-
Automatic Cannulation of Femoral Vessels in a Porcine Shock Model
Authors:
Nico Zevallos,
Cecilia G. Morales,
Andrew Orekhov,
Tejas Rane,
Hernando Gomez,
Francis X. Guyette,
Michael R. Pinsky,
John Galeotti,
Artur Dubrawski,
Howie Choset
Abstract:
Rapid and reliable vascular access is critical in trauma and critical care. Central vascular catheterization enables high-volume resuscitation, hemodynamic monitoring, and advanced interventions like ECMO and REBOA. While peripheral access is common, central access is often necessary but requires specialized ultrasound-guided skills, posing challenges in prehospital settings. The complexity arises…
▽ More
Rapid and reliable vascular access is critical in trauma and critical care. Central vascular catheterization enables high-volume resuscitation, hemodynamic monitoring, and advanced interventions like ECMO and REBOA. While peripheral access is common, central access is often necessary but requires specialized ultrasound-guided skills, posing challenges in prehospital settings. The complexity arises from deep target vessels and the precision needed for needle placement. Traditional techniques, like the Seldinger method, demand expertise to avoid complications. Despite its importance, ultrasound-guided central access is underutilized due to limited field expertise. While autonomous needle insertion has been explored for peripheral vessels, only semi-autonomous methods exist for femoral access. This work advances toward full automation, integrating robotic ultrasound for minimally invasive emergency procedures. Our key contribution is the successful femoral vein and artery cannulation in a porcine hemorrhagic shock model.
△ Less
Submitted 17 June, 2025;
originally announced June 2025.
-
Natural Language Programming in Medicine: Administering Evidence Based Clinical Workflows with Autonomous Agents Powered by Generative Large Language Models
Authors:
Akhil Vaid,
Joshua Lampert,
Juhee Lee,
Ashwin Sawant,
Donald Apakama,
Ankit Sakhuja,
Ali Soroush,
Sarah Bick,
Ethan Abbott,
Hernando Gomez,
Michael Hadley,
Denise Lee,
Isotta Landi,
Son Q Duong,
Nicole Bussola,
Ismail Nabeel,
Silke Muehlstedt,
Silke Muehlstedt,
Robert Freeman,
Patricia Kovatch,
Brendan Carr,
Fei Wang,
Benjamin Glicksberg,
Edgar Argulian,
Stamatios Lerakis
, et al. (5 additional authors not shown)
Abstract:
Generative Large Language Models (LLMs) hold significant promise in healthcare, demonstrating capabilities such as passing medical licensing exams and providing clinical knowledge. However, their current use as information retrieval tools is limited by challenges like data staleness, resource demands, and occasional generation of incorrect information. This study assessed the potential of LLMs to…
▽ More
Generative Large Language Models (LLMs) hold significant promise in healthcare, demonstrating capabilities such as passing medical licensing exams and providing clinical knowledge. However, their current use as information retrieval tools is limited by challenges like data staleness, resource demands, and occasional generation of incorrect information. This study assessed the potential of LLMs to function as autonomous agents in a simulated tertiary care medical center, using real-world clinical cases across multiple specialties. Both proprietary and open-source LLMs were evaluated, with Retrieval Augmented Generation (RAG) enhancing contextual relevance. Proprietary models, particularly GPT-4, generally outperformed open-source models, showing improved guideline adherence and more accurate responses with RAG. The manual evaluation by expert clinicians was crucial in validating models' outputs, underscoring the importance of human oversight in LLM operation. Further, the study emphasizes Natural Language Programming (NLP) as the appropriate paradigm for modifying model behavior, allowing for precise adjustments through tailored prompts and real-world interactions. This approach highlights the potential of LLMs to significantly enhance and supplement clinical decision-making, while also emphasizing the value of continuous expert involvement and the flexibility of NLP to ensure their reliability and effectiveness in healthcare settings.
△ Less
Submitted 22 August, 2024; v1 submitted 5 January, 2024;
originally announced January 2024.
