Search NASA⌕ Search

SEARCH · Search NASA

Results for “biomedical imaging”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 37 records · Page 2
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

Why is the winner the best?

International benchmarking competitions have become fundamental for the comparative performance assessment of image analysis methods. However, little attention has been given to investigating what can be learnt from these competitions. Do they really generate scientific progress? What are common and successful participation strategies? What makes a solution superior to a competing method? To address this gap in the literature, we performed a multi- center study with all 80 competitions that were conducted in the scope of IEEE ISBI 2021 and MICCAI 2021. Statistical analyses performed based on comprehensive descriptions of the submitted algorithms linked to their rank as well as the underlying participation strategies revealed common characteristics of winning solutions. These typically include the use of multi-task learning (63%) and/or multi-stage pipelines (61%), and a focus on augmentation (100%), im- age preprocessing (97%), data curation (79%), and post- processing (66%). The “typical” lead of a winning team is a computer scientist with a doctoral degree, five years of experience in biomedical image analysis, and four years of experience in deep learning. Two core general development strategies stood out for highly-ranked teams: the reflection of the metrics in the method design and the focus on analyzing and handling failure cases. According to the organizers, 43% of the winning algorithms exceeded the state of the art but only 11% completely solved the respective domain problem. The insights of our study could help researchers (1) improve algorithm development strategies when approaching new problems, and (2) focus on open research questions revealed by this work.

Eisenmann, Matthias↗
DOE OSTI2021

Inorganic Halide Perovskitoid TlPbI 3 for Ionizing Radiation Detection

Room temperature semiconductor detector (RTSD) materials for γ-ray and X-ray radiation are in great demand for the nonproliferation of nuclear materials as well as for biomedical imaging applications. Halide perovskites have attracted great attention as emerging and promising RTSD materials. In this contribution, the material synthesis, purification, crystal growth, crystal structure, photoluminescence properties, ionizing radiation detection performance, and electronic structure of the inorganic halide perovskitoid compound TlPbI 3 are reported on. This compound crystallizes in the ABX 3 non-perovskite crystal structure with a high density of d = 6.488 g·cm –3 , has a wide bandgap of 2.25 eV, and melts congruently at a low temperature of 360 °C without phase transitions, which allows for facile growth of high quality crystals with few thermally-activated defects. High-quality TlPbI 3 single crystals of centimeter-size are grown using the vertical Bridgman method using purified raw materials. A high electrical resistivity of ≈10 12 Ω·cm is readily obtainable, and detectors made of TlPbI 3 single crystals are highly photoresponsive to Ag K α X-rays (22.4 keV), and detects 122 keV γ-rays from 57 Co radiation source. The electron mobility-lifetime product µ e τ e was estimated at 1.8 × 10 –5 cm 2 ·V –1 . A high relative static dielectric constant of 35.0 indicates strong capability in screening carrier scattering and charged defects in TlPbI 3 .

36 MATERIALS SCIENCE↗
DOE OSTI2025

Colossal Terahertz Emission with Ultrafast Tunability Based on Van der Waals Ferroelectric NbOI 2

Abstract Terahertz (THz) technology is critical for quantum material physics, biomedical imaging, ultrafast electronics, and next‐generation wireless communications. However, standing in the way of widespread applications is the scarcity of efficient ultrafast THz sources with on‐demand fast modulation and easy on‐chip integration capability. Here the discovery of colossal THz emission is reported from a van der Waals (vdW) ferroelectric semiconductor NbOI 2 . Using THz emission spectroscopy, a THz generation efficiency an order of magnitude higher than that of ZnTe, a standard nonlinear crystal for ultrafast THz generation is observed. The underlying generation mechanisms associated are further uncovered with its large ferroelectric polarization by studying the THz emission dependence on excitation wavelength, incident polarization, and fluence. Moreover, the ultrafast coherent amplification and annihilation of the THz emission and associated coherent phonon oscillations by employing a double‐pump scheme are demonstrated. These findings combined with first‐principles calculations, inform a new understanding of the THz light–matter interaction in emergent vdW ferroelectrics and pave the way to develop high‐performance THz devices on them for quantum materials sensing and ultrafast electronics.

