BoneHost is a high-performance interactive platform for 3D human bone analysis. Developed from my diploma thesis on statistical shape analysis, it combines GPU-accelerated WebGL (Unity) with a machine learning backend. The system operates entirely browser-based, providing an intuitive research environment for medical and anatomical studies.
Linux (Ubuntu) Infrastructure
Serves as a high-performance, bare-metal foundation for the entire BoneHost ecosystem.
Orchestrates the core runtime environments, including C++ compilers, web servers, and container daemons.
Docker Containerization
Facilitates the modular execution of backend services, WebGL components, and ML modules.
Enhances system reliability and maintainability through strict service isolation.
Orchestration via Docker Compose
Manages complex service dependencies via unified, declarative configuration.
Standardizes deployment workflows, ensuring parity between development and production.
WebGL & Unity Frontend
Delivers low-latency, hardware-accelerated 3D visualization directly in the browser.
High-fidelity interactivity (rotation, scaling, selection) is powered by optimized C# logic.
Mathematical Framework & ML
Linear Algebra: Essential framework for coordinate transformations and statistical shape analysis.
Logistic Regression: Core classification logic for predicting ancestry based on osteological features.
Full-Stack Development (C++ / C#)