MBFormer: A Transformer model for 3D Time-Series Data Processing to Improve Bound Bubble Detection in Nondestructive Ultrasound Molecular Imaging
Bibliographic record
Abstract
Ultrasound molecular imaging (USMI) is an approach that detects targeted microbubbles (MBs) that are bound to specific biomarkers of disease; however, effectively suppressing free-floating MBs remains challenging. In this study, we adopted 3D transformer to analyze time-series ultrasound dataset to achieve enhanced differentiation of bound MBs from free-floating counterparts.We present MBFormer, or microbubble transformer, a nondestructive USMI transformer model that processes time-series 3D spatio-temporal ultrasound videos. The architecture of our model consists of a hierarchical transformer framework that comprises (1) an encoder with the attention mechanism but without positional embeddings, (2) a lightweight decoder featuring multilayer perceptrons (MLPs), and (3) a probability head. The model was validated through an in vivo study employing a transgenic mouse model of spontaneous breast cancer. B-mode and contrast enhanced ultrasound (CEUS) videos were captured utilizing the Vevo2100 system (FUJIFILM Visualsonics, Toronto, Canada). The ground truth data was generated via robust principal component analysis (RPCA)-processed differential targeted enhancement (DTE).We assessed the performance of MBFormer against the baseline models: a CNN-based USMI approach and SegFormer3D, an established 3D transformer model for medical image segmentation. MBFormer demonstrated improved detectability of bound MBs, achieving an area under the curve (AUC) of 0.951, a correlation coefficient (CC) of 0.981, and a continuous dice coefficient (CDC) of 0.748, surpassing the baseline models. These findings suggest that our 3D time-series data processing using the proposed transformer architecture effectively suppresses free-floating MBs while enhancing the detectability of bound MBs.
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 distilled prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. Learned from the 10,348 direct Codex labels and 10,348 direct Gemma labels. Candidate is the union of thresholded teacher heads; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels or direct frontier model labels.
Codex and Gemma teacher scores by category
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.001 | 0.000 |
| Meta-epidemiology (narrow) | 0.001 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.000 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.002 |
| Open science | 0.001 | 0.000 |
| Research integrity | 0.000 | 0.000 |
| Insufficient payload (model declined to judge) | 0.000 | 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 teacher head, 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".