Subedi, Sujan [Department of Materials Science and↗
DOE OSTI2022

Widespread opportunities for materials engineering of nanocrystals: Synthetically tailorable effects and methodologies

Since their initial development, semiconductor nanocrystals (NCs) and metal nanoparticles (NPs) have become widely utilized in a variety of applications, including solar energy harvesting, lighting, catalysis, biomedical imaging, and targeted cancer treatments. Because of their reduced dimensionality, NCs and NPs have vastly different optical and electronic properties than their bulk counterparts, and the properties often depend heavily on the NC or NP surface. Surface defects and ligands alter the optoelectronic properties of NCs and NPs and affect their interaction with the environment, allowing tailoring of their properties without changing the overall chemical composition. In this review, we discuss research progress focusing on optoelectronic properties of NPs and NCs and their respective applications. We highlight how different surface treatments and ligand functionalization of nanomaterials can influence these applications.

36 MATERIALS SCIENCE↗
DOE OSTI2023

Chiral Spectroscopy of Nanostructures

Chirality is ubiquitous in the universe and in living creatures over detectable length scales from the subatomic to the galactic, as exemplified in the two extremes by subatomic particles (neutrinos) and spiral galaxies. Between them are living creatures that display multiple levels of chirality emerging from hierarchically assembled asymmetric building blocks. Not too far from the bottom of this pyramid are the foundational building blocks with chiral atomic centers on sp 3 carbon atoms exemplified by L-amino acids and D-sugars that are self-assembled into higher-order structures with increasing dimensions forming highly complex, amazingly functional, and energy-efficient living systems. The organization and materials employed in their construction inspired scientists to replicate complex living systems via the self-assembly of chiral components. Multiple studies pointed to unexpected and unique electromagnetic properties of chiral structures with nanoscale and microscale dimensions, including giant circular dichroism and collective circularly polarized scattering that their constituent units did not possess. To address the wide variety of chiral geometries observed in continuous materials, singular particles, and their complex systems, multiple analytic techniques are needed. Simultaneously, their spectroscopic properties create a pathway to multiple applications. For example, mirror-asymmetric vibrations at chiral centers formed by sp 3 carbon atoms lead to optical activity for the infrared (IR) wavelength regions. At the same time, understanding the optical activity in, for example, the IR region enables biomedical applications because multiple modalities of biomedical imaging and vibrational optical activity (VOA) of biomolecules are known for IR range. In turn, VOA can be realized in both absorption and emission modalities due to large magnetic transition moments, as vibrational circular dichroism (VCD) or Raman optical activity (ROA) spectroscopy. In addition to the VOA, in the range of longer wavelengths, lattice vibrational mode or phononic behavior occurs in chiral crystals and nanoassemblies, which can be readily detected by terahertz circular dichroism (TCD) spectroscopy. Meanwhile, chiral self-assembly can induce circularly polarized light emission (CPLE) regardless of the existence of chirality in coassembled fluorophores. The CPLE from self-assembled chiral materials is particularly interesting because the CPLE can originate from both circularly polarized luminescence and circularly polarized scattering (CPS). Furthermore, because self-assembled nanostructures often exhibit stronger optical activity than their building blocks owing to dimension and resonance effects, the optical activity of single assembled nanostructures can be investigated by using microscopic technology combined with chiral optics. Here, we describe the state of the art for spectroscopic methods for the comprehensive analysis of chiral nanomaterials at various photon wavelengths, addressed with special attention given to new tools emerging both for materials with self-organized hierarchical chirality and single-particle spectroscopy.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2020

Predicting Dexter Energy Transfer Interactions from Molecular Orbital Overlaps

Dexter energy transfer (DET), also known as spin forbidden energy transfer (e.g., 3 D 1 A → 1 D 3 A), is undergoing renewed interest because of its utility in energy science, photocatalysis, and biomedical imaging. The DET reactions enable exciton fission and fusion and have been recently used to facilitate nanoscale imaging of protein assemblies on cell surfaces. While DET and Förster energy transfer rates are often well described using Fermi’s golden rule, the donor–acceptor couplings for DET may be bridge-state-mediated and depend on the electronic structure of and interactions among the donor, bridge, and acceptor. We provide a simple approach to approximate DET couplings using a product of molecular orbital overlap factors. This approach may be used to guide the design of DET systems with targeted energy-transfer kinetics.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2021

Magnetic and dielectric property control in the multivalent nanoscale perovskite Eu 0.5 Ba 0.5 TiO 3

