MAP. In conclusion, the ARC was successful in its programmed objective of reaching and maintaining the target MAP for extended periods of time in vivo, a critical next step toward improving hemorrhage treatment in the pre-hospital environment.


Deep learning models for interpretation of point of care ultrasound in military working dogs

June 2024

·

25 Reads

·

1 Citation

Introduction Military working dogs (MWDs) are essential for military operations in a wide range of missions. With this pivotal role, MWDs can become casualties requiring specialized veterinary care that may not always be available far forward on the battlefield. Some injuries such as pneumothorax, hemothorax, or abdominal hemorrhage can be diagnosed using point of care ultrasound (POCUS) such as the Global FAST® exam. This presents a unique opportunity for artificial intelligence (AI) to aid in the interpretation of ultrasound images. In this article, deep learning classification neural networks were developed for POCUS assessment in MWDs. Methods Images were collected in five MWDs under general anesthesia or deep sedation for all scan points in the Global FAST® exam. For representative injuries, a cadaver model was used from which positive and negative injury images were captured. A total of 327 ultrasound clips were captured and split across scan points for training three different AI network architectures: MobileNetV2, DarkNet-19, and ShrapML. Gradient class activation mapping (GradCAM) overlays were generated for representative images to better explain AI predictions. Results Performance of AI models reached over 82% accuracy for all scan points. The model with the highest performance was trained with the MobileNetV2 network for the cystocolic scan point achieving 99.8% accuracy. Across all trained networks the diaphragmatic hepatorenal scan point had the best overall performance. However, GradCAM overlays showed that the models with highest accuracy, like MobileNetV2, were not always identifying relevant features. Conversely, the GradCAM heatmaps for ShrapML show general agreement with regions most indicative of fluid accumulation. Discussion Overall, the AI models developed can automate POCUS predictions in MWDs. Preliminarily, ShrapML had the strongest performance and prediction rate paired with accurately tracking fluid accumulation sites, making it the most suitable option for eventual real-time deployment with ultrasound systems. Further integration of this technology with imaging technologies will expand use of POCUS-based triage of MWDs.


Evaluation of Deep Learning Model Architectures for Point-of-Care Ultrasound Diagnostics

April 2024

·

12 Reads

·

3 Citations

Bioengineering

Point-of-care ultrasound imaging is a critical tool for patient triage during trauma for diagnosing injuries and prioritizing limited medical evacuation resources. Specifically, an eFAST exam evaluates if there are free fluids in the chest or abdomen but this is only possible if ultrasound scans can be accurately interpreted, a challenge in the pre-hospital setting. In this effort, we evaluated the use of artificial intelligent eFAST image interpretation models. Widely used deep learning model architectures were evaluated as well as Bayesian models optimized for six different diagnostic models: pneumothorax (i) B- or (ii) M-mode, hemothorax (iii) B- or (iv) M-mode, (v) pelvic or bladder abdominal hemorrhage and (vi) right upper quadrant abdominal hemorrhage. Models were trained using images captured in 27 swine. Using a leave-one-subject-out training approach, the MobileNetV2 and DarkNet53 models surpassed 85% accuracy for each M-mode scan site. The different B-mode models performed worse with accuracies between 68% and 74% except for the pelvic hemorrhage model, which only reached 62% accuracy for all model architectures. These results highlight which eFAST scan sites can be easily automated with image interpretation models, while other scan sites, such as the bladder hemorrhage model, will require more robust model development or data augmentation to improve performance. With these additional improvements, the skill threshold for ultrasound-based triage can be reduced, thus expanding its utility in the pre-hospital setting.


Design and testing of ultrasound probe adapters for a robotic imaging platform

March 2024

·

152 Reads

·

2 Citations

Medical imaging-based triage is a critical tool for emergency medicine in both civilian and military settings. Ultrasound imaging can be used to rapidly identify free fluid in abdominal and thoracic cavities which could necessitate immediate surgical intervention. However, proper ultrasound image capture requires a skilled ultrasonography technician who is likely unavailable at the point of injury where resources are limited. Instead, robotics and computer vision technology can simplify image acquisition. As a first step towards this larger goal, here, we focus on the development of prototypes for ultrasound probe securement using a robotics platform. The ability of four probe adapter technologies to precisely capture images at anatomical locations, repeatedly, and with different ultrasound transducer types were evaluated across more than five scoring criteria. Testing demonstrated two of the adapters outperformed the traditional robot gripper and manual image capture, with a compact, rotating design compatible with wireless imaging technology being most suitable for use at the point of injury. Next steps will integrate the robotic platform with computer vision and deep learning image interpretation models to automate image capture and diagnosis. This will lower the skill threshold needed for medical imaging-based triage, enabling this procedure to be available at or near the point of injury.