-
Patient-specific computational forecasting of prostate cancer growth during active surveillance using an imaging-informed biomechanistic model
Authors:
Guillermo Lorenzo,
Jon S. Heiselman,
Michael A. Liss,
Michael I. Miga,
Hector Gomez,
Thomas E. Yankeelov,
Alessandro Reali,
Thomas J. R. Hughes
Abstract:
Active surveillance (AS) is a suitable management option for newly-diagnosed prostate cancer (PCa), which usually presents low to intermediate clinical risk. Patients enrolled in AS have their tumor closely monitored via longitudinal multiparametric magnetic resonance imaging (mpMRI), serum prostate-specific antigen tests, and biopsies. Hence, the patient is prescribed treatment when these tests i…
▽ More
Active surveillance (AS) is a suitable management option for newly-diagnosed prostate cancer (PCa), which usually presents low to intermediate clinical risk. Patients enrolled in AS have their tumor closely monitored via longitudinal multiparametric magnetic resonance imaging (mpMRI), serum prostate-specific antigen tests, and biopsies. Hence, the patient is prescribed treatment when these tests identify progression to higher-risk PCa. However, current AS protocols rely on detecting tumor progression through direct observation according to standardized monitoring strategies. This approach limits the design of patient-specific AS plans and may lead to the late detection and treatment of tumor progression. Here, we propose to address these issues by leveraging personalized computational predictions of PCa growth. Our forecasts are obtained with a spatiotemporal biomechanistic model informed by patient-specific longitudinal mpMRI data. Our results show that our predictive technology can represent and forecast the global tumor burden for individual patients, achieving concordance correlation coefficients ranging from 0.93 to 0.99 across our cohort (n=7). Additionally, we identify a model-based biomarker of higher-risk PCa: the mean proliferation activity of the tumor (p=0.041). Using logistic regression, we construct a PCa risk classifier based on this biomarker that achieves an area under the receiver operating characteristic curve of 0.83. We further show that coupling our tumor forecasts with this PCa risk classifier enables the early identification of PCa progression to higher-risk disease by more than one year. Thus, we posit that our predictive technology constitutes a promising clinical decision-making tool to design personalized AS plans for PCa patients.
△ Less
Submitted 29 September, 2023;
originally announced October 2023.
-
Patient-specific, mechanistic models of tumor growth incorporating artificial intelligence and big data
Authors:
Guillermo Lorenzo,
Syed Rakin Ahmed,
David A. Hormuth II,
Brenna Vaughn,
Jayashree Kalpathy-Cramer,
Luis Solorio,
Thomas E. Yankeelov,
Hector Gomez
Abstract:
Despite the remarkable advances in cancer diagnosis, treatment, and management that have occurred over the past decade, malignant tumors remain a major public health problem. Further progress in combating cancer may be enabled by personalizing the delivery of therapies according to the predicted response for each individual patient. The design of personalized therapies requires patient-specific in…
▽ More
Despite the remarkable advances in cancer diagnosis, treatment, and management that have occurred over the past decade, malignant tumors remain a major public health problem. Further progress in combating cancer may be enabled by personalizing the delivery of therapies according to the predicted response for each individual patient. The design of personalized therapies requires patient-specific information integrated into an appropriate mathematical model of tumor response. A fundamental barrier to realizing this paradigm is the current lack of a rigorous, yet practical, mathematical theory of tumor initiation, development, invasion, and response to therapy. In this review, we begin by providing an overview of different approaches to modeling tumor growth and treatment, including mechanistic as well as data-driven models based on ``big data" and artificial intelligence. Next, we present illustrative examples of mathematical models manifesting their utility and discussing the limitations of stand-alone mechanistic and data-driven models. We further discuss the potential of mechanistic models for not only predicting, but also optimizing response to therapy on a patient-specific basis. We then discuss current efforts and future possibilities to integrate mechanistic and data-driven models. We conclude by proposing five fundamental challenges that must be addressed to fully realize personalized care for cancer patients driven by computational models.
△ Less
Submitted 28 August, 2023;
originally announced August 2023.
-
Toward Robotically Automated Femoral Vascular Access
Authors:
Nico Zevallos,
Evan Harber,
Abhimanyu,
Kirtan Patel,
Yizhu Gu,
Kenny Sladick,
Francis Guyette,
Leonard Weiss,
Michael R. Pinsky,
Hernando Gomez,
John Galeotti,
Howie Choset
Abstract:
Advanced resuscitative technologies, such as Extra Corporeal Membrane Oxygenation (ECMO) cannulation or Resuscitative Endovascular Balloon Occlusion of the Aorta (REBOA), are technically difficult even for skilled medical personnel. This paper describes the core technologies that comprise a teleoperated system capable of granting femoral vascular access, which is an important step in both of these…
▽ More
Advanced resuscitative technologies, such as Extra Corporeal Membrane Oxygenation (ECMO) cannulation or Resuscitative Endovascular Balloon Occlusion of the Aorta (REBOA), are technically difficult even for skilled medical personnel. This paper describes the core technologies that comprise a teleoperated system capable of granting femoral vascular access, which is an important step in both of these procedures and a major roadblock in their wider use in the field. These technologies include a kinematic manipulator, various sensing modalities, and a user interface. In addition, we evaluate our system on a surgical phantom as well as in-vivo porcine experiments. These resulted in, to the best of our knowledge, the first robot-assisted arterial catheterizations; a major step towards our eventual goal of automatic catheter insertion through the Seldinger technique.