We report nanoscale Eu 0.5 Ba 0.5 TiO 3 , a multiferroic in the bulk and candidate in the search to quantify the electric dipole moment of the electron. Eu 0.5 Ba 0.5 TiO 3 , in the form of nanoparticles and other nanostructures is interesting for nanocomposite integration, biomedical imaging and fundamental research, based upon the prospect of polarizability, f-orbital magnetism and tunable optical/radio luminescence. We developed a [non-hydrolytic]sol–[H 2 O-activated]gel route, derived from in-house metallic Ba (s) /Eu (s) alkoxide precursors and Ti{(OCH(CH 3 ) 2 } 4 . Two distinct nanoscale compounds of Ba:Ti:Eu with the parent perovskite crystal structure were produced, with variable dielectric, magnetic and optical properties, based on altering the oxidizing/reducing conditions. Eu 0.5 Ba 0.5 TiO 3 prepared under air/O 2 atmospheres produced a spherical core–shell nanostructure (30–35 nm), with perovskite Eu 0.5 Ba 0.5 TiO 3 nanocrystal core-insulating oxide shell layer (~3 nm), presumed a pre-pyrochlore layer abundant with Eu 3+ . Fluorescence spectroscopy shows a high intensity 5 D 0 → 7 F 2 transition at 622 nm and strong red fluorescence. The core/shell structure demonstrated excellent capacitive properties: assembly into dielectric thin films gave low conductivity (2133 GΩ mm -1 ) and an extremely stable, low loss permittivity of ε eff ~25 over a wide frequency range (tan δ < 0.01, 100 kHz–2 MHz). Eu 0.5 Ba 0.5 TiO 3 prepared under H 2 /argon produced more irregular shaped nanocrystals (20–25) nm, with a thin film permittivity around 4 times greater ( ε eff 101, tan δ < 0.05, 10 kHz–2 MHz, σ ~59.54 kΩ mm -1 ). Field-cooled magnetization values of 0.025 emu g -1 for EBTO-Air and 0.84 emu g -1 for EBTO-Argon were observed. X-ray photoelectron spectroscopy analysis reveals a complex interplay of Eu II/III /Ti III/IV configurations which contribute to the observed ferroic and fluorescence behavior.

77 NANOSCIENCE AND NANOTECHNOLOGY↗
DOE OSTI2025

Femtosecond wavelength-tunable laser system using gain managed nonlinear amplifier

Optical parametric amplifiers require a seed energy at the spectral range of interest for amplification. In the case of pulsed laser systems, seed pulse characteristics such as spectral energy density and spectral uniformity influence laser system design and performance. In this letter, we present the design and modeling of a few micro-joules, wavelength-tunable, femtosecond parametric amplifier system. We employ a gain-managed nonlinear fiber amplifier output as the seed pulse. This seed pulse is relatively uniform across its bandwidth from 1010 to 1180 nm. Its spectral energy density averages 600 pJ/nm, 30 times higher than conventional counterparts. The overall amplification gain is 175, and the conversion efficiency is 17.5%. The inherent quasi-linear chirp of the seed pulse is utilized to alter the output wavelength. Our tunable femtosecond laser is an enabling technology for high-demand applications such as time-resolved spectroscopy, nonlinear and advanced material research, and biomedical imaging and microscopy modalities.

47 OTHER INSTRUMENTATION↗
DOE OSTI2022

Freestanding high-aspect-ratio gold masks for low-energy, phase-based x-ray microscopy

High-resolution, x-ray phase contrast microscopy, a key technique with promising potential in biomedical imaging and diagnostics, is based on narrow-slit high-aspect-ratio gold gratings. Here we present the development, fabrication details, and experimental testing of the freestanding 10 μm thick gold membrane masks with an array of 0.9–1.5 μm void slit apertures for a novel low-energy x-ray microscope. The overall mask size is 4 mm × 4 mm, with a grating pitch of 7.5 μm, 6.0–6.6 μm wide gold bars are supported by 3 μm wide crosslinks at 400 μm intervals. The fabrication process is based on gold electroplating into a silicon mold coated with various thin films to form a voltage barrier, plating base, and sacrificial layer, followed by the mold removal to obtain the freestanding gold membrane with void slit apertures. We discuss key aspects for the materials and processes, including gold structures homogeneity, residual stresses, and prevention of collapsing of the grid elements. We further demonstrate the possibility to obtain high-resolution, high contrast 2D images of biological samples using an incoherent, rotating anode x-ray tube.

