A5060546007- Advanced Radiotherapy Techniques
- Radiomics and Machine Learning in Medical Imaging
- Medical Imaging Techniques and Applications
- Head and Neck Cancer Studies
- Advanced X-ray and CT Imaging
- Radiation Therapy and Dosimetry
- Lung Cancer Diagnosis and Treatment
- Radiation Dose and Imaging
- Glioma Diagnosis and Treatment
- AI in cancer detection
- Management of metastatic bone disease
- Advances in Oncology and Radiotherapy
- Brain Tumor Detection and Classification
- Quantum Mechanics and Applications
- Quantum Information and Cryptography
- Medical Imaging and Analysis
- Advanced Thermodynamics and Statistical Mechanics
- Medical Image Segmentation Techniques
- Advanced MRI Techniques and Applications
- Viral Infections and Vectors
- Boron Compounds in Chemistry
- Zoonotic diseases and public health
- Big Data and Business Intelligence
- Pancreatic and Hepatic Oncology Research
- Advanced Neural Network Applications
The University of Texas Health Science Center at Houston
2021-2025
The University of Texas MD Anderson Cancer Center
2017-2025
University of Washington
2024
University of Georgia
2023
University of Edinburgh
2022
Purpose Radiation therapy treatment planning is a time‐consuming and iterative manual process. Consequently, plan quality varies greatly between within institutions. Artificial intelligence shows great promise in improving reducing times. This technical note describes our participation the American Association of Physicists Medicine Open Knowledge‐Based Planning Challenge (OpenKBP), competition to accurately predict radiation dose distributions. Methods A three‐dimensional (3D) densely...
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...
To develop a deep learning model that generates consistent, high-quality lymph node clinical target volumes (CTV) contours for head and neck cancer (HNC) patients, as an integral part of fully automated radiation treatment planning workflow.
The Radiation Planning Assistant (RPA) is a system developed for the fully automated creation of radiotherapy treatment plans, including volume-modulated arc therapy (VMAT) plans patients with head/neck cancer and 4-field box cervical cancer. It combination specially in-house software that uses an application programming interface to communicate commercial planning system. also interfaces secondary dose verification software. necessary inputs are Treatment Plan Order, approved by radiation...
Abstract Manually delineating upper abdominal organs at risk (OARs) is a time-consuming task. To develop deep-learning-based tool for accurate and robust auto-segmentation of these OARs, forty pancreatic cancer patients with contrast-enhanced breath-hold computed tomographic (CT) images were selected. We trained three-dimensional (3D) U-Net ensemble that automatically segments all organ contours concurrently the self-configuring nnU-Net framework. Our tool’s performance was assessed on...
Objective.To establish an open framework for developing plan optimization models knowledge-based planning (KBP).Approach.Our includes radiotherapy treatment data (i.e. reference plans) 100 patients with head-and-neck cancer who were treated intensity-modulated radiotherapy. That also high-quality dose predictions from 19 KBP that developed by different research groups using out-of-sample during the OpenKBP Grand Challenge. The input to four fluence-based mimicking form 76 unique pipelines...
Abstract Background Recent studies have shown deep learning techniques are able to predict three‐dimensional (3D) dose distributions of radiotherapy treatment plans. However, their use in prediction for treatments with varied prescription doses including simultaneous integrated boost (SIB), that is, using multiple within the same plan, and benefit improving plan quality should be validated. Purpose To investigate feasibility potential distribution volumetric modulated arc therapy (VMAT) SIB...
Pancreatic gross tumor volume (GTV) delineation is challenging due to their variable morphology and uncertain ground truth. Previous deep learning-based auto-segmentation methods have struggled handle tasks with truth not accommodated stylistic customizations. We aim develop a human-in-the-loop pancreatic GTV segmentation tool using Tversky ensembles by leveraging uncertainty estimation techniques. In this study, we utilized total of 282 patients from the pancreas task Medical Segmentation...
The sharpness of the kernels used for image reconstruction in computed tomography affects values quantitative features. We sought to identify that produce similar feature enable a more effective comparison images produced using scanners from different manufactures. also investigated new filter designed change kernel-related component frequency spectrum postreconstruction initial kernel preferred kernel. A radiomics texture phantom was imaged GE, Philips, Siemens, and Toshiba. Images were...
Abstract Background In recent years, deep‐learning models have been used to predict entire three‐dimensional dose distributions. However, the usability of predictions improve plan quality should be further investigated. Purpose To develop a model high‐quality distributions for volumetric modulated arc therapy (VMAT) plans patients with gynecologic cancer and evaluate their in driving improvements. Methods A total 79 VMAT female pelvis were train (47 plans), validate (16 test plans) 3D dense...
A large number of surveys have been sent to the medical physics community addressing many clinical topics for which physicist is, or may be, responsible. Each survey provides an insight into practice relevant community. The goal this study was create a summary these giving snapshot patterns. Surveys used in were created using SurveyMonkey and distributed between February 6, 2013 January 2, 2018 via MEDPHYS MEDDOS listserv groups. format included questions that multiple choice free response....
