Apart from routine residency rotations that are listed in my Curriculum Vitae, I developed clinical codes and software in Emily Couric Clinical Cancer Center at the University of Virginia.

Clinical Development

1. Developed mechanism to automate patient-specific IMRT-QA plan generation in our institution (Python and Pinnacle Scripting)
   • In our institution, once an approved plan exported from TPS to record and verify system, the corresponding QA plan is automatically created in a specific institution (Patient_QA) with the assigned plan revision number by TPS. The associated planar dose profiles for IMRT QA is created.
2. Developed GPU-accelerated software that performs robustness analysis for a given clinical plan.
   • Supported uncertainties: Daily setup, independent rigid motion of regions of interest uncertainty, and basic intra-fraction motion uncertainty models
   • Scripted from Pinnacle treatment planning system.
3. Developed a software that automated composite report generation in our institution
   • reads in pertinent DICOM files exported from Velocity AI.
   • calculates DVH using an in-house DVH computation engine
   • creates a standardized fusion/composite dose report
   • automatically performs EUD and EQD2 analysis for multiple treatment courses scenarios according to normal tissue tolerances established in our institution
4. Developed a software that automatically generates standardized written directive for all brachytherapy cases from Mosaiq schedule database (python, pyodbc, sqlalchemy)
5. Data mining codes to develop standardized reporting system for IORT cases using Eclipse/BrachyVision Scripting API and SQL queries from Mosaiq database.
6. Developed clinical scripts in Pinnacle TPS
7. Developed clinical signal processing and film analysis codes for annual and monthly QA for Truebeam, Trilogy, and Tomotherapy treatment modalities.
8. UVA's custom delineation features in Oncospace: the radiation oncology department of the university of Virginia is one of the collaborators in Johns Hopkins's oncospace program. The oncospace provides an intuitive human interface designed to ask clinically relevant questisons of larger inter-institutional patient data, also it provides convenient ways to perform statistical analysis. In oncospace DB, it is possible save custom features for different delineated regions of interest for a given plan in database table. I developed some python codes to extract UVA's ROI features which were not in the standard schema of the oncospace DB and store it in the database.