Summary of performance metrics for single-class shrapnel segmentation.
Summary of performance metrics by class for the segmentation models.
Using AI Segmentation Models to Improve Foreign Body Detection and Triage from Ultrasound Images

January 2024

·

100 Reads

·

1 Citation

Bioengineering

Medical imaging can be a critical tool for triaging casualties in trauma situations. In remote or military medicine scenarios, triage is essential for identifying how to use limited resources or prioritize evacuation for the most serious cases. Ultrasound imaging, while portable and often available near the point of injury, can only be used for triage if images are properly acquired, interpreted, and objectively triage scored. Here, we detail how AI segmentation models can be used for improving image interpretation and objective triage evaluation for a medical application focused on foreign bodies embedded in tissues at variable distances from critical neurovascular features. Ultrasound images previously collected in a tissue phantom with or without neurovascular features were labeled with ground truth masks. These image sets were used to train two different segmentation AI frameworks: YOLOv7 and U-Net segmentation models. Overall, both approaches were successful in identifying shrapnel in the image set, with U-Net outperforming YOLOv7 for single-class segmentation. Both segmentation models were also evaluated with a more complex image set containing shrapnel, artery, vein, and nerve features. YOLOv7 obtained higher precision scores across multiple classes whereas U-Net achieved higher recall scores. Using each AI model, a triage distance metric was adapted to measure the proximity of shrapnel to the nearest neurovascular feature, with U-Net more closely mirroring the triage distances measured from ground truth labels. Overall, the segmentation AI models were successful in detecting shrapnel in ultrasound images and could allow for improved injury triage in emergency medicine scenarios.


Toward Smart, Automated Junctional Tourniquets—AI Models to Interpret Vessel Occlusion at Physiological Pressure Points

January 2024

·

47 Reads

Bioengineering

Hemorrhage is the leading cause of preventable death in both civilian and military medicine. Junctional hemorrhages are especially difficult to manage since traditional tourniquet placement is often not possible. Ultrasound can be used to visualize and guide the caretaker to apply pressure at physiological pressure points to stop hemorrhage. However, this process is technically challenging, requiring the vessel to be properly positioned over rigid boney surfaces and applying sufficient pressure to maintain proper occlusion. As a first step toward automating this life-saving intervention, we demonstrate an artificial intelligence algorithm that classifies a vessel as patent or occluded, which can guide a user to apply the appropriate pressure required to stop flow. Neural network models were trained using images captured from a custom tissue-mimicking phantom and an ex vivo swine model of the inguinal region, as pressure was applied using an ultrasound probe with and without color Doppler overlays. Using these images, we developed an image classification algorithm suitable for the determination of patency or occlusion in an ultrasound image containing color Doppler overlay. Separate AI models for both test platforms were able to accurately detect occlusion status in test-image sets to more than 93% accuracy. In conclusion, this methodology can be utilized for guiding and monitoring proper vessel occlusion, which, when combined with automated actuation and other AI models, can allow for automated junctional tourniquet application.


Figure 2. Image processing and algorithm training pathway. M-mode images were obtained from the synthetic phantom, and coordinates of the time capture window were identified. A rolling window method was performed to isolate 100 individual panels from each capture window. The individual panels were used as the initial training set (No Data Aug box). A basic augmentation step was performed to the original dataset to create the second training set (+X-Y Flip, +Zoom Aug box). The third image preprocessing step used a porcine M-mode image to normalize the histogram for the phantom panels (+Histogram Normalization box). The final augmentation step included brightness and contrast augmentation to generate additional images (+Brightness, +Contrast Aug box).
Figure 3. B-mode image comparison with euthanized porcine subjects and synthetic phantom: (A) Baseline thoracic image acquired from porcine subject; (B) Baseline image acquired from rib and lung phantom apparatus; (C) Post-injury image acquired from same porcine subject; (D) Pneumothorax positive image acquired from synthetic phantom apparatus.
Training Ultrasound Image Classification Deep-Learning Algorithms for Pneumothorax Detection Using a Synthetic Tissue Phantom Apparatus

September 2022

·

95 Reads

·

17 Citations

Journal of Imaging

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.