Abstract The purpose of this study is to investigate the dosimetric impact multi‐leaf collimator (MLC) positioning errors on a Varian Halcyon for both random and systematic errors, evaluate effectiveness portal dosimetry quality assurance in catching clinically significant changes caused by these errors. Both were purposely added 11 physician‐approved head neck volumetric modulated arc therapy (VMAT) treatment plans, yielding total 99 unique plans. Plans then delivered preclinical linear...
Abstract A consensus species tree is reconstructed from 11 gene trees for human, bat, and pangolin beta coronaviruses samples taken early in the pandemic (prior to April 1, 2020). Using coalescent theory, shallow (short branches relative hosts) provides evidence of recent flow events between bat predating zoonotic transfer humans. The was also used reconstruct ancestral sequence human SARS-CoV-2, which 2 nucleotides different Wuhan sequence. time most common ancestor estimated be Dec 8, 2019...
Journal Article Study of a method based on TLD detectors for in-phantom dosimetry in BNCT Get access G. Gambarini, Gambarini *Corresponding author: grazia.gambarini@mi.infn.it Search other works by this author on: Oxford Academic PubMed Google Scholar V. Klamert, Klamert S. Agosteo, Agosteo C. Birattari, Birattari Gay, Gay Rosi, Rosi L. Scolari Radiation Protection Dosimetry, Volume 110, Issue 1-4, 1 August 2004, Pages 631–636, https://doi.org/10.1093/rpd/nch109 Published: 01 2004
The Radiation Planning Assistant (RPA) is a system developed for the fully automated creation of radiotherapy treatment plans, including volume-modulated arc therapy (VMAT) plans patients with head/neck cancer and 4-field box cervical cancer. It combination specially in-house software that uses an application programming interface to communicate commercial planning system. also interfaces secondary dose verification software. necessary inputs are Treatment Plan Order, approved by radiation...
Radiomics studies require large patient cohorts, which often include patients imaged using different imaging protocols. We aimed to determine the impact of variability in protocol parameters and interscanner a phantom that produced feature values similar those patients. Positron emission tomography (PET) scans Hoffman brain were acquired on GE Discovery 710, Siemens mCT, Philips Vereos scanners. A standard-protocol scan was each machine, then parameter could be changed altered individually....
Target delineation for radiation therapy is a time-consuming and complex task. Autocontouring gross tumor volumes (GTVs) has been shown to increase efficiency. However, there limited literature on post-operative target delineation, particularly CT-based studies. To this end, we trained autocontouring model contour the GTV of pediatric patients with medulloblastoma.One hundred four retrospective CT scans were used train auto-contouring model. Eighty then preselected visibility, continuity,...
Planning for palliative radiotherapy is performed without the advantage of MR or PET imaging in many clinics. Here, we investigated CT-only GTV delineation treatment head and neck cancer. Two multi-institutional datasets palliative-intent plans were retrospectively acquired: a set 102 non-contrast-enhanced CTs 96 contrast-enhanced CTs. The nnU-Net auto-segmentation network was chosen its strength medical image segmentation, five approaches separately trained: (1) heuristic-cropped,...
Abstract Background The delineation of clinical target volumes (CTVs) for radiotherapy nasopharyngeal cancer is complex and varies based on the location extent disease. Purpose current study aimed to develop an auto‐contouring solution following one protocol guidelines (NRG‐HN001) that can be adjusted meet other guidelines, such as RTOG‐0225 2018 International guidelines. Methods used 2‐channel 3‐dimensional U‐Net nnU‐Net framework auto‐contour 27 normal structures in head neck (H&N)...
We conducted our study to develop a tool capable of automatically detecting dental artifacts in CT scan on slice-by-slice basis and assess the dosimetric impact implementing into Radiation Planning Assistant (RPA), web-based platform designed fully automate radiation therapy treatment planning process.We developed an automatic artifact identification assessed its use RPA. Three users manually annotated 83,676 head-and-neck (HN) slices (549 patients). Majority-voting was applied individual...
To investigate the inter- and intra-fraction motion associated with use of a low-cost tape immobilization technique as an alternative to thermoplastic masks for whole-brain treatments. The results this study may be interest clinical staff severely limited resources (e.g., in low-income countries) also when treating patients who cannot tolerate standard masks. Setup reproducibility eight healthy volunteers was assessed two different techniques. (a) One strip placed across volunteer's forehead...
Two-dimensional radiotherapy is often used to treat cervical cancer in low- and middle-income countries, but treatment planning can be challenging time-consuming. Neural networks offer the potential greatly decrease time through automation, impact of wide range hyperparameters set during training on model accuracy has not been exhaustively investigated. In current study, we evaluated effect several convolutional neural network architectures 2D field delineation.Six commonly deep learning...
Strong magnetic fields affect radiation dose deposition in MRI-guided therapy systems, particularly at interfaces between tissues of differing densities such as those the thorax. In this study, we evaluated impact a 1.5 T field on radiation-induced lung damage C57L/J mice. We irradiated 140 mice to whole thorax with parallel-opposed Co-60 beams doses 0, 9.0, 10.0, 10.5, 11.0, 12.0, or 13.0 Gy (20 per group). Ten group were while was applied transverse beam and ten set 0 T. compared survival...