Li Chen

Download Resume

Email: clatfd@gmail.com
Phone: (+1) 206-349-4629
Website: clatfd.cn | LinkedIn
No work sponsorship required

SUMMARY

Medical imaging and AI researcher with 13+ years of experience developing machine learning algorithms for medical image and temporal video analysis. Proven track record of driving the development lifecycle of commercial pulmonary ultrasound AI applications from algorithm design through product milestones and clinical validation. Expertise in ultrasound video analysis, spatiotemporal modeling, weakly and self-supervised learning, graph neural networks, and medical research.

EXPERIENCE

Ultrasound AI Group, Philips Research / Philips Ultrasound

Research Scientist · Oct. 2021–Present · Cambridge, MA · Managed by Dr. Alvin Chen

• Lung ultrasound AI for Philips Lumify (point-of-care imaging device)

  • Spearheaded the full-cycle medical imaging software development of commercial ultrasound AI features (consolidation and pleural effusion detection) on the Philips Lumify device, including concept selection, function-proven milestones, clinical validation toward commercial deployment.
  • Built a teacher-student mutual-learning framework that learns from frame-labeled, video-labeled, and/or unlabeled data, handles differences in annotation styles, and reduces estimated labeling effort for 15,000 videos from 15,000 hours to about 250 hours.
  • Contrastive self-supervised pretraining on 27,063 unlabeled videos from 528 patients raised test AUC from 0.65 to 0.88 with only 5% of labeled data (IEEE ISBI 2023).
  • In a multi-reader, multi-case (MRMC) study, AI-assisted reading met non-inferiority and superiority endpoints, improved reader-averaged AUC by 0.101/0.046 for consolidation/effusion, and reduced reading time by 16.6% (12.5 seconds).

• Graph neural networks (GNN) for ultrasound reasoning at multiple levels

  • Detection: A GNN-based pseudo-label selection method for weakly semi-supervised detection, achieving mAP@0.5 of 0.586/0.439 on breast/lung datasets with limited box labels (IEEE Transactions on Medical Imaging, conditionally accepted).
  • Video: GADA aggregates frame-level detections and spatiotemporal relationships for video classification. It achieved AUCs of 0.943/0.960 for consolidation/effusion on a six-site dataset of 9,900+ lung ultrasound videos (ICCV 2025 Workshop).
  • Patient: ZAGNet models relationships among findings across lung zones, improving patient-level AUC from 0.677/0.815 (pooling) to 0.803/0.893 for consolidation/effusion (IEEE IUS 2026).
  • Filed a series of invention disclosures and patents (12 with major contributions), strengthening Philips' intellectual property portfolio.

Vascular Imaging Lab (Radiology) & Information Processing Lab (ECE), University of Washington

Research Assistant · Sep. 2016–June 2021 · Seattle, WA · Advised by Prof. Chun Yuan (Radiology) and Prof. Jenq-Neng Hwang (ECE)

• Quantitative intracranial artery modeling and vascular feature extraction

  Developed iCafe (60,000+ lines of C++/Qt code) to trace and label intracranial arteries and extract quantitative vascular features. This work led to five first-author journal papers and seven conference publications spanning technical development, validation, and medical applications, including an Editor's Pick. iCafe is used by 34+ sites in 9+ countries for medical research.

• Automated vessel wall segmentation and quantification

  Proposed a deep-learning vessel wall segmentation algorithm in a polar coordinate system. The fully automated workflow was validated on 3.5 million popliteal artery images, equivalent to 70 years of expert manual review. Winner of the American Heart Association/Amazon Web Services Prize Competition.

Medical Imaging Lab, Fudan University

Undergraduate Researcher, Project Lead · Oct. 2013–June 2016 · Advised by Dr. Huiliang Shang and Dr. Yi Guo

• Developed algorithms for tongue image analysis and prescription data-mining (paper 1, 2, 3, 4).
• Capstone project representing vasculatures as circuits for registration (Best Fudan undergraduate thesis).

