A5015909611- Fetal and Pediatric Neurological Disorders
- Domain Adaptation and Few-Shot Learning
- Prenatal Screening and Diagnostics
- Artificial Intelligence in Healthcare and Education
- Cleft Lip and Palate Research
- Congenital Heart Disease Studies
- AI in cancer detection
- Congenital Diaphragmatic Hernia Studies
- Advanced Neural Network Applications
- Radiomics and Machine Learning in Medical Imaging
- Generative Adversarial Networks and Image Synthesis
- Ultrasound Imaging and Elastography
- Neonatal and fetal brain pathology
- COVID-19 diagnosis using AI
- Neonatal Respiratory Health Research
- Medical Image Segmentation Techniques
- Radiology practices and education
- Autopsy Techniques and Outcomes
- Pregnancy and preeclampsia studies
- Flow Measurement and Analysis
- Electrical and Bioimpedance Tomography
- Infant Development and Preterm Care
- Face recognition and analysis
- Blood transfusion and management
- Medical Imaging and Analysis
King's College London
2016-2025
Guy's and St Thomas' NHS Foundation Trust
2016-2024
St Thomas' Hospital
2018-2024
University College London
2022
University College Hospital
2022
City, University of London
2021
Imperial College London
2020
The King's College
2018
Kings Health Partners
2018
Memorial University of Newfoundland
1993-1995
Identifying and interpreting fetal standard scan planes during 2-D ultrasound mid-pregnancy examinations are highly complex tasks, which require years of training. Apart from guiding the probe to correct location, it can be equally difficult for a non-expert identify relevant structures within image. Automatic image processing provide tools help experienced as well inexperienced operators with these tasks. In this paper, we propose novel method based on convolutional neural networks,...
Abstract The current approach to fetal anomaly screening is based on biometric measurements derived from individually selected ultrasound images. In this paper, we introduce a paradigm shift that attains human-level performance in measurement by aggregating automatically extracted biometrics every frame across an entire scan, with no need for operator intervention. We use neural network classify each of video recording. then measure where appropriate anatomy visible. Bayesian method estimate...
Summary Inadequate zinc intake may lead to poor growth and developmental outcome in very-low-birth-weight (VLBW; < 1,500 g) infants. Fifty-two infants (mean birth weight, 1,117 287 g; mean gestational age, 29 ± 2.9 weeks) were randomly allocated two groups. SUPP received a regular term formula plus supplements (4.4 mg/L; final content, 11 mg/L); PLAC the same placebo (final 6.7 mg/L). Infants started their at 1,853 109 g, consumed for 6 months. All subjects evaluated 3, 6, 9, 12 0.75 months...
Introduction: The use of artificial intelligence (AI) in medical imaging and radiotherapy has been met with both scepticism excitement. However, clinical integration AI is already well-underway. Many authors have recently reported on the knowledge perceptions radiologists/medical staff students however there a paucity information regarding radiographers. Published literature agrees that likely to significant impact radiology practice. As radiographers are at forefront service delivery, an...
Artificial intelligence (AI)-enabled applications are increasingly being used in providing healthcare services, such as medical imaging support. Sufficient and appropriate education for professionals is required successful AI adoption. Although, currently, there training programmes radiologists, formal radiographers lacking. Therefore, this study aimed to evaluate discuss a postgraduate-level module on developed the UK radiographers.A participatory action research methodology was applied,...
In this work, we apply an attention-gated network to real-time automated scan plane detection for fetal ultrasound screening. Scan in is a challenging problem due the poor image quality resulting low interpretability both clinicians and algorithms. To solve this, propose incorporating self-gated soft-attention mechanisms. A mechanism generates gating signal that end-to-end trainable, which allows contextualise local information useful prediction. The proposed attention generic it can be...
Measurement of head biometrics from fetal ultrasonography images is key importance in monitoring the healthy development fetuses. However, accurate measurement relevant anatomical structures subject to large inter-observer variability clinic. To address this issue, an automated method utilizing Fully Convolutional Networks (FCN) proposed determine measurements circumference (HC) and biparietal diameter (BPD). An FCN was trained on approximately 2000 2D ultrasound with annotations provided by...
