Speaker
Description
Single-photon counting detectors with direct conversion to electrical signals enable highly accurate, noise-free, real-time particle detection, which is essential for high-precision applications such as dosimetry and beam diagnostics. The Timepix family of hybrid semiconductor pixelated detectors, known for their wide energy range detection capability and high-resolution positional accuracy, is well-suited for applications that require precise radiation measurement and monitoring. Their small size and low power consumption make them highly suitable for mobile and remote applications, including space radiation monitoring and particle physics research. Timepix3 detectors have found use in applications as diverse as medical radiotherapy, space radiation monitoring on satellites, radiation safety in nuclear power plants, and particle physics in accelerator environments.
This talk will present the design and implementation of a versatile, user-friendly data analysis tool called TraX Engine, which supports fast, reliable data analysis for Timepix-family detectors across multiple platforms, including a graphical user interface as a desktop application, a Python API, command-line tool, and a web-based interface. With cross-platform availability, it is accessible for both research and practical applications, with a focus on educational use through its online web portal.
The modular processing workflow addresses the unique demands of Timepix-family detector data across three primary levels:
1) Pre-processing – In this phase, adjacent pixels activated by a single particle are clustered, then calibration and corrections are applied. Clusterization extracts essential particle parameters, including energy, time-of-arrival, and spatial size. Cluster formation criteria vary by mode, with frame-based clustering relying on spatial proximity, and data-driven mode using both spatial and temporal criteria.
2) Processing – Machine learning-based algorithms in the presented analytical tool enhance particle classification and radiation field recognition even further. Particle identification techniques classify particle clusters by estimating particle type, energy, and morphological properties. Neural networks developed specifically for Timepix and Timepix3 detectors with various sensors classify particles into categories such as protons, photons, electrons, and ions. A hierarchical classification model provides further detail, distinguishing particle types and energies, such as low- to high-energy protons and helium ions.
3) Post-processing – Advanced data analysis methods, including directional and coincidence analysis, support complex assessments of radiation fields. Coincidence analysis identifies correlated particle events within defined time windows, essential in multi-detector setups or high-energy particle beam experiments where simultaneous particle detection provides critical insights. Directional analysis calculates particle trajectories by determining azimuth and elevation angles, a key factor in proton therapy and space radiation studies. Spatial mapping generates detailed visualizations of parameters like energy deposition and particle flux, aiding applications in beam profiling, field visualization, and material analysis.
The advanced analytical and processing capabilities makes TraxEngine a versatile tool across several fields, including:
Nuclear Power Plants – Accurate radiation monitoring in nuclear facilities is critical, where the TraxEngine supports real-time tracking of radiation fields and particle flux, ensuring reliable operation of dosimeters and radiation safety protocols.
Space and Aircraft Radiation Monitoring – Compact and energy-efficient, the MiniPIX Space radiation monitor continuously measures space radiation with minimal power usage. Equipped with the data-analysis software, it tracks dose rates, flux, and provides alerts for solar events, contributing to astronaut safety and spacecraft system protection.
Medical Applications – In proton therapy, this tool enables precise measurements of beam characteristics, including flux, energy deposition, and spatial distribution, ensuring accurate targeting of cancerous tissues. In imaging applications, such as gamma detection with Compton cameras, it aids in reconstructing photon origin, supporting internal radiation imaging critical to cancer treatment.
Particle Physics and Accelerator Environments – At particle accelerators, Timepix-family detectors are valuable for beam diagnostics, tracking particle interactions, and monitoring time structures in high-energy beams. Their comppactness and low power requirements make them ideal for placement near active beams and in hard-to-reach areas within accelerator facilities, where traditional detectors may be impractical. TraX Engine enables particle flux analysis, time structure monitoring, and temporal characterization, crucial for high-precision measurements in accelerator physics.
Educational and Practical Utility – The web-based portal facilitates remote data analysis and visualization, making it a useful tool in educational settings. Users can upload and process data from real-world scenarios, such as space radiation or medical imaging, providing hands-on experience in radiation field analysis.
The presented data analysis tool offers a comprehensive, user-friendly solution for real-time analysis of complex radiation fields from Timepix-family detectors. Through advanced pre-processing, machine learning-based particle classification, and post-processing techniques, it has great potential in precise radiation field analysis in nuclear power, space exploration, particle physics, medical applications and others. The system's compatibility across graphical user interface, Python API, command-line, and web platforms ensures broad accessibility for research, operational, and educational purposes. To illustrate its effectiveness, data from ongoing projects such as space radiation monitoring and medical beam diagnostics will be presented, highlighting its application in environments with diverse and complex radiation fields.
