A5030148633- Radiomics and Machine Learning in Medical Imaging
- Advanced Radiotherapy Techniques
- Medical Imaging Techniques and Applications
- Photoacoustic and Ultrasonic Imaging
- Ultrasound and Hyperthermia Applications
- Advanced MRI Techniques and Applications
- Glioma Diagnosis and Treatment
- MRI in cancer diagnosis
- Advanced X-ray and CT Imaging
- Infrared Thermography in Medicine
- Hepatocellular Carcinoma Treatment and Prognosis
- Advanced NMR Techniques and Applications
- AI in cancer detection
- Liver Disease Diagnosis and Treatment
- Brain Tumor Detection and Classification
- Nanoplatforms for cancer theranostics
- Advanced Neural Network Applications
- Medical Imaging and Analysis
- Mathematical Biology Tumor Growth
- Medical Image Segmentation Techniques
- Brain Metastases and Treatment
- Enzyme Structure and Function
- Cancer Genomics and Diagnostics
- Atomic and Subatomic Physics Research
- Pancreatic and Hepatic Oncology Research
The University of Texas MD Anderson Cancer Center
2016-2025
Universidad de Alcalá
2020-2024
The University of Texas Health Science Center at Houston
2011-2024
Scripps MD Anderson Cancer Center
2024
Universidad de Los Andes, Chile
2024
University of Sharjah
2024
Royal Brompton Hospital
2023
The University of Texas at Austin
2006-2022
Rice University
2020-2022
ORCID
2022
Landmark point-pairs provide a strategy to assess deformable image registration (DIR) accuracy in terms of the spatial underlying anatomy depicted medical images. In this study, we propose augment publicly available database (www.dir-lab.com) images with large sets manually identified anatomic feature pairs between breath-hold computed tomography (BH-CT) for DIR evaluation. Ten BH-CT were randomly selected from COPDgene study cases. Each patient had received CT imaging entire thorax supine...
Purpose To evaluate a fully automated machine learning algorithm that uses pretherapeutic quantitative CT image features and clinical factors to predict hepatocellular carcinoma (HCC) response transcatheter arterial chemoembolization (TACE). Materials Methods Outcome information from 105 patients receiving first-line treatment with TACE was evaluated retrospectively. The primary endpoint time progression (TTP) based on follow-up radiologic criteria (modified Response Evaluation Criteria in...
Purpose: This study aimed to use deep learning-based dose prediction assess head and neck (HN) plan quality identify suboptimal plans. Methods: A total of 245 VMAT HN plans were created using RapidPlan knowledge-based planning (KBP). subset 112 high-quality was selected under the supervision an radiation oncologist. We trained a 3D Dense Dilated U-Net architecture predict 3-dimensional distributions 3-fold cross-validation on 90 Model inputs included CT images, target prescriptions, contours...
Hyperpolarized [1-(13)C]-pyruvate has shown tremendous promise as an agent for imaging tumor metabolism with unprecedented sensitivity and specificity. Imaging hyperpolarized substrates by magnetic resonance is unlike traditional MRI because signals are highly transient their spatial distribution varies continuously over observable lifetime. Therefore, new approaches needed to ensure optimal measurement under these circumstances. Constrained reconstruction algorithms can integrate prior...
Purpose To evaluate the uncertainty of radiomics features from contrast-enhanced breath-hold helical CT scans non-small cell lung cancer for both manual and semi-automatic segmentation due to intra-observer, inter-observer, inter-software reliability. Methods Three radiation oncologists manually delineated tumors twice 10 using two software tools (3D-Slicer MIM Maestro). Additionally, three observers without formal clinical training were instructed use tools, Lesion Sizing Toolkit (LSTK)...
Artificial intelligence (AI) is the most revolutionizing development in health care industry current decade, with diagnostic imaging having greatest share such development. Machine learning and deep (DL) are subclasses of AI that show breakthrough performance image analysis. They have become state art field classification recognition. deals extraction important characteristic features from images, whereas DL uses neural networks to solve problems better performance. In this review, we...