77 NANOSCIENCE AND NANOTECHNOLOGY↗
DOE OSTI2025

Physics-guided dual implicit neural representations for source separation

Significant challenges exist in efficient data analysis of most advanced experimental and observational techniques because the collected signals often include unwanted contributions, such as background and signal distortions, that can obscure the physically relevant information of interest. To address this, we have developed a self-supervised machine-learning approach for source separation using a dual implicit neural representation framework that jointly trains two neural networks: one for approximating distortions of the physical signal of interest and the other for learning the effective background contribution. Our method learns directly from the raw data by minimizing a reconstruction-based loss function without requiring labeled data or pre-defined dictionaries. We demonstrate the effectiveness of our framework by considering a challenging case study involving large-scale simulated, as well as experimental, momentum-energy-dependent inelastic neutron scattering data in a four-dimensional parameter space, characterized by heterogeneous background contributions and unknown distortions to the target signal. The method is found to successfully separate physically meaningful signals from a complex or structured background even when the signal characteristics vary across all four dimensions of the parameter space. An analytical approach that informs the choice of the regularization parameter is presented. Our method offers a versatile framework for addressing source separation problems across diverse domains, ranging from superimposed signals in astronomical measurements to structural features in biomedical image reconstructions.

47 OTHER INSTRUMENTATION↗
DOE OSTI2023

Material Identification From Radiographs Without Energy Resolution

We propose a method for performing material identification from radiographs without energy-resolved measurements. Material identification has a wide variety of applications, including in biomedical imaging, nondestructive testing, and security. While existing techniques for radiographic material identification make use of dual energy sources, energy-resolving detectors, or additional (e.g., neutron) measurements, such setups are not always practical— requiring additional hardware and complicating imaging. We tackle material identification without energy resolution, allowing standard X-ray systems to provide material identification information without requiring additional hardware. Assuming a setting where the geometry of each object in the scene is known and the materials come from a known set of possible materials, we pose the problem as a combinatorial optimization with a loss function that accounts for the presence of scatter and an unknown gain and propose a branch and bound algorithm to efficiently solve it. We present experiments on both synthetic data and real, experimental data with relevance to security applications— thick, dense objects imaged with MeV X-rays. We show that material identification can be efficient and accurate, for example, in a scene with three shells (two copper, one aluminum), our algorithm ran in six minutes on a consumer-level laptop and identified the correct materials as being among the top 10 best matches out of 8,000 possibilities.

36 MATERIALS SCIENCE↗
DOE OSTI2025

Quantum Transfer Learning to Boost Dementia Detection

Dementia is a devastating condition with profound implications for individuals, families, and healthcare systems. Early and accurate detection of dementia is critical for timely intervention and improved patient outcomes. While classical machine learning and deep learning approaches have been explored extensively for dementia prediction, these solutions often struggle with high-dimensional biomedical data and large-scale datasets, quickly reaching computational and performance limitations. To address this challenge, quantum machine learning (QML) has emerged as a promising paradigm, offering faster training and advanced pattern recognition capabilities. This work aims to demonstrate the potential of quantum transfer learning (QTL) to enhance the performance of a weak classical deep learning model applied to a binary classification task for dementia detection. Besides, we show the effect of noise on the QTL-based approach, investigating the reliability and robustness of this method. Using the OASIS 2 dataset, we show how quantum techniques can transform a suboptimal classical model into a more effective solution for biomedical image classification, highlighting their potential impact on advancing healthcare technology.

Bhowmik, Sounak [University of Tennessee, Knoxvill↗
DOE OSTI2020

Development of High-QE GaN Photodectors for Liquid Noble Particle Detectors

Detection of the handful of photons generated when a dark matter particle interacts in a liquid Xenon detector requires a new type of photodetector with very high detection efficiency at ultraviolet wavelengths, very low radioactive backgrounds, and essentially no measureable readout noise (achieved with devices with high intrisic gain). Such devices would find wide applicability to not only high energy physics experiments, but to a number of commercial applications including, e.g., biomedical imaging of fluorescent tags. We developed photodetectors based on amorphous GaN thin film structures (photocathodes) grown by molecular beam epitaxy on quartz windows. We also developed approaches to integrate these photocathode device structures with vacuum electron multipliers. All materials and fabrication techniques were developed to minimize radioactive backgrounds in materials, so these detectors could be used for very low count rate experiments, such as the future liquid-Xenon dark matter detectors, and future liquid-Argon neutrino detectors. We demonstrated growth of very high quantum efficiency amorphous GaN cathodes grown on a number of different substrates (Stainless steel, p-Silicon) reaching detection efficiencies of 80% at 220nm. The key to this growth was to use nitrogen plasma-assistend MBEs and very low growth temperatures. We also discovered that thinner structures (e.g.,40nm a-GaN thin films) had the best performance. We successfully demonstrated techniques for the vacuum transfer of these photocathode structures to a novel chamber that used hydraulics and bellow to form a cold indium bond needed to integrate these cathodes into sealed-tube devices. A complete, stand-along, device could not be completed during this proposal period due to difficulties in growing structures on the thick quartz window. Subsequent work is aimed at addressing this difficulty, by growing a buffer layer on the quartz substrate.