△ Less
Submitted 6 July, 2021;
originally announced July 2021.
-
Optimal control of cytotoxic and antiangiogenic therapies on prostate cancer growth
Authors:
Pierluigi Colli,
Hector Gomez,
Guillermo Lorenzo,
Gabriela Marinoschi,
Alessandro Reali,
Elisabetta Rocca
Abstract:
Prostate cancer can be lethal in advanced stages, for which chemotherapy may become the only viable therapeutic option. While there is no clear clinical management strategy fitting all patients, cytotoxic chemotherapy with docetaxel is currently regarded as the gold standard. However, tumors may regain activity after treatment conclusion and become resistant to docetaxel. This situation calls for…
▽ More
Prostate cancer can be lethal in advanced stages, for which chemotherapy may become the only viable therapeutic option. While there is no clear clinical management strategy fitting all patients, cytotoxic chemotherapy with docetaxel is currently regarded as the gold standard. However, tumors may regain activity after treatment conclusion and become resistant to docetaxel. This situation calls for new delivery strategies and drug compounds enabling an improved therapeutic outcome. Combination of docetaxel with antiangiogenic therapy has been considered a promising strategy. Bevacizumab is the most common antiangiogenic drug, but clinical studies have not revealed a clear benefit from its combination with docetaxel. Here, we capitalize on our prior work on mathematical modeling of prostate cancer growth subjected to combined cytotoxic and antiangiogenic therapies, and propose an optimal control framework to robustly compute the drug-independent cytotoxic and antiangiogenic effects enabling an optimal therapeutic control of tumor dynamics. We describe the formulation of the optimal control problem, for which we prove the existence of at least a solution and determine the necessary first order optimality conditions. We then present numerical algorithms based on isogeometric analysis to run a preliminary simulation study over a single cycle of combined therapy. Our results suggest that only cytotoxic chemotherapy is required to optimize therapeutic performance and we show that our framework can produce superior solutions to combined therapy with docetaxel and bevacizumab. We also illustrate how the optimal drug-naïve cytotoxic effects computed in these simulations may be successfully leveraged to guide drug production and delivery strategies by running a nonlinear least-square fit of protocols involving docetaxel and a new design drug.
△ Less
Submitted 9 July, 2020;
originally announced July 2020.
-
A deep learning framework for solution and discovery in solid mechanics
Authors:
Ehsan Haghighat,
Maziar Raissi,
Adrian Moure,
Hector Gomez,
Ruben Juanes
Abstract:
We present the application of a class of deep learning, known as Physics Informed Neural Networks (PINN), to learning and discovery in solid mechanics. We explain how to incorporate the momentum balance and constitutive relations into PINN, and explore in detail the application to linear elasticity, and illustrate its extension to nonlinear problems through an example that showcases von~Mises elas…
▽ More
We present the application of a class of deep learning, known as Physics Informed Neural Networks (PINN), to learning and discovery in solid mechanics. We explain how to incorporate the momentum balance and constitutive relations into PINN, and explore in detail the application to linear elasticity, and illustrate its extension to nonlinear problems through an example that showcases von~Mises elastoplasticity. While common PINN algorithms are based on training one deep neural network (DNN), we propose a multi-network model that results in more accurate representation of the field variables. To validate the model, we test the framework on synthetic data generated from analytical and numerical reference solutions. We study convergence of the PINN model, and show that Isogeometric Analysis (IGA) results in superior accuracy and convergence characteristics compared with classic low-order Finite Element Method (FEM). We also show the applicability of the framework for transfer learning, and find vastly accelerated convergence during network re-training. Finally, we find that honoring the physics leads to improved robustness: when trained only on a few parameters, we find that the PINN model can accurately predict the solution for a wide range of parameters new to the network---thus pointing to an important application of this framework to sensitivity analysis and surrogate modeling.
