Responsibilities*
The Research Investigator will contribute to ongoing research efforts in biomedical data sciences, focusing on the development of computer-supported decision systems for clinical use. Responsibilities include designing and implementing new methods in data processing and analysis, such as signal and image processing, machine learning, and advanced mathematical techniques (e.g., LLM, ML, digital twins). The investigator will create scalable algorithms for analyzing large-scale data, emphasizing the use of tensor-based approaches to improve accuracy in medical/dental decision-making.
Work will consist of:
Research and Development (50%): Conducting research in data analysis, developing and implementing algorithms, and creating innovative solutions for medical applications.
Data Integration and Analysis (20%): Integrating various types of clinical and biomedical data, ensuring robust analysis across multiple data sources using advanced techniques.
Collaboration and Project Management (10%): Working with multidisciplinary teams, participating in project administration, and managing research initiatives.
Mentoring (10%): Supporting undergraduate and graduate students in their research activities.
Publication and Grant Writing (10%): Preparing and submitting peer-reviewed publications; writing grant proposals to seek funding for ongoing projects.
AI/Deep Learning
Hands-on experience in PyTorch for deep learning model development.
Practical expertise in CT image segmentation and classification, especially using Unet architectures.
Experience in deep reinforcement learning methods, including deep Q-learning, decision transformers, and decision mamba.
Familiarity with multimodal transformers, large language models, and vision-language models such as CLIP.
Use of HuggingFace ecosystem for model deployment and research.
Solid understanding and application of tensor methods in deep learning, generative models, graphical models, and causality.
Machine Learning
Proficiency with scikit-learn for classical machine learning tasks.
Background in data preprocessing and feature engineering.
Experience with random forests, clustering, boosting, ensemble learning methods, and interpretable machine learning approaches.
Application of tensor methods and signal processing techniques to biomedical or image data.
Programming & Computational Skills
Strong programming skills in Python are essential.
Demonstrated use of GitHub for collaborative software development and version control.
Experience working with High Performance Computing (HPC) clusters and Linux environments.
Familiarity with database tools such as SQLite; experience with Matlab and R for data analysis.
Communication, Scientific Writing, & Leadership
Proven ability to present research findings at scientific conferences, industry meetings, and to cross-functional teams.
Experience writing and editing scientific manuscripts, grant proposals, and undertaking peer reviews.
Ability to work effectively in multidisciplinary teams comprising scientists, engineers, and clinicians.
Demonstrated leadership, including mentoring and training junior scientists and students.
Preferred Qualifications:
Prior experience in medical imaging or healthcare AI applications.
Published work in leading journals or conferences.
Track record of successful grant funding or contributions to large-scale projects.
Positions that are eligible for hybrid or mobile/remote work mode are at the discretion of the hiring department. Work agreements are reviewed annually at a minimum and are subject to change at any time, and for any reason, throughout the course of employment. Learn more about the work modes.
U-M EEO StatementThe University of Michigan is an equal employment opportunity employer.