EDUCATION

• Ph.D. in Electrical Engineering, University of Washington, Seattle, WA · Sep. 2016–Aug. 2021 · GPA: 3.89/4.0
• B.S. in Electronic Engineering, Fudan University, Shanghai, China · Sep. 2012–June 2016 · GPA: 3.6/4.0 · Ranking: 9%

PUBLICATION

1. Li Chen, Shubham Patil, Rashid Al Mukaddim, et.al. ZAGNet: Zone-Aware Graph Aggregation Network for Patient-Level Lung Ultrasound Diagnosis. The 2026 IEEE International Ultrasonics Symposium, 2026, October 4–8 (Raleigh, North Carolina). (pdf)

2. Li Chen, Naveen Balaraju, Jochen Kruecker, et.al. GADA: Graph Attention-based Detection Aggregation for Ultrasound Video Classification. ICCV 2025 Workshop on Computer Vision for Automated Medical Diagnosis, 2025, Oct 19 – 23th (Hawaii Convention Center, Honolulu. Hawai'i). DOI: 10.1109/ICCVW69036.2025.00098 (pdf)

3. Jiahong Ouyang, Li Chen*, Gary Y. Li, et.al. Weakly Semi-Supervised Detection in Lung Ultrasound Videos. The 28th biennial international conference on Information Processing in Medical Imaging, 2023, San Carlos de Bariloche, Argentina (18 to 23 June 2023). DOI: 10.1007/978-3-031-34048-2_16 (pdf)

4. Li Chen*, Jonathan Rubin*†, Jiahong Ouyang, et.al. Contrastive self-supervised learning for spatio-temporal analysis of lung ultrasound videos. IEEE International Symposium on Biomedical Imaging (ISBI), 2023, Cartagena, Colombia (April 18 - 21, 2023). DOI: 10.1109/isbi53787.2023.10230816 (pdf)

5. Li Chen, Wenjin Liu, Niranjan Balu, et.al. Deep Open Snake Tracker for Vessel Tracing. MICCAI 2021, the 24th International Conference on Medical Image Computing and Computer Assisted Intervention, 2021, Strasbourg, France (September 27th to October 1st). DOI: 10.1007/978-3-030-87231-1_56 (pdf)

6. Li Chen, Thomas Hatsukami, Jenq-Neng Hwang, et.al. Automated Intracranial Artery Labeling using a Graph Neural Network and Hierarchical Refinement. MICCAI 2020, the 23rd International Conference on Medical Image Computing and Computer Assisted Intervention, 2020, Lima, Peru (October 4 - 8, 2020). DOI: 10.1007/978-3-030-59725-2_8 (pdf) (ppt) (code)

7. Li Chen, Jie Sun, Gador Canton, et.al. Automated Artery Localization and Vessel Wall Segmentation of Vessel Wall Images using Tracklet Refinement and Polar Conversion . IEEE Access, 2020, 8, Pages 217603-217614. DOI: 10.1109/ACCESS.2020.3040616 (pdf) (code)

8. Li Chen, Gador Canton, Wenjin Liu, et.al. Fully automated and Robust Analysis Technique for Popliteal Artery Vessel Wall Evaluation (FRAPPE) using Neural Network Models from Standardized Knee MRI. Magnetic resonance in medicine (IF:3.6), 2020, 84, Pages 2147–2160. DOI: 10.1002/mrm.28237 (pdf)

9. Li Chen, Mahmud Mossa‐Basha, Niranjan Balu, et.al. Development of a quantitative intracranial vascular features extraction tool on 3D MRA using semiautomated open‐curve active contour vessel tracing. Magnetic resonance in medicine (IF:3.9), 2018, 79 (6), Pages 3229-3238. DOI: 10.1002/mrm.26961 (Editor's pick) (pdf) (ppt)

For a complete list of publications, see the full publication list in my Curriculum Vitae.

TECHNICAL SKILLS

• Programming: Python, C/C++, MATLAB
• Machine learning: PyTorch, TensorFlow, Keras, self-supervised learning, weak/semi-supervised learning
• Computer vision & medical imaging: Object detection, segmentation, tracing, OpenCV, ITK, VTK
• Advanced modeling: Graph neural networks, spatiotemporal modeling, multimodal clinical AI
• Tools: Docker, Git, Linux

OTHER

• 10+ years as a reviewer for leading conferences (MICCAI) and journals (IEEE TMI) in medical imaging and AI.
• Chair for the Deep Learning section in Philips' intellectual property review group.
• Organizer of the Carotid Vessel Wall Segmentation Challenge (MICCAI/SMRA 2021).
• 9 Free Apps in Windows App Store, downloaded over 63,000 times.
• All publications and projects: clatfd.cn/cv.