BME Research Seminar: Ultrasound Foundation Modelling and Its Clinical Applications

BME Research Seminar: Ultrasound Foundation Modelling and Its Clinical Applications

Speaker: Prof. Yi Guo, Professor, School of Science and Technology, Fudan University

Overview of the Lecture

Medical ultrasound is widely used in clinical diagnosis and treatment due to its advantages of non-invasive and non-radiation and real-time imaging. Intelligence analysis and diagnosis of ultrasound images is an important part of intelligent medical research. Inadequate generality across different organs and tasks constrains the application of ultrasound (US) image analysis methods in smart healthcare. Building a universal US foundation model holds the potential to address these issues. Nevertheless, the development of such foundational models encounters intrinsic challenges in US analysis, i.e., insufficient databases, low quality, and ineffective features. In this talk, we will present a universal US foundation model, named USFM, generalized to diverse tasks and organs towards label efficient US image analysis. First, a largescale Multi-organ, Multi-center, and Multi-device US database were built, comprehensively contains over two million US images. Data unification is achieved through a new method of high-resolution reconstruction and domain generalization. Then a self-supervised approach is used to build an ultrasound base model, which can extract ultrasound image features efficiently and accurately. The pre-trained USFM can be plug-and-play implemented with existing frameworks. In the clinical application method, a human-in-loop interaction is used to quickly apply the foundation model to multiple ultrasound downstream tasks, achieving better performance than sota methods.

About the Speaker

Prof. Guo is a Professor in the School of Science and Technology at Fudan University. She was selected in the National High-level Talent Youth Program. She is engaged in the research of ultrasound imaging, medical image analysis and intelligent ultrasound equipment. She has hosted the National Natural Science funds and Shanghai Science and Technology Commission Young Scientific Talents Yangfan Program. She has published over 80 publications in leading international journals and conferences. She has 14 issued patents and software copyright with over 20 patents pending in the areas of ultrasound imaging, medical image analysis and computer-aided diagnosis. She won the Shanghai Natural Science Award, the Gold Medal of Excellent Invention Selection Competition, and the Excellent Product Award of Hi-Tech Fair. She also served as the challenge chair for ultrasound image enhancement held in conjunction with MICCAI.