Deep-learning-assisted and GPU-accelerated vector Doppler imaging with aliasing-resistant velocity estimation
Bibliographic record
Abstract
Vector flow imaging is a diagnostic ultrasound modality that is suited for the visualization of complex blood flow dynamics. One popular way of realizing vector flow imaging at high frame rates over 1000 fps is to apply multi-angle vector Doppler estimation principles in conjunction with plane wave pulse-echo sensing. However, this approach is susceptible to flow vector estimation errors attributed to Doppler aliasing, which is prone to arise when a low pulse repetition frequency (PRF) is inevitably used due to the need for finer velocity resolution or because of hardware constraints. Existing dealiasing solutions tailored for vector Doppler may have high computational demand that makes them unfeasible for practical applications. In this paper, we present the use of deep learning and graphical processing unit (GPU) computing principles to devise a fast vector Doppler estimation framework that is resilient against aliasing artifacts. Our new framework works by using a convolutional neural network (CNN) to detect aliased regions in vector Doppler images and subsequently applying an aliasing correction algorithm only at these affected regions. The framework's CNN was trained using 15,000 in vivo vector Doppler frames acquired from the femoral and carotid arteries, including healthy and diseased conditions. Results show that our framework can perform aliasing segmentation with an average precision of 90 % and can render aliasing-free vector flow maps with real-time processing throughputs (25-100 fps). Overall, our new framework can improve the visualization quality of vector Doppler imaging in real-time.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.000 | 0.000 |
| Bibliometrics | 0.000 | 0.000 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.000 | 0.001 |
| Open science | 0.001 | 0.001 |
| Research integrity | 0.000 | 0.001 |
| Insufficient payload (model declined to judge) | 0.001 | 0.000 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".