Praveen Dassanayake
Data Scientist/ MITACS Research Fellow
Based in London, Ontario, Canada
About
Data Scientist and Research Analyst with 6+ years of experience leveraging advanced analytics, statistical modeling, and data storytelling to deliver high-impact solutions across healthcare, imaging, and regulatory science. My background spans experimental design, data wrangling, and predictive modeling, supported by a PhD in Applied Neuro Science Analytics and formal training in Management Accountancy. Proficient in SQL, Excel, Power BI, and Python (Pandas, NumPy, scikit-learn, matplotlib, seaborn) for building automated ETL pipelines, data visualizations, and dashboard reporting aligned with product and financial KPIs. Developed a Python-based deep learning pipeline using nnU-Net for automated pig organ segmentation on multi-contrast MRI (T1/T2), applying image preprocessing.
Skills
Projects
Experience
2024-2027
Data Scientist/ MITACS Research Fellow @ Lawson Research Institute
Developed a Python-based deep learning segmentation model using nnU-Net for multi-contrast MRI datasets, building scalable data preprocessing and training pipelines. Evaluated performance using Dice and Jaccard metrics to ensure model accuracy and reproducibility. Designed an automated PET–MRI analytics pipeline for image registration, normalization, and similarity analysis, enabling quantitative modeling, dosimetry calculations, and data-driven decision support through structured reporting and cross-functional collaboration.
2019-2023
Doctoral Researcher (Applied Analytics)
Collaborated with Siemens Healthineers to develop a scalable PET/MRI analytics framework (caliPER) enabling automated tracer quantification using Patlak modeling. Built Python-based image processing pipelines for segmentation, registration, and partial volume correction, integrating backend logic into syngo.Via and MeVisLab environments for high-throughput analysis. Developed statistical validation tools to benchmark image-derived input functions using regression analysis, ICC, Pearson correlation, and Bland–Altman methods. Implemented voxel-level kinetic modeling and cohort-level hypothesis testing (ANOVA, t-tests, effect sizes) to support reproducible clinical insights. Led nonlinear optimization and simulation workflows in MATLAB for kinetic parameter estimation.
2017-2019
Master's Researcher (Biostatistics)
Led the design and execution of statistical analysis plans (SAPs) for MRI-based biomarker studies during my Master’s research, applying one-way and two-way ANOVA, Pearson correlation, and linear regression in RStudio to evaluate treatment effects on iron uptake, protein expression, and MR relaxation rates. Translated experimental hypotheses into structured analytical models, ensuring data integrity, statistical rigor, and reproducible reporting. Co-authored peer-reviewed research by driving quantitative analysis of MRI-derived biomarkers and ICP-MS trace element data, integrating regression modeling and hypothesis testing (t-tests, ANOVA) to validate imaging outcomes.
Education
2019-2023
PhD in Medical Biophysics
Development and validation of less invasive quantification methods for imaging neuroinflammation in frontotemporal dementia patients using PET reversible and irreversible tracers. Western University, London, Ontario, Canada.
2017-2019
Masters in Medical Biophysics and Molecular Imaging
Thesis: Monocyte MRI relaxation rates are regulated by extracellular iron and hepcidin Western University, London, Ontario, Canada.
2011-2014
Advanced Diploma in Management Accountancy
Charted Institute of Management Accountants London, EC2, United Kingdom
2013-2016
Bachelor's Degree in Physics
4th Year Thesis: Development of an electronic stethoscope to monitor abnormalities in the heart rhythm of patients with cardiac issues. University of Sri Jayewardenepura, Colombo, Sri Lanka