Detecting acoustic shadows in ultrasound images is important many clinical and engineering applications. Real-time feedback of can guide sonographers to a standardized diagnostic viewing plane with minimal artifacts provide additional information for other automatic image analysis algorithms. However, automatically detecting shadow regions using learning-based algorithms challenging because pixel-wise ground truth annotation subjective time consuming. In this paper, we propose weakly...
Deep neural networks exhibit limited generalizability across images with different entangled domain features and categorical features. Learning generalizable that can form universal decision boundaries domains is an interesting difficult challenge. This problem occurs frequently in medical imaging applications when attempts are made to deploy improve deep learning models image acquisition devices, parameters or if some classes unavailable new training databases. To address this problem, we...
IntroductionRadiographer reporting is accepted practice in the UK. With a national shortage of radiographers and radiologists, artificial intelligence (AI) support may help minimise backlog unreported images. Modern AI not well understood by human end-users. This have ethical implications impact trust these systems, due to over- under-reliance. study investigates perceptions about AI, gathers information explain how they interact with future identifies features perceived as necessary for...
Congenital heart disease (CHD) is common and associated with impaired early brain development neurodevelopmental outcomes, yet the exact mechanisms underlying these associations are unclear.
Advances in artificial intelligence (AI) have demonstrated potential to improve medical diagnosis. We piloted the end-to-end automation of mid-trimester screening ultrasound scan using AI-enabled tools.A prospective method comparison study was conducted. Participants had both standard and AI-assisted US scans performed. The AI tools automated image acquisition, biometric measurement, report production. A feedback survey captured sonographers' perceptions scanning.Twenty-three subjects were...
Abstract Background Artificial intelligence (AI) has the potential to improve prenatal detection of congenital heart disease. We analysed performance current national screening programme in detecting hypoplastic left syndrome (HLHS) compare with our own AI model. Methods Current was calculated from local and sources. models were trained using four‐chamber ultrasound views fetal heart, a ResNet classifier. Results Estimated sensitivity specificity for HLHS 94.3% 99.985%, respectively....
Abstract Background Persistent, high rates of maternal mortality amongst ethnic minorities is one the UK’s starkest examples racial disparity. With greater risks adverse outcomes during maternity care, minority women are subjected to embedded, structural and systemic discrimination throughout healthcare service. Methods Fourteen semi-structured interviews were undertaken with who had recent experience UK care. Data pertaining ethnicity race subject iterative, inductive coding, constant...
Artificial intelligence (AI) has shown promise in improving the performance of fetal ultrasound screening detecting congenital heart disease (CHD). The effect giving AI advice to human operators not been studied this context. Giving additional information about model workings, such as confidence scores for predictions, may be a way further performance. Our aims were investigate whether improved overall diagnostic accuracy (using single CHD lesion an exemplar), and determine what, if any,...
Abstract This study explores the potential of 3D Slice-to-Volume Registration (SVR) motion-corrected fetal MRI for craniofacial assessment, traditionally used only brain analysis. In addition, we present first description an automated pipeline based on Attention UNet trained segmentation, followed by surface refinement. Results printing selected models are also presented. Qualitative analysis multiplanar volumes, SVR output and segmentations outputs, were assessed with computer printed...
ABSTRACT Objective To use artificial intelligence (AI) to automatically extract video clips of the fetal heart from a stream ultrasound video, and assess performance these when used for remote second review. Methods Using dataset previous clinical trial AI assist in scanning, was scans 48 fetuses which diagnosis known: 24 normal with congenital disease (CHD). These, manually still saved images, were shown random order expert clinicians, who asked detect cardiac abnormalities. Results The...
Background: The detailed assessment of fetal brain maturation and development involves morphological evaluation, gyration analysis, reliable biometric measurements. Manual measurements on conventional 2-D magnetic resonance imaging (MRI) are affected by motion there is no clear consensus regarding definitions for parameters anatomical landmark placements, making consistent reference plane slice selection challenging. Automated biometry with 3-D slice-to-volume reconstruction (SVR) has the...