Hepatocellular carcinoma (HCC) is the most common primary liver neoplasm, and its incidence has doubled over past two decades owing to increasing risk factors. Despite surveillance, HCC cases are diagnosed at advanced stages can only be treated using transarterial chemo-embolization (TACE) or systemic therapy. TACE failure may occur with reaching up 60% of cases, leaving patients a financial emotional burden. Radiomics emerged as new tool capable predicting tumor response from pre-procedural...
Undersampling of gliomas at first biopsy is a major clinical problem, as accurate grading determines all subsequent treatment. We submit technological solution to reduce the problem undersampling by estimating marker tumor proliferation (Ki-67) using MR imaging data inputs, against stereotactic histopathology gold standard. was performed with anatomic, diffusion, permeability, and perfusion sequences, in untreated glioma patients prospective trial. Stereotactic biopsies were harvested from...
Medical imaging deep learning models are often large and complex, requiring specialized hardware to train evaluate these models. To address such issues, we propose the PocketNet paradigm reduce size of by throttling growth number channels in convolutional neural networks. We demonstrate that, for a range segmentation classification tasks, architectures produce results comparable that conventional networks while reducing parameters multiple orders magnitude, using up 90% less GPU memory,...
Abstract Purpose Pediatric patients with medulloblastoma in low‐ and middle‐income countries (LMICs) are most treated 3D‐conformal photon craniospinal irradiation (CSI), a time‐consuming, complex treatment to plan, especially resource‐constrained settings. Therefore, we developed tested CSI autoplanning tool for varying patient lengths. Methods materials Autocontours were generated deep learning model trained:tested (80:20 ratio) on 143 pediatric CT scans (patient ages: 2–19 years, median =...
Purpose Several methods in MRI use the phase information of complex signal and require unwrapping (e.g., B0 field mapping, chemical shift imaging, velocity measurements). In this work, an algorithm was developed focusing on needs requirements MR temperature imaging applications. Methods The proposed method performs fully automatic using a list all pixels sorted by magnitude descending order creates merges clusters unwrapped until entire image is unwrapped. evaluated simulated phantom data...
Abstract BACKGROUND Laser Interstitial Thermal Therapy (LITT) has been used to treat recurrent brain metastasis after stereotactic radiosurgery (SRS). Little is known about how best assess the efficacy of treatment, specifically ability LITT control local tumor progression post-SRS. OBJECTIVE To evaluate predictive factors associated with recurrence LITT. METHODS Retrospective study consecutive patients metastases treated Based on radiological aspects, lesions were divided into progressive...
Historically, clinician-derived contouring of tumors and healthy tissues has been crucial for radiotherapy (RT) planning. In recent years, advances in artificial intelligence (AI), predominantly deep learning (DL), have rapidly improved automated RT applications, particularly routine organs-at-risk 1–3. Despite research efforts actively promoting its broader acceptance, clinical adoption auto-contouring is not yet standard practice.
Currently, a majority of institution-specific automatic MRI-based contouring algorithms are trained, tested, and validated on one contrast weighting (i.e., T2-weighted), however their actual performance within this across different repetition times, TR, echo TE) is under-investigated poorly understood. As result, external institutions with scan protocols for the same may experience sub-optimal performance. The purpose study was to develop method evaluate robustness varying MRI weightings....
Physics-informed neural networks (PINNs) have gained significant attention for solving forward and inverse problems related to partial differential equations (PDEs). While advancements in loss functions network architectures improved PINN accuracy, the impact of collocation point sampling on their performance remains underexplored. Fixed methods, such as uniform random equispaced grids, can fail capture critical regions with high solution gradients, limiting effectiveness complex PDEs....
In recent years, there has been an increasing interest in using deep learning and neural networks to tackle scientific problems, particularly solving partial differential equations (PDEs). However, many network-based methods, such as physics-informed networks, depend on automatic differentiation the sampling of collocation points, which can result a lack interpretability lower accuracy compared traditional numerical methods. To address this issue, we propose two approaches for discontinuous...