Improved low-count quantitative PET reconstruction with an iterative neural network

TL;DR

This paper introduces an iterative neural network, BCD-Net, tailored for low-count PET image reconstruction, demonstrating significant improvements in image quality and noise reduction over traditional regularization methods, with good generalization to different data.

Contribution

The paper presents a modified BCD-Net architecture for PET MBIR that effectively enhances image quality in low-count scenarios, outperforming classical regularizers like TV and NLM.

Findings

BCD-Net significantly improves CNR and RMSE in low-count PET images.

The method generalizes well to different datasets and activity distributions.

Improvements are validated on both simulated and clinical phantom data.

Abstract

Image reconstruction in low-count PET is particularly challenging because gammas from natural radioactivity in Lu-based crystals cause high random fractions that lower the measurement signal-to-noise-ratio (SNR). In model-based image reconstruction (MBIR), using more iterations of an unregularized method may increase the noise, so incorporating regularization into the image reconstruction is desirable to control the noise. New regularization methods based on learned convolutional operators are emerging in MBIR. We modify the architecture of an iterative neural network, BCD-Net, for PET MBIR, and demonstrate the efficacy of the trained BCD-Net using XCAT phantom data that simulates the low true coincidence count-rates with high random fractions typical for Y-90 PET patient imaging after Y-90 microsphere radioembolization. Numerical results show that the proposed BCD-Net significantly improves CNR and RMSE of the reconstructed images compared to MBIR methods using non-trained regularizers, total variation (TV) and non-local means (NLM). Moreover, BCD-Net successfully generalizes to test data that differs from the training data. Improvements were also demonstrated for the clinically relevant phantom measurement data where we used training and testing datasets having very different activity distributions and count-levels.