Computational fluid dynamics (CFD) can be used to simulate vascular haemodynamics and analyse potential treatment options. CFD has shown to be beneficial in improving patient outcomes. However, the implementation of CFD for routine clinical use is yet to be realised. Barriers for CFD include high computational resources, specialist experience needed for designing simulation set-ups, and long processing times. The aim of this study was to explore the use of machine learning (ML) to replicate conventional aortic CFD with automatic and fast regression models. Data used to train/test the model consisted of 3,000 CFD simulations performed on synthetically generated 3D aortic shapes. These subjects were generated from a statistical shape model (SSM) built on real patient-specific aortas (N = 67). Inference performed on 200 test shapes resulted in average errors of 6.01% ±3.12 SD and 3.99% ±0.93 SD for pressure and velocity, respectively. Our ML-based models performed CFD in ∼0.075 seconds (4,000x faster than the solver). This proof-of-concept study shows that results from conventional vascular CFD can be reproduced using ML at a much faster rate, in an automatic process, and with reasonable accuracy.
7 November 2022 in Journal of Cardiovascular Magnetic Resonance
Computational fluid dynamics (CFD) is increasingly used for the assessment of blood flow conditions in patients with congenital heart disease (CHD). This requires patient-specific anatomy, typically obtained from segmented 3D cardiovascular magnetic resonance (CMR) images. However, segmentation is time-consuming and requires expert input. This study aims to develop and validate a machine learning (ML) method for segmentation of the aorta and pulmonary arteries for CFD studies. 90 CHD patients were retrospectively selected for this study. 3D CMR images were manually segmented to obtain ground-truth (GT) background, aorta and pulmonary artery labels. These were used to train and optimize a U-Net model, using a 70-10-10 train-validation-test split. Segmentation performance was primarily evaluated using Dice score. CFD simulations were set up from GT and ML segmentations using a semi-automatic meshing and simulation pipeline. Mean pressure and velocity fields across 99 planes along the vessel centrelines were extracted, and a mean average percentage error (MAPE) was calculated for each vessel pair (ML vs GT). A second observer (SO) segmented the test dataset for assessment of inter-observer variability. Friedman tests were used to compare ML vs GT, SO vs GT and ML vs SO metrics, and pressure/velocity field errors. The network’s Dice score (ML vs GT) was 0.945 (interquartile range: 0.929–0.955) for the aorta and 0.885 (0.851–0.899) for the pulmonary arteries. Differences with the inter-observer Dice score (SO vs GT) and ML vs SO Dice scores were not statistically significant for either aorta or pulmonary arteries (p = 0.741, p = 0.061). The ML vs GT MAPEs for pressure and velocity in the aorta were 10.1% (8.5–15.7%) and 4.1% (3.1–6.9%), respectively, and for the pulmonary arteries 14.6% (11.5–23.2%) and 6.3% (4.3–7.9%), respectively. Inter-observer (SO vs GT) and ML vs SO pressure and velocity MAPEs were of a similar magnitude to ML vs GT (p > 0.2). ML can successfully segment the great vessels for CFD, with errors similar to inter-observer variability. This fast, automatic method reduces the time and effort needed for CFD analysis, making it more attractive for routine clinical use.
Congenital heart disease (CHD) is the most common defect at birth. Effective training for clinical professionals is essential in order to provide a high standard of care for patients. Visual aids for teaching complex CHD have remained mostly unchanged in recent years, with traditional methods such as diagrams and specimens still essential for delivering educational content. Diagrams and other 2D visualisations for teaching are in most cases artistic illustrations with no direct relation to true, 3D medical data. Specimens are rare, difficult for students to access and are limited to specific institutions. Digital, patient-specific models could potentially address these problems within educational programmes. Virtual Reality (VR) can facilitate the access to digital models and enhance the educational experience. In this study, we recorded and analysed the sentiment of clinical professionals towards VR when learning about CHD. A VR application (VheaRts) containing a set of patient-specific models was developed in-house. The application was incorporated into a specialised cardiac morphology course to assess the feasibility of integrating such a tool, and to measure levels of acceptance. Attendees were clinical professionals from a diverse range of specialities. VR allowed users to interact with six different patient-derived models immersed within a 3D space. Feedback was recorded for 58 participants. The general response towards the use of VR was overwhelmingly positive, with 88% of attendees rating 4 or 5 for ‘helpfulness of VR in learning CHD’ (5-points Likert scale). Additionally, 70% of participants with no prior VR experience rated 4 or 5 for ‘intuitiveness and ease of use’. Our study indicates that VR has a high level of acceptance amongst clinical trainees when used as an effective aid for learning congenital heart disease. Additionally, we noted three specific use-cases where VR offered novel teaching experiences not possible with conventional methods.
Computational fluid dynamics (CFD) can be used to simulate vascular haemodynamics and analyse potential treatment options. CFD has shown to be beneficial in improving patient outcomes. However, the implementation of CFD for routine clinical use is yet to be realised. Barriers for CFD include high computational resources, specialist experience needed for designing simulation set-ups, and long processing times. The aim of this study was to explore the use of machine learning (ML) to replicate conventional aortic CFD with automatic and fast regression models. Data used to train/test the model consisted of 3,000 CFD simulations performed on synthetically generated 3D aortic shapes. These subjects were generated from a statistical shape model (SSM) built on real patient-specific aortas (N=67). Inference performed on 200 test shapes resulted in average errors of 6.01% +/-3.12 SD and 3.99% +/-0.93 SD for pressure and velocity, respectively. Our ML-based models performed CFD in +/-0.075 seconds (4,000x faster than the solver). This proof-of-concept study shows that results from conventional vascular CFD can be reproduced using ML at a much faster rate, in an automatic process, and with high accuracy.
Accurate assessment of tumour anatomy is critical for neuroblastoma surgical planning. This study evaluated whether patient-specific virtual reality (VR) models improve the accuracy of pre-operative anatomical assessment compared with conventional computed tomography (CT) imaging. CT scans from nine patients with abdominal neuroblastoma were segmented to create patient-specific VR models, which were then imported into an in-house VR application. Eleven paediatric surgeons (n=5 consultants; n=6 registrars) independently evaluated conventional CT images and VR reconstructions to assess vascular and organ involvement, image-defined risk factors (IDRFs), and operative strategies. Anatomical accuracy was quantified against intraoperative findings recorded in the International Neuroblastoma Surgical Report Form. Subjective ratings were collected using Likert scales. Mixed-effects models and non-parametric statistical analyses were conducted (p<0.05). VR significantly improved overall anatomical accuracy, with a mean increase of 12.4 percentage points (95% CI 8.5-16.3; p<0.0001). Improvements were observed for vascular involvement (+12.5%, p<0.001), organ involvement (+9.6%, p<0.01), and IDRF identification (+13.7%, p<0.0001). Improvements in anatomical accuracy with VR were similar between consultants and registrars. VR did not significantly alter operative planning decisions or predicted tumour resectability. Surgeons reported greater confidence in anatomical understanding and surgical planning with VR than CT (p<0.001). The VR platform received high ratings for usability, integration into the preoperative planning workflow, and perceived clinical value. VR improved surgeons' anatomical assessment accuracy and confidence compared with traditional CT scans, highlighting VR's potential as a useful supplement to conventional imaging for preoperative evaluation in complex abdominal neuroblastoma.