MlPET: A Localized Neural Network Approach for Probabilistic Post-Reconstruction PET Image Analysis Using Informed Priors

TL;DR

MlPET introduces a neural network-based method for probabilistic PET image analysis that enhances resolution and reduces noise, outperforming standard methods in phantom tests and enabling faster, more reliable imaging.

Contribution

The paper presents MlPET, a novel localized neural network approach that replaces traditional sampling methods for probabilistic PET analysis, incorporating scanner-specific priors for improved image quality.

Findings

Higher contrast recovery coefficients than standard PET

Reduction of point spread function FWHM from ~2mm to below 1mm

Achieves comparable image quality at significantly shorter acquisition times

Abstract

We develop and evaluate MlPET, a fast localized machine learning approach for probabilistic PET image analysis addressing the noise-resolution trade-off in conventional reconstructions. MlPET replaces computationally demanding Markov chain Monte Carlo sampling with a localized neural network trained to estimate posterior mean voxel activity from small image neighborhoods. The method incorporates scanner-specific point spread functions, spatially correlated noise modeling, and flexible priors. Performance was evaluated on NEMA IEC phantom data from three PET systems (GE Discovery MI, Siemens Biograph Vision 600, and Quadra) under varying reconstruction settings and acquisition times. On phantom data, MlPET achieved contrast recovery coefficients consistently higher than standard PET and close to 1.0 (including 10 mm spheres), while reducing background noise and improving spatial definition. Effective pointspread function full width at half maximum decreased from approximately 2 mm in standard PET to below 1 mm with MlPET, a 2.5 fold reduction in blur. Comparable image quality was obtained at 40-80 s acquisition time with MlPET versus 900 s with conventional PET. MlPET provides an efficient approach for quantitative probabilistic post-reconstruction PET analysis. By combining informed priors with neural network speed, it achieves noise suppression and resolution enhancement without altering reconstruction algorithms. The method shows promise for improved small-lesion detectability and quantitative reliability in clinical PET imaging. Future studies will evaluate performance on patient data.