Related Experiment Video
Updated: Nov 7, 2025
04:23
A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images
Published on: April 21, 2023
2.1K
Reverse Scan Conversion and Efficient Deep Learning Network Architecture for Ultrasound Imaging on a Mobile Device
Kunkyu Lee1, Min Kim2, Changhyun Lim2
1Department of Electronic Engineering, Sogang University, Seoul 04107, Korea.
Sensors (Basel, Switzerland)
|April 30, 2021
Summary
This study introduces a novel system for point-of-care ultrasound (POCUS) that combines real-time imaging with AI-driven guidance on mobile devices. A new dataset creation method enhances diagnostic accuracy for inexperienced users in emergency settings.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Ultrasound Technology
Background:
- Point-of-care ultrasound (POCUS) is crucial in emergencies but faces challenges with untrained users.
- Cloud-based AI systems for POCUS are limited by security, network, and energy concerns.
Purpose of the Study:
- To develop an integrated POCUS system for simultaneous imaging and AI-guided diagnosis on mobile devices.
- To improve the accuracy of deep learning models for POCUS by creating a specialized training dataset.
Main Methods:
- Proposed a novel structure for integrated ultrasound imaging and mobile device-based AI guidance.
- Introduced a reverse scan conversion (RSC) method for generating an ultrasound training dataset.
Main Results:
- The integrated system achieved simultaneous ultrasound imaging and deep learning at up to 42.9 frames per second.
- The RSC method improved image classification accuracy by over 3%.
Conclusions:
- The proposed integrated POCUS system effectively addresses limitations of cloud-based AI.
- The RSC method enhances deep learning model performance, improving diagnostic support for POCUS users.