07 ISOTOPE AND RADIATION SOURCES↗
DOE OSTI2022

Data-Driven and Multiscale Modeling of DNA-Templated Dye Aggregates

Dye aggregates are of interest for excitonic applications, including biomedical imaging, organic photovoltaics, and quantum information systems. Dyes with large transition dipole moments (μ) are necessary to optimize coupling within dye aggregates. Extinction coefficients (ε) can be used to determine the μ of dyes, and so dyes with a large ε (>150,000 M –1 cm –1 ) should be engineered or identified. However, dye properties leading to a large ε are not fully understood, and low-throughput methods of dye screening, such as experimental measurements or density functional theory (DFT) calculations, can be time-consuming. In order to screen large datasets of molecules for desirable properties (i.e., large ε and μ), a computational workflow was established using machine learning (ML), DFT, time-dependent (TD-) DFT, and molecular dynamics (MD). ML models were developed through training and validation on a dataset of 8802 dyes using structural features. A Classifier was developed with an accuracy of 97% and a Regressor was constructed with an R2 of above 0.9, comparing between experiment and ML prediction. Using the Regressor, the ε values of over 18,000 dyes were predicted. The top 100 dyes were further screened using DFT and TD-DFT to identify 15 dyes with a μ relative to a reference dye, pentamethine indocyanine dye Cy5. Two benchmark MD simulations were performed on Cy5 and Cy5.5 dimers, and it was found that MD could accurately capture experimental results. The results of this study exhibit that our computational workflow for identifying dyes with a large μ for excitonic applications is effective and can be used as a tool to develop new dyes for excitonic applications.

77 NANOSCIENCE AND NANOTECHNOLOGY↗
DOE OSTI2025

Modeling inter‐reader variability in clinical target volume delineation for soft tissue sarcomas using diffusion model

Abstract Background Accurate delineation of the clinical target volume (CTV) is essential in the radiotherapy treatment of soft tissue sarcomas. However, this process is subject to inter‐reader variability due to the need for clinical assessment of risk and extent of potential microscopic spread. This can lead to inconsistencies in treatment planning, potentially impacting treatment outcomes. Most existing automatic CTV delineation methods do not account for this variability and can only generate a single CTV for each case. Purpose This study aims to develop a deep learning‐based technique to generate multiple CTV contours for each case, simulating the inter‐reader variability in the clinical practice. Methods We employed a publicly available dataset consisting of fluorodeoxyglucose positron emission tomography (FDG‐PET), x‐ray computed tomography (CT), and pre‐contrast T1‐weighted magnetic resonance imaging (MRI) scans from 51 patients with soft tissue sarcoma, along with an independent validation set containing five additional patients. An experienced reader drew a contour of the gross tumor volume (GTV) for each patient based on multi‐modality images. Subsequently, two additional readers, together with the first one, were responsible for contouring three CTVs in total based on the GTV. We developed a diffusion model‐based deep learning method that is capable of generating arbitrary number of different and plausible CTVs to mimic the inter‐reader variability in CTV delineation. The proposed model incorporates a separate encoder to extract features from the GTV masks, leveraging the critical role of GTV information in accurate CTV delineation. Results The proposed diffusion model demonstrated superior performance with the highest Dice Index (0.902 compared to values below 0.881 for state‐of‐the‐art models) and the best generalized energy distance (GED) (0.209 compared to values exceeding 0.221 for state‐of‐the‐art models). It also achieved the second‐highest recall and precision metrics among the compared ambiguous image segmentation models. Results from both datasets exhibited consistent trends, reinforcing the reliability of our findings. Additionally, ablation studies exploring different model structures and input configurations highlighted the significance of incorporating prior GTV information for accurate CTV delineation. Conclusions The proposed diffusion model successfully generates multiple plausible CTV contours for soft tissue sarcomas, effectively capturing inter‐reader variability in CTV delineation.

