2025 ASEE Annual Conference & Exposition

BOARD #107: An Efficient Approach for Muscle Segmentation and 3D Modeling Using Keypoint Tracking in MRI Scans

Presented at Computing and Information Technology Division (CIT) Poster Session

Magnetic resonance imaging (MRI) enables non-invasive, high-resolution analysis of muscle structures. However, automated segmentation remains limited by high computational costs, reliance on large training datasets, and reduced accuracy in segmenting smaller muscles. Convolutional neural network (CNN)-based methods, while powerful, often suffer from substantial computational overhead, limited generalizability, and poor interpretability across diverse populations. This study proposes a training-free segmentation approach based on keypoint tracking, which integrates keypoint selection with Lucas-Kanade optical flow. The proposed method achieves a mean Dice similarity coefficient (DSC) ranging from 0.6 to 0.7, depending on the keypoint selection strategy, performing comparably to state-of-the-art CNN-based models while substantially reducing computational demands and enhancing interpretability. This scalable framework presents a robust and explainable alternative for muscle segmentation in clinical and research applications.

Authors

Mengyuan Liu Northeastern University [biography]

Mengyuan Liu is with the Khoury College of Computer Sciences at Northeastern University. She currently serves as vice president of the Multimedia Information Group Lab at San Jose (MIG Lab@SJ). Her research interests lie in the application of computer vision and artificial intelligence techniques to medical imaging analysis.
Dr. Jeongkyu Lee http://orcid.org/0000-0003-1954-4502 Northeastern University [biography]

Dr. Jeongkyu Lee is currently a full teaching professor in the Khoury College of Computer Science at Northeastern University Silicon Valley campus. He received his Ph.D. in Computer Science from the University of Texas at Arlington in 2006. Previously, he earned a BS in Mathematics Education from Sungkyunkwan University and an MS in Computer Science from Sogang University, both located in South Korea. Before pursuing his doctorate, he worked as a database administrator for seven years with companies including Hana Bank and IBM Korea. He has demonstrated expertise in multimedia databases, medical image analysis, and video data mining. Prof. Lee's proficiency extends to machine learning, where he focuses on deep learning applications for video data and machine learning. His commitment to education is evident through his teaching of various computer science courses, ranging from data structures to advanced database systems. He founded the Multimedia Information Group Lab at San Jose (MIG Lab@SJ), which drives research in medical image analysis, SQL auto-grading, and enhancement of prediction models for ServoSphere projects.

Note

The full paper will be available to logged in and registered conference attendees once the conference starts on June 22, 2025, and to all visitors after the conference ends on June 25, 2025

« View session