A5031932354- Photoacoustic and Ultrasonic Imaging
- Ultrasound and Hyperthermia Applications
- Medical Imaging Techniques and Applications
- Ultrasound and Cavitation Phenomena
- Radiomics and Machine Learning in Medical Imaging
- Ultrasound Imaging and Elastography
- Advanced Radiotherapy Techniques
- Pregnancy and preeclampsia studies
- AI in cancer detection
- Prenatal Screening and Diagnostics
- Advanced X-ray and CT Imaging
- Neonatal Respiratory Health Research
- Lanthanide and Transition Metal Complexes
- Head and Neck Cancer Studies
- Surgical Simulation and Training
- Cardiovascular Health and Disease Prevention
- Assisted Reproductive Technology and Twin Pregnancy
- Glioma Diagnosis and Treatment
- Digital Radiography and Breast Imaging
- MRI in cancer diagnosis
- Management of metastatic bone disease
- Hemodynamic Monitoring and Therapy
- Medical Image Segmentation Techniques
- Renal and Vascular Pathologies
- Smart Systems and Machine Learning
Mirada Medical (United Kingdom)
2020-2023
University of Oxford
2017-2021
John Radcliffe Hospital
2018
Royal Surrey County Hospital
2015
University of Patras
2014
University of Edinburgh
2009-2012
We present a new technique to fully automate the segmentation of an organ from 3D ultrasound (3D-US) volumes, using placenta as target organ. Image analysis tools estimate volume do exist but are too time consuming and operator dependant. Fully automating process would potentially allow use placental screen for increased risk pregnancy complications. The was segmented 2,393 first trimester 3D-US volumes semiautomated technique. This quality controlled by three operators produce...
Placental volume measured with 3D ultrasound in the first trimester has been shown to be correlated adverse pregnancy outcomes. This could potentially used as a screening test predict "at risk" pregnancy. However, manual segmentation whilst previously accurate and repeatable is very time consuming semi-automated methods still require operator input. To generate tool, fully automated placental required. In this work, deep convolutional neural network (cNN), DeepMedic, was trained using output...
Background and purposeAuto-contouring performance has been widely studied in development commissioning studies radiotherapy, its impact on clinical workflow assessed that context. This study aimed to evaluate the manual adjustment of auto-contouring routine practice identify improvements regarding model user interaction, improve efficiency auto-contouring.Materials methodsA total 103 head neck cancer cases, contoured using a commercial deep-learning contouring system subsequently checked...
Volumetric placental measurement using 3-D ultrasound has proven clinical utility in predicting adverse pregnancy outcomes. However, this metric cannot currently be employed as part of a screening test due to lack robust and real-time segmentation tools. We present multiclass (MC) convolutional neural network (CNN) developed segment the placenta, amniotic fluid, fetus. The ground-truth data set consisted 2093 labeled volumes augmented by 300 with fetus annotated. A two-pathway, hybrid (HB)...
Background and purposeUser-adjustments after deep-learning (DL) contouring in radiotherapy were evaluated to get insight real-world editing during clinical practice. This study assessed the amount, type spatial regions of auto-contouring for organs-at-risk (OARs) routine workflow patients thorax region.Materials methodsA total 350 lung cancer 362 breast patients, contoured between March 2020 2021 using a commercial DL-contouring method followed by manual adjustments retrospectively analyzed....
Background Three-dimensional (3D) fractional moving blood volume (FMBV) derived from 3D power Doppler US has been proposed for noninvasive approximation of perfusion. However, FMBV never applied in animals against a ground truth. Purpose To determine the correlation between and reference standard fluorescent microspheres (FMS) measurement renal perfusion porcine model. Materials Methods From February 2017 to September 2017, adult pigs were administered FMS before after at baseline (100%)...
Large numbers of acoustic signals from single lipid-shelled Definity® microbubbles have been measured using a calibrated microacoustic system and two population response observed. Theoretical results based on the Mooney–Rivlin strain softening shell model used to identify these populations as primary resonant off-primary scatter. An experimentally size distribution was provide initial resting radius for simulations, responses agree well with experimental data. In this way, or behavior...
PurposeAtlas-based and deep-learning contouring (DLC) are methods for automatic segmentation of organs-at-risk (OARs). The European Particle Therapy Network (EPTN) published a consensus-based atlas delineation OARs in neuro-oncology. In this study, geometric dosimetric evaluation automatically-segmented neuro-oncological was performed using CT- MR-models following the EPTN-contouring atlas.MethodsImage data from 76 patients were included. Two atlas-based models (CT-atlas MR-atlas) one...
The Kretzfile format is used to store 3D ultrasound data, from GE Voluson scanners. geometry in these hies a toroidal coordinate system. Cartesian coordinates are required allow application of advanced image libraries like ITK and scikit-image. We present transformation utilities convert Kretzfiles cartesian coordinates. Previous work (SlicerHeart, 2017) has enabled the reading kretz files approximate transformations. This will enable medical imaging researchers investigate clinically ultrasound.
Power Doppler ultrasound imaging provides a non-invasive method to explore tissue vascularity in real-time. It has farreaching clinical utility, for example assess the perfusion of organs such as placenta. In this study, fully automated was proposed standardize power signals placental bed estimate by technique known fractional moving blood volume (FMBV). The uterine vasculature extracted region growing algorithm from 3D scan and further localized based on placenta segmentation obtained...
Radiological imaging is fundamental within the healthcare industry and has become routinely adopted for diagnosis, disease monitoring treatment planning. With advent of digital modalities rapid growth in both diagnostic therapeutic imaging, ability to be able harness this large influx data paramount importance. The Oncology Medical Image Database (OMI-DB) was created provide a centralized, fully annotated dataset research. database contains processed unprocessed images, associated data,...
The study of single microbubble (MB) acoustics has relevance to signal and image processing ultrasound (US) contrast imaging. aim this work is the effect consecutive pulses on MB response. RF backscatter data from 223 lipid shelled Definity® MBs were collected using a flow phantom modified Philips Sonos 5500 scanner. results 2 imaging (1.6MHz, 550kPa, 6 cycles, 1kHz PRF) are presented compared similar theoretical model Definity (Keller-Miksis equation, Mooney-Rivlin shell model). Bubble...
Ultrasound contrast agents are gas filled microbubbles which produced enhanced echoes in ultrasound imaging thus allowing the acquisition of detailed information on path blood. It is theoretically known that size a vessel affects behavior microbubble, could potentially be used to discriminate different sized vessels. This would useful monitoring neovascularization tumor growth or treatment. However, currently it not possible identify diameter by any means signal processing microbubble...
Nonlinear microbubble oscillations can be exploited by signal processing to provide contrast enhancement in ultrasound imaging. High-speed optical investigations have established that microbubbles resonate in-vitro, but resonance has not yet been in-vivo. Here we aim establish the acoustic signature of resonance, which will help identify its occurrence in-vivo and further optimize processing. Resonant a transient amplitude response, is unique resonance. Large numbers signals from single...
The response of the rigid shelled contrast agent biSphere® has been previously studied optically and acoustically over a range mechanical index (MI). Free bubble models have used to model behaviour microbubbles (MBs). In this work new theoretical is proposed compared with available ones in terms agreement experimental acoustic responses from single MBs.