Machine learning classifiers evaluated for PET scanner background rejection

Researchers have systematically evaluated machine learning classifiers, including XGBoost, AdaBoost, and neural networks, for background rejection in LAFOV PET scanners. The study focused on event-by-event classification using Monte Carlo simulations of the Siemens Biograph Vision Quadra scanner. Results indicate that a 4-feature model, incorporating attenuation factor, time difference, and energy-related variables, generalizes better across different phantom geometries than a more complex 6-feature model. While these machine learning approaches outperform traditional methods, further gains in image quality may require larger input representations or specific handling of scattered coincidences. AI

RANK_REASON Research paper detailing a systematic evaluation of machine learning classifiers for a specific application. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CV →

• AdaBoost
• artificial neural network
• Konrad Klimaszewski
• LAFOV PET
• NEMA IEC
• Siemens Biograph Vision Quadra
• XGBoost

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →