• Statistical support and data analysis of preclinical and translational imaging studies in discovery and early drug development. Designed, analyzed and reported preclinical and early clinical studies with functional imaging component• Contribution to machine learning efforts at Merck: feature identification / screening in learn and confirm paradigm for biomarker identification cancer staging, heart failure classification from real world claims data, prediction in clinical safety, unsupervised learning/visualization for high dimensional problems with categorical and continuous data with applicability to clinical trials, real world evidence and CMC, assessed data complexity in classification and developed algorithms for prediction of heart failure and neurodegeneration in animal models, organized Data Science Symposium and Data Science competition at Merck • Developed clustering and visualization for high-dimensional imaging data in proof-of-concept imaging studies• Performed repeatability and reproducibility analysis and longitudinal data analysis in imaging based on linear mixed effects model. • Carried out simulations to estimate variability of test-retest outcome measures for quantitative imaging endpoints• Developed processing pipelines for preclinical and clinical studies with functional Magnetic Resonance Imaging (fMRI)• Quantified longitudinal change in Positron Emission Tomography (PET) receptor occupancy studies and beta amyloid tracer studies• Mentored new hires, summer interns and contractors working on image processing and statistical analysis of imaging data• Effectively communicated the results of the research in internal presentations, scientific meetings and publications

• Quantified low dimensional endpoints from stimulus driven and resting state functional MR imaging and spectroscopic data in temporal and frequency domain using model as well as data driven methods such as clustering and Independent Component Analysis • Developed clustering and visualization for high-dimensional imaging data in proof-of-concept imaging studies• Performed repeatability and reproducibility analysis and longitudinal data analysis in imaging based on linear mixed effects model• Investigated variability of test-retest outcome measures for quantitative imaging endpoints• Developed processing pipelines for functional Magnetic Resonance Imaging (fMRI)• Mentored new hires, biometricians, summer interns and contractors working on image processing and statistical analysis of imaging data• Effectively communicated the results of the research in internal presentations, scientific meetings and publications

• Carried out research in machine learning (classification and regression) for biomarker discovery from high-dimensional biomedical data in metabolomics and chemometrics (MR spectroscopy), genomics (gene microarrays) and proteomics (mass spectroscopy)• Developed dimensionality reduction, visualization, feature selection/extraction and data complexity assessment techniques for prediction of different species of pathogenic fungi from high-dimensional spectroscopic data• Co-developed methods for classification error estimation and selection bias in under-sampled classification problems, mostly for spectroscopic data• Developed multiple classifiers systems for classification of spectroscopic and gene expression data from neurodegenerative diseases and oncology• Co-developed improved methods for prediction of osteoporosis from administrative data• Supervised students and served as a lecturer for several undergraduate and graduate courses

• Developed methods for exploratory analysis of functional MRI data• Implemented modifications to fuzzy k-means clustering. Developed minimum spanning tree ranking for visualization and exploration of fMRI time-series. Carried out comparisons between Principal Component Analysis (PCA), clustering and general linear model in fMRI• Co-developed signal processing methods for fMRI based on wavelets

Taught intro math courses and supervised studentsCourses taught: Linear Algebra I, Linear Algebra II, Introduction to Calculus

Taught courses and supervised master studentsCourses taught: Topics in applied linear algebra and optimization

• Developed methods for exploratory data analysis (EDA) for fMRI based on PCA / factor analysis and cluster analysis. Pioneered application of EDA in fMRI

• Carried out research in image segmentation of MR images• Developed a software package for semi-automated segmentation of cardiac MRI