△ Less
Submitted 6 May, 2020; v1 submitted 14 February, 2020;
originally announced March 2020.
-
The divergence-conforming immersed boundary method: Application to vesicle and capsule dynamics
Authors:
Hugo Casquero,
Carles Bona-Casas,
Deepesh Toshniwal,
Thomas J. R. Hughes,
Hector Gomez,
Yongjie Jessica Zhang
Abstract:
We extend the recently introduced divergence-conforming immersed boundary (DCIB) method [1] to fluid-structure interaction (FSI) problems involving closed co-dimension one solids. We focus on capsules and vesicles, whose discretization is particularly challenging due to the higher-order derivatives that appear in their formulations. In two-dimensional settings, we employ cubic B-splines with perio…
▽ More
We extend the recently introduced divergence-conforming immersed boundary (DCIB) method [1] to fluid-structure interaction (FSI) problems involving closed co-dimension one solids. We focus on capsules and vesicles, whose discretization is particularly challenging due to the higher-order derivatives that appear in their formulations. In two-dimensional settings, we employ cubic B-splines with periodic knot vectors to obtain discretizations of closed curves with C^2 inter-element continuity. In three-dimensional settings, we use analysis-suitable bi-cubic T-splines to obtain discretizations of closed surfaces with at least C^1 inter-element continuity. Large spurious changes of the fluid volume inside closed co-dimension one solids is a well-known issue for IB methods. The DCIB method results in volume changes orders of magnitude lower than conventional IB methods. This is a byproduct of discretizing the velocity-pressure pair with divergence-conforming B-splines, which lead to negligible incompressibility errors at the Eulerian level. The higher inter-element continuity of divergence-conforming B-splines is also crucial to avoid the quadrature/interpolation errors of IB methods becoming the dominant discretization error. Benchmark and application problems of vesicle and capsule dynamics are solved, including mesh-independence studies and comparisons with other numerical methods.
△ Less
Submitted 22 January, 2020;
originally announced January 2020.
-
Mathematical analysis and simulation study of a phase-field model of prostate cancer growth with chemotherapy and antiangiogenic therapy effects
Authors:
Pierluigi Colli,
Hector Gomez,
Guillermo Lorenzo,
Gabriela Marinoschi,
Alessandro Reali,
Elisabetta Rocca
Abstract:
Cytotoxic chemotherapy is a common treatment for advanced prostate cancer. These tumors are also known to rely on angiogenesis, i.e., the growth of local microvasculature via chemical signaling produced by the tumor. Thus, several clinical studies have been investigating antiangiogenic therapy for advanced prostate cancer, either as monotherapy or combined with standard cytotoxic protocols. Howeve…
▽ More
Cytotoxic chemotherapy is a common treatment for advanced prostate cancer. These tumors are also known to rely on angiogenesis, i.e., the growth of local microvasculature via chemical signaling produced by the tumor. Thus, several clinical studies have been investigating antiangiogenic therapy for advanced prostate cancer, either as monotherapy or combined with standard cytotoxic protocols. However, the complex genetic alterations promoting prostate cancer growth complicate the selection of the best chemotherapeutic approach for each patient's tumor. Here, we present a mathematical model of prostate cancer growth and chemotherapy that may enable physicians to test and design personalized chemotherapeutic protocols in silico. We use the phase-field method to describe tumor growth, which we assume to be driven by a generic nutrient following reaction-diffusion dynamics. Tumor proliferation and apoptosis (i.e., programmed cell death) can be parameterized with experimentally-determined values. Cytotoxic chemotherapy is included as a term downregulating tumor net proliferation, while antiangiogenic therapy is modeled as a reduction in intratumoral nutrient supply. Another equation couples the tumor phase field with the production of prostate-specific antigen, which is an extensively used prostate cancer biomarker. We prove the well-posedness of our model and we run a series of representative simulations using an isogeometric method to explore untreated tumor growth as well as the effects of cytotoxic chemotherapy and antiangiogenic therapy, both alone and combined. Our simulations show that our model captures the growth morphologies of prostate cancer as well as common outcomes of cytotoxic and antiangiogenic mono and combined therapy. Our model also reproduces the usual temporal trends in tumor volume and prostate-specific antigen evolution observed in previous studies.
△ Less
Submitted 26 July, 2019;
originally announced July 2019.
