Angel Pineda

Institute: Hofstra University

Date: September 26, 2024

Title: Task-based assessment for neural network reconstructions: optimizing undersampled MRI based on human observer signal detection

Abstract: Neural networks are being used to reconstruct undersampled magnetic resonance imaging (MRI). Because of the complexity of the artifacts in the reconstructed images, task-based approaches of image quality are needed. Common metrics for evaluating image quality like the normalized root mean squared error (NRMSE) and structural similarity (SSIM) are global metrics which average out impact of subtle features in the images. Using measures of image quality which incorporate a subtle signal for a specific task allow for image quality assessment which locally evaluates the effect of undersampling on a signal detection. We used a U-Net to reconstruct under-sampled images with 2x, 3x, 4x and 5x fold 1-D undersampling rates. Cross validation was performed with both structural similarity (SSIM) and mean squared error (MSE) losses. A two alternative forced choice (2-AFC) observer study was carried out for detecting a subtle signal (small blurred disk) from images. We found that for both loss functions, the human observer performance on the 2-AFC studies led to a choice of a 2x undersampling but the SSIM and NRMSE led to a choice of a 3x undersampling. For this task, SSIM and NRMSE overestimate the achievable undersampling using a U-Net before a steep loss of image quality when compared to the performance of human observers in the detection of a subtle lesion.  In this talk, I will also give an overview of my academic journey which includes a PhD in applied mathematics and a postdoctoral fellowship in a radiology department followed by faculty positions in mathematics departments which emphasize teaching and have sizable fractions of students from underrepresented groups in mathematics.