Dong, Yafei [Yale Biomedical Imaging Institute Yal↗
DOE OSTI2024

Coupling a recurrent neural network to SPAD TCSPC systems for real-time fluorescence lifetime imaging

Fluorescence lifetime imaging (FLI) has been receiving increased attention in recent years as a powerful diagnostic technique in biological and medical research. However, existing FLI systems often suffer from a tradeoff between processing speed, accuracy, and robustness. Inspired by the concept of Edge Artificial Intelligence (Edge AI), we propose a robust approach that enables fast FLI with no degradation of accuracy. This approach couples a recurrent neural network (RNN), which is trained to estimate the fluorescence lifetime directly from raw timestamps without building histograms, to SPAD TCSPC systems, thereby drastically reducing transfer data volumes and hardware resource utilization, and enabling real-time FLI acquisition. We train two variants of the RNN on a synthetic dataset and compare the results to those obtained using center-of-mass method (CMM) and least squares fitting (LS fitting). Results demonstrate that two RNN variants, gated recurrent unit (GRU) and long short-term memory (LSTM), are comparable to CMM and LS fitting in terms of accuracy, while outperforming them in the presence of background noise by a large margin. To explore the ultimate limits of the approach, we derive the Cramer-Rao lower bound of the measurement, showing that RNN yields lifetime estimations with near-optimal precision. To demonstrate real-time operation, we build a FLI microscope based on an existing SPAD TCSPC system comprising a 32 x 32 SPAD sensor named Piccolo. Four quantized GRU cores, capable of processing up to 4 million photons per second, are deployed on the Xilinx Kintex-7 FPGA that controls the Piccolo. Powered by the GRU, the FLI setup can retrieve real-time fluorescence lifetime images at up to 10 frames per second. The proposed FLI system is promising and ideally suited for biomedical applications, including biological imaging, biomedical diagnostics, and fluorescence-assisted surgery, etc.

47 OTHER INSTRUMENTATION↗
DOE OSTI2021

Diagnosing Malaria Patients with Plasmodium falciparum and vivax Using Deep Learning for Thick Smear Images

We propose a new framework, PlasmodiumVF-Net, to analyze thick smear microscopy images for a malaria diagnosis on both image and patient-level. Our framework detects whether a patient is infected, and in case of a malarial infection, reports whether the patient is infected by Plasmodium falciparum or Plasmodium vivax. PlasmodiumVF-Net first detects candidates for Plasmodium parasites using a Mask Regional-Convolutional Neural Network (Mask R-CNN), filters out false positives using a ResNet50 classifier, and then follows a new approach to recognize parasite species based on a score obtained from the number of detected patches and their aggregated probabilities for all of the patient images. Reporting a patient-level decision is highly challenging, and therefore reported less often in the literature, due to the small size of detected parasites, the similarity to staining artifacts, the similarity of species in different development stages, and illumination or color variations on patient-level. We use a manually annotated dataset consisting of 350 patients, with about 6000 images, which we make publicly available together with this manuscript. Our framework achieves an overall accuracy above 90% on image and patient-level.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2021

Mixture Model Framework for Traumatic Brain Injury Prognosis Using Heterogeneous Clinical and Outcome Data

Prognoses of Traumatic Brain Injury (TBI) outcomes are neither easily nor accurately determined from clinical indicators. This is due in part to the heterogeneity of damage inflicted to the brain, ultimately resulting in diverse and complex outcomes. Using a data-driven approach on many distinct data elements may be necessary to describe this large set of outcomes and thereby robustly depict the nuanced differences among TBI patients’ recovery. In this work, we develop a method for modeling large heterogeneous data types relevant to TBI. Our approach is geared toward the probabilistic representation of mixed continuous and discrete variables with missing values. The model is trained on a dataset encompassing a variety of data types, including demographics, blood-based biomarkers, and imaging findings. In addition, it includes a set of clinical outcome assessments at 3, 6, and 12 months post-injury. The model is used to stratify patients into distinct groups in an unsupervised learning setting. We use the model to infer outcomes using input data, and show that the collection of input data reduces uncertainty of outcomes over a baseline approach. In addition, we quantify the performance of a likelihood scoring technique that can be used to self-evaluate the extrapolation risk of prognosis on unseen patients.

97 MATHEMATICS AND COMPUTING↗