Citations (12)


... In future applications, nanosensor arrays analysing the volatolome will be integrated into wearable self-powered healthcare devices for clinical decision of prehospitalization [316]. This will also promote the development of telemedicine using monitoring platforms enabling wireless communication [317,318]. ...

Reference:

Volatolomics for Anticipated Diagnosis of Cancers with Chemoresistive Vapour Sensors: A Review
Overview of Wearable Healthcare Devices for Clinical Decision Support in the Prehospital Setting

Sensors

... Our team has previously developed a dual-input fuzzy logic (DFL) controller and validated its ability to resuscitate to a target mean arterial pressure (MAP) in a physical testbed that mimicked different hemorrhage scenarios [20]. Additionally, an adaptive resuscitation controller (ARC) for fluid administration, also developed by our team, has been shown to successfully resuscitate and stabilize swine in a clinically relevant large animal study of hemorrhagic shock [21]. ...

In vivo evaluation of an adaptive resuscitation controller using whole blood and crystalloid infusates for hemorrhagic shock

... There are no studies quantifying and comparing the echogenicity of the different uterine intraluminal contents in domestic animals. Ultrasound parameters quantification has been used in machine learning applications to assess liver, kidneys and abdominal or thoracic free fluid in dogs [19][20][21]. ...

Deep learning models for interpretation of point of care ultrasound in military working dogs

... There are also studies that summarize the progress in ultrasound applications with AI, including utilizing convolutional neural networks for diagnostic applications and a robotic arm for assistance in casualty classification in pre-hospital settings [18]. We previously explored the use of deep learning AI through the exploration and evaluation of a wide range of trained binary classification diagnostic models to detect injury at eFAST scan sites in swine subjects [19]. Having diagnostic models to interpret medical images only addresses part of the challenge with performing eFAST exams. ...

Evaluation of Deep Learning Model Architectures for Point-of-Care Ultrasound Diagnostics

Bioengineering

... For this, AI and robotics can be applied to the eFAST exam, utilizing computer vision AI to guide a robotic platform to the relevant scan points of the eFAST exam. We have previously shown that a robotic imaging platform can traverse a wide range of eFAST scan points, and assessed different US probe holder designs for this application [20]. ...

Design and testing of ultrasound probe adapters for a robotic imaging platform

... Future work suggests expanding the dataset, including multiple labelers, and validating the models with in vivo medical images to enhance clinical abilities. In conclusion, both models are efficient and effective for segmenting foreign bodies in ultrasound images in ultrasound images [18]. ...

Using AI Segmentation Models to Improve Foreign Body Detection and Triage from Ultrasound Images

Bioengineering

... As such, AI/ML-based CDS tools have been widely developed for medical applications and have been thoroughly reviewed elsewhere [113][114][115][116]. Some relevant CDS applications for prehospital medicine include automated US-based detection of abdominal hemorrhage [117,118], pneumothorax [119,120], and COVID-19 [36,37]. By redesigning these imaging devices to be in a wearable format, data can be continuously collected "hands-free", opening the possibility to a wider range of CDS applications as the imaging WHD technology becomes more readily available. ...

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

Journal of Imaging

... The ARC is a novel controller that uses the MAP as the target variable; it was developed using the retrospective analysis of volume responsiveness in swine [28] and further tuned on a hardware-in-loop testbed [29]. For its operation, an initial bolus infusion of 100 mL was given during the first minute of resuscitation to obtain a baseline volume responsiveness measurement. ...

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

Bioengineering

... The US display provides track lines to delineate the expected insertion path. Our team previously developed laboratory test platforms [14] for vascular access device evaluation as well as an ex vivo swine model [15]. We expanded on the testing pipeline to evaluate the Vu-Path™ ACVAD in a swine model as well as conduct a human usability assessment. ...

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

Journal of Personalized Medicine

... Although certain initial conditions and target parameters may be input by the user, once the closed-loop controller is activated, all its outputs are calculated and executed by the controller itself until the system is stopped. There are also "hybrid" controllers that are unique, where they primarily function as closed-loop but make allowances for the user to manually override output values or change targets without having to stop and reset the controller (Avital et al., 2022). ...

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

Journal of Personalized Medicine