Super-resolution ultrasound localization microscopy (ULM) is an acoustic analog to optical super-resolution microscopy such as photoactivated localization microscopy (PALM) or stochastic optical reconstruction microscopy (STORM). It was recently introduced for microvasculature imaging beyond the inherent acoustic spatial resolution limit while conserving the imaging depth of conventional ultrasound. The primary idea of ULM is to localize microbubbles (MBs) flowing in the vascular networks to achieve super-resolution, and then track the localized MBs over time to measure blood flow velocity. ULM improves the conventional ultrasound spatial resolution by approximately tenfold and it showed promising results in various tissues including brain, kidney, liver, and tumor.
1. Capsule Endoscopy System Development with IMU Sensors
Our research focuses on the development of a state-of-the-art capsule endoscopy system, utilizing IMU sensors. This innovative approach is designed to overcome the limitations of traditional flexible endoscopies, such as discomfort and limited diagnostic range. By incorporating IMU sensors, our capsule endoscopy system offers a more comfortable and comprehensive diagnostic tool for patients, enabling detailed exploration of the gastrointestinal tract.
2. Endoscopic Position Localization Using IMU Sensors
A critical aspect of our research involves the precise localization of the capsule endoscope using IMU sensors. Although these sensors are a powerful tool, they face challenges in accurate position estimation due to various error factors. Our work addresses these challenges by leveraging advanced deep learning techniques, specifically the Transformer model, to enhance the accuracy of the IMU sensor-based navigation system.
Innovative Approach Using Deep Learning
This study introduces an innovative deep learning model-based approach for keypoint detection on X-ray and ultrasound images of Developmental Dysplasia of the Hip (DDH). It delves into the analysis of the performance of various deep learning models for this purpose. A significant aspect of this research is the introduction and evaluation of diverse data augmentation techniques to address the scarcity of medical data in this field.
The application of deep learning models to ultrasound and X-ray images for hip joint keypoint detection offers a promising advancement in the diagnostic process of DDH. By improving both the objectivity and productivity in the diagnosis, this method holds the potential to significantly benefit early and accurate detection of DDH, ultimately contributing to better patient outcomes.
1. Development of a Board with High-Voltage Pulser and ADC Chips
Our research team is dedicated to developing an advanced board that integrates high-performance high-voltage pulser and analog-to-digital converter (ADC) chips. This initiative is designed to significantly improve the efficiency and accuracy of ultrasound imaging systems. The board is engineered to effectively transfer high-voltage pulses and seamlessly convert the received analog signals into a digital format. Such capabilities are essential for the precise acquisition and processing of ultrasound data, thereby enhancing the overall quality and reliability of ultrasound imaging."
2. Board Control and Data Transmission with Zybo Z7-20 (FPGA)
We are advancing a board control system based on the Zybo Z7-20 Field-Programmable Gate Array (FPGA). This system facilitates real-time data processing and management. Supporting high-speed data transmission, it ensures rapid transfer and processing of ultrasound imaging data. Leveraging the flexibility and scalability of the FPGA, we are providing a robust solution adaptable to various applications.
3. Development of a PC Visualization Tool
To effectively analyze and interpret ultrasound imaging data, we are developing a user-friendly PC visualization tool. This tool presents complex data in a visually comprehensible manner, aiding researchers and medical professionals in more accurate diagnosis and analysis.