My name is Marlene Lin, currently a neuroimaging data analyst at the Rabinovici Lab at UCSF Memory & Aging Center, supporting research on neurodegenerative diseases. I work with Python, R, and MATLAB for image processing, data wrangling, statistical modeling, and machine learning—focusing on understanding disease heterogeneity and its implications for clinical care.
AI-Guided Surgical Blood Transfusion Need Prediction
Read More
Data Sciences Help Desk Dashboard
Read More
Breast Lesion Classification with MRS
Read More
About Me
I have broad experience analyzing data across domains and thrive in diverse analytical environments. As an operations intern at Kaisa Group, I evaluated complex population and health data to support a key study on telemedicine resource distribution. At UCLA and UCSF, I contributed to projects that derived diagnostic and clinical decision-making insights from imaging or EHR data using rigorous statistical analysis and machine learning models.
Beyond technical proficiency, I further built my problem-solving and collaboration skills while working in these cross-functional teams and as a teaching assistant with the UCSF Library’s data science team. I strive to foster effective working environments through development and documentation of pipelines and tools that optimize workflows. I believe my proactive, detail-oriented approach would allow me to contribute meaningfully to the team.
Read More
Projects
Research
Tau-PET subtypes in Early-onset Alzheimer’s Disease
CAPSTONE project with Rabinovici Lab, UCSF Memory and Aging Center
Identify Alzheimer’s Disease subtypes based on the topographic distribution of tau by applying robust data-driven clustering methods on baseline tau-PET of sporadic early-onset patients from the Longitudinal Early-Onset Alzheimer’s Disease Study
AI-Guided Surgical Blood Transfusion Need Prediction
June 2024 with Dr. Priya Ramaswamy, UCSF Department of Anesthesia
Enhanced a gradient boosting model to predict surgical blood transfusion needs and benchmarked model against the Maximum Surgical Blood Order Schedule and clinician order performance. Supported silent prospective validation and model deployment.
Prevalence of Diabetes Screening, Pre-diabetes Testing, and Nutrition Counseling
March 2024 with Dr. Yoshimi Fukuoka, UCSF School of Nursing
Characterize the prevalence of encountered for diabetes screening, testing for pre-diabetes, and nutrition counseling among different demographic groups using electronic health record data queried from the UCSF Info Commons
Re-implemented a deep-learning model for bone age regression with an attention-guided localization network and label distribution learning-based regression network. Incorporated ethnicity information from the Digital Hand Atlas data to improve the accuracy of bone age prediction.
Thesis: Ensemble Learning for Breast Cancer Lesion Classification: A Pilot Validation Using Correlated Spectroscopic Imaging and Diffusion-Weighted Imaging. Metabolites. 2023; 13(7):835. https://doi.org/10.3390/metabo13070835
Relevant Coursework:
Biotechnology and Society
Digital Image Processing
System and Signal
Statistical Programming
Mathematical Statistics
Linear Models
Data Analysis & Regressions
Experimental Design
University of California, Los Angeles (2019 - 2023)
Bachelor of Science in Computational and System Biology
Thesis: Ensemble Learning for Breast Cancer Lesion Classification: A Pilot Validation Using Correlated Spectroscopic Imaging and Diffusion-Weighted Imaging. Metabolites. 2023; 13(7):835. https://doi.org/10.3390/metabo13070835
Co-teach and assist with R and Python programming workshops on topics including data visualization, statistical analysis, and machine learning.
Offer 1:1 programming and data analysis help to UCSF community members during weekly data science help desk.
Assist with other projects including updating online course materials, creating subject guides and help articles, and presenting library resources to UCSF members.
Enhanced a gradient boosting model for blood transfusion predictions through feature engineering, model optimization, and threshold tuning, achieving a marked improvement in precision.
Benchmarked model performance against the Maximum Surgical Blood Order Schedule and clinician request patterns using both retrospective data and silent prospective validation, providing actionable insights for model deployment to support pre-operative clinical decision-making.
Managed data visualization and report preparation for the Learning Health System meetings, collaborating with data scientists, clinicians, and health IT stakeholders to align on project goals.
Developed MATLAB applications for in-depth processing and annotations of multidimensional Magnetic Resonance Spectra data.
Tuned parameters and assessed the performance of various MR spectroscopy reconstruction methods, reducing MR spectroscopy scan time by at least two-fold.
Characterized brain MR Spectroscopy data of obstructive sleep apnea and pediatric AIDS patients.
Construct ensemble learning models to classify breast tumors based on 5D MR Spectroscopy quantitation, results published in Ajin et. al, Metabolites. 2023; 13(7):835.
Encouraged growth-oriented mindsets among students and created an equitable STEM learning environment by fostering student collaboration as a peer educator.
Effectively explained course concepts and expanded peers’ understanding of the relevant materials in discussions, course workshops, and review sessions.
Formulated ways to improve course structure and helped maintain clarity within the course by consistently communicating with instructors in weekly meetings.
Directed detailed regional analysis of the healthcare sector in major cities of China to provide insights on the department’s Direct-to-Patient service deployment.
Collected and analyzed data on medical resources distribution, international and domestic health product sales, and health expenditures of different communities.
Conducted on-site investigations of various pharmacy chains and delivered relevant reports to support the department’s business acquisition plan.
Trivia
Thank you for scrolling all the way through :)
Header: Sayram Hu, Xinjiang, 2008-08
Fun fact: my friend got me a “already thinking about my next coffee” cup from Typo but I reserved it exclusively for tea drinking! (recipe: brown sugar syrup line the cup, hojicha, a spalsh of sikhye, and a sprinkle of salt. Let it sit for 10 minutes and enjoy ☕🧘)
