Fall 2022:I worked with Dr. Joshua Yung, Dr. Moiz Ahmad, and Dr. Christopher Walker to utilize aseries of 3T contrast-enhanced 3D T1 Gradient Echo (GRE) images and 3D Turbo Spin Echo (3DTSE)brain images evaluated by neuro-radiologists for image quality and image artifacts along withcomputed MRQy quantitative metrics to develop a decision tree classifier (with bagging ensemble) toclassify images based on their quantitative metrics in comparison to the radiologist evaluations. Thiswork was accepted as a Digital Poster at the International Society for Magnetic Resonance inMedicine (ISMRM) 2023 annual meeting.Spring 2023:I worked with Dr. Joshua Niedzielski applying multi-parametric MRI (mpMRI) usingIVIM and DCE to predict treatment response in Pancreatic cancer. Apart from the quantitativeparameters resulting from the IVIM and DCE analysis, I also looked at radiomics features as aframework for future machine learning applications. The mpMRI work was accepted as an OralPresentation at the 2023 ASTRO Annual Meeting and the radiomics work was accepted as anePoster at the 2023 AAPM Annual Meeting.Summer 2023:I worked with Dr. Mark Pagel on multiple projects including developingMagnetic Resonance Fingerprinting (MRF) and multi-slice MRI methods for Contrast EnhancedSaturation Transfer (CEST) MRI on a 7T Bruker system along with applying machine learningapproaches to predict pH in-vivo using CEST MRI.

- Worked with University of Florida to implement organ vasculature models for blood dynamics simulations- Maintained computer code on combining flow dynamics with three-dimensional organ vasculature models and dose delivery in Python- Expanded the code to include simulations of other organs such as the brain, lung, and heart- Validated the results using treatment outcome data- Incorporated tetrahedral mesh representations of vasculature models into a Monte Carlo framework- Utilized the model to predict lymphocyte (ALC) depletion at various fractionation and dose rate combinations

Laboratory for Computational Physiology (Dr. Roger G. Mark):Supported and developed open-source software for analyzing, processing, and sharing multi-modal clinical data including structured data, waveforms, and images.PhysioNet (https://github.com/MIT-LCP/physionet-build):- Developed and maintained clinical data web platform using Python (Django), JavaScript, HTML, and CSS- Developed API to allow creating, saving, and editing annotations of waveforms using GraphQL- Developed interactive and customizable waveform visualizations using Plotly for JavaScriptWFDB-Python (https://github.com/MIT-LCP/wfdb-python):- Translated the original project written in C and developed Python software to analyze waveforms in MAT, WAV, EDF, and WFDB format using Numpy, Pandas, and Struct.pack() / Struct.unpack()- Provided interactive and customizable waveform visualization using Matplotlib- Created a package demo using Jupyter Notebooks for PythonServers:- Managed daily server activities, usage, and updates using Nagios and NRPE on different operating systems including Fedora, CentOS, Debian, and Windows- Updated packages and techniques included CERTBOT for SSL certificates, RSYNC and RSNAPSHOT for backups, NTP for time synchronization, and NGINX for sever configuration and logs- Developed password authentication for Wikimedia page using LDAP

Assisted in a research project mentored by Dr. Radi Masri with the goal of determining the signals and responses of neurons to various levels of stimuli. In this project, I was assigned to use Matlab to analyze videos and determine the neuronal signals and signatures as well as determine various neuro-mathematical processes. This project involved the heavy use of large datasets due to the high spatial and temporal resolution of the video (.avi) files. In parallel with this research project, we will be submitting a grant proposal to continue the project throughout the year of 2018 as well as developing and publishing a research paper to document our results.

Analyzed MRI imaging data using FreeSurfer to reconstruct biologically and topologically correct segments of the human brain. Implemented advanced statistical analysis (e.g. ANOVA, regression, etc.) and machine learning techniques (e.g. TensorFlow, Keras, etc.) to identify indicators for Alzheimer’s disease, Parkinson’s disease, and a control group.

Redesigning and building a high-level Python wrapper in FORTRAN to drive the automated testing and provide a high-level user interface for the OpenFAST project. The Python wrapper involved user interface design through an API and general object-oriented programming along with data modeling and text parsing. Additionally, lower level classes and functions were developed which are used in an abstract manner throughout the software.

Assisted in the development of assays to reduce the quantity of primer-dimers to increase the efficiency of wet testing. The results of this work will be used to tackle bio-environmental problems including water sanitation, structural damage control, and food quality assurance. I implemented C++ and Python to computational derive the optimal solutions to primer development and optimization. Future work for this project will include developing a platform for users to customize their preferences, settings, and desired output parameters to increase the utility of primer design.

Participated in the research and development of a transducing device utilizing lead zirconate titanate (PZT) plates and the piezoelectric effect in order to detect tumors under simulated conditions.

Official note taker for UMBC for the course Probability and Statistics for Scientists and Engineers and assisted in the delivery of notes to students with disabilities.

Assisted in the design and development of a cam device used to simulate the motion of the connection between the femur and tibia in order to generate electricity from an electromagnetic device inserted into a knee transplant.

Assisted in developing software for a website displaying live data extracted from windturbines. Implemented Python, JavaScript, HTML, and CSS for data visualization, analysis, andwebsite development. Utilized Anaconda, Sphinx, GitHub, Read the Docs, and reStructuredTextto provide accurate and precise levels of documentation. Results and findings will be published in a 2018 national Department of Energy (DOE) report and showcased in a live demonstration.

Advocated renewable technology through designing a website associated with the Maryland Department of Transportation (MDOT) and collaborated with Exelon, Baltimore City Hall, Under Armour, etc. on electric vehicle charging station implementation.