Enhancing Annotation Efficiency with Machine Learning: Automated Partitioning of a Lung Ultrasound Dataset by View

VanBerlo, Bennett; Smith, Delaney; Tschirhart, Jared; VanBerlo, Blake; Wu, Daniel; Ford, Alex; McCauley, Joseph; Wu, Benjamin; Chaudhary, Rushil; Dave, Chintan; Ho, Jordan; Deglint, Jason; Li, Brian; Arntfield, Robert

doi:10.3390/diagnostics12102351

Cited by 5 publications

(4 citation statements)

References 35 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Before experimentation, we split the labeled portion of ParenchymalLUS by anonymous patient identifier into training, validation, and test sets. The unlabeled portion of ParenchymalLUS was assembled by gathering 20000 videos from the unlabeled pool of videos in the database that were predicted to contain a parenchymal view of the lungs by a previously trained lung US view classifier (VanBerlo et al, 2022a). All videos from the same patient were in either the labeled or the unlabeled subset.…”

Section: A-line Vs B-line Classification (Ab)mentioning

confidence: 99%

Intra-video positive pairs in self-supervised learning for ultrasound

VanBerlo,

Wong,

Hoey

et al. 2024

Front. Imaging

Self Cite

View full text Add to dashboard Cite

IntroductionSelf-supervised learning (SSL) is a strategy for addressing the paucity of labelled data in medical imaging by learning representations from unlabelled images. Contrastive and non-contrastive SSL methods produce learned representations that are similar for pairs of related images. Such pairs are commonly constructed by randomly distorting the same image twice. The videographic nature of ultrasound offers flexibility for defining the similarity relationship between pairs of images.MethodsWe investigated the effect of utilizing proximal, distinct images from the same B-mode ultrasound video as pairs for SSL. Additionally, we introduced a sample weighting scheme that increases the weight of closer image pairs and demonstrated how it can be integrated into SSL objectives.ResultsNamed Intra-Video Positive Pairs (IVPP), the method surpassed previous ultrasound-specific contrastive learning methods' average test accuracy on COVID-19 classification with the POCUS dataset by ≥ 1.3%. Detailed investigations of IVPP's hyperparameters revealed that some combinations of IVPP hyperparameters can lead to improved or worsened performance, depending on the downstream task.DiscussionGuidelines for practitioners were synthesized based on the results, such as the merit of IVPP with task-specific hyperparameters, and the improved performance of contrastive methods for ultrasound compared to non-contrastive counterparts.

show abstract

Section: A-line Vs B-line Classification (Ab)mentioning

confidence: 99%

Intra-video positive pairs in self-supervised learning for ultrasound

VanBerlo,

Wong,

Hoey

et al. 2024

Front. Imaging

Self Cite

View full text Add to dashboard Cite

show abstract

“…Access to the data was permitted via ethics approval granted by Western University (REB 116838). The dataset had been previously labeled for the View, AB, and PE tasks by competent LUS interpreters during prior work [11,36]. The labeled portion of the local dataset was split by patient identifier into a training set (70%), validation set (15%), and test set (15%), and the external dataset was reserved for testing only.…”

Section: B Datamentioning

confidence: 99%

Self-Supervised Pretraining Improves Performance and Inference Efficiency in Multiple Lung Ultrasound Interpretation Tasks

Vanberlo,

Li,

Hoey

et al. 2023

IEEE Access

Self Cite

View full text Add to dashboard Cite

In this study, we investigated whether self-supervised pretraining could produce a neural network feature extractor applicable to multiple classification tasks in B-mode lung ultrasound analysis. When fine-tuning on three lung ultrasound tasks, pretrained models resulted in an improvement of the average across-task area under the receiver operating characteristic curve (AUC) by 0.032 and 0.061 on local and external test sets respectively. Compact nonlinear classifiers trained on features outputted by a single pretrained model did not improve performance across all tasks; however, they reduced inference time by 49% compared to the serial execution of separate fine-tuned models. When training using 1% of the available labels, pretrained models consistently outperformed fully supervised models, with a maximum observed test AUC increase of 0.396 for the task of view classification. Overall, the results indicate that self-supervised pretraining is a useful strategy for producing initial weights for lung ultrasound classifiers.INDEX TERMS Multi-task, self-supervised learning, ultrasound

show abstract

“…In the past few years, the analysis of medical images based on ML has gained significant importance in scientific research. In particular, with the progress of computer vision, researchers are encouraged to develop various systems for the analysis, correlation, and interpretation of medical images [ 15 , 16 , 17 ] such as convolutional neural networks for brain image segmentation [ 18 , 19 , 20 , 21 ]; for brain tumor detection and classification [ 22 , 23 , 24 ]; medical image registration, fusion, and annotation [ 25 , 26 , 27 , 28 , 29 ]; computer-aided diagnosis (CAD) systems [ 30 , 31 , 32 , 33 , 34 , 35 ]; and the automatic detection of micro-bleeds in a medical image [ 36 , 37 , 38 ].…”

Section: Introductionmentioning

confidence: 99%

Smart Visualization of Medical Images as a Tool in the Function of Education in Neuroradiology

Simović¹,

Lutovac-Banduka

Lekić

et al. 2022

Diagnostics

View full text Add to dashboard Cite

The smart visualization of medical images (SVMI) model is based on multi-detector computed tomography (MDCT) data sets and can provide a clearer view of changes in the brain, such as tumors (expansive changes), bleeding, and ischemia on native imaging (i.e., a non-contrast MDCT scan). The new SVMI method provides a more precise representation of the brain image by hiding pixels that are not carrying information and rescaling and coloring the range of pixels essential for detecting and visualizing the disease. In addition, SVMI can be used to avoid the additional exposure of patients to ionizing radiation, which can lead to the occurrence of allergic reactions due to the contrast media administration. Results of the SVMI model were compared with the final diagnosis of the disease after additional diagnostics and confirmation by neuroradiologists, who are highly trained physicians with many years of experience. The application of the realized and presented SVMI model can optimize the engagement of material, medical, and human resources and has the potential for general application in medical training, education, and clinical research.

show abstract

Enhancing Annotation Efficiency with Machine Learning: Automated Partitioning of a Lung Ultrasound Dataset by View

Cited by 5 publications

References 35 publications

Intra-video positive pairs in self-supervised learning for ultrasound

Intra-video positive pairs in self-supervised learning for ultrasound

Self-Supervised Pretraining Improves Performance and Inference Efficiency in Multiple Lung Ultrasound Interpretation Tasks

Smart Visualization of Medical Images as a Tool in the Function of Education in Neuroradiology

Contact Info

Product

Resources

About