Publications | Ishak Ayad

2024

UnWave-Net: Unrolled Wavelet Network for Compton Tomography Image Reconstruction

Ishak Ayad, Cécilia Tarpau, Javier Cebiero, and Mai K. Nguyen

In Proceedings of the International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI), 2024

MICCAI Young Scientist Award

Early Acceptance - Top 11%

Abs PDF Code

Computed tomography (CT) is a widely used medical imaging technique to scan internal structures of a body, typically involving collimation and mechanical rotation. Compton scatter tomography (CST) presents an interesting alternative to conventional CT by leveraging Compton physics instead of collimation to gather information from multiple directions. While CST introduces new imaging opportunities with several advantages such as high sensitivity, compactness, and entirely fixed systems, image reconstruction remains an open problem due to the mathematical challenges of CST modeling. In contrast, deep unrolling networks have demonstrated potential in CT image reconstruction, despite their computationally intensive nature. In this study, we investigate the efficiency of unrolling networks for CST image reconstruction. To address the important computational cost required for training, we propose UnWave-Net, a novel unrolled wavelet-based reconstruction network. This architecture includes a non-local regularization term based on wavelets, which captures long-range dependencies within images and emphasizes the multi-scale components of the wavelet transform. We evaluate our approach using a CST of circular geometry which stays completely static during data acquisition, where UnWave-Net facilitates image reconstruction in the absence of a specific reconstruction formula. Our method outperforms existing approaches and achieves state-of-theart performance in terms of SSIM and PSNR, and offers an improved computational efficiency compared to traditional unrolling networks.

Oral
QN-Mixer: A Quasi-Newton MLP-Mixer Model for Sparse-View CT Reconstruction

Ishak Ayad, Nicolas Larue, and Mai K. Nguyen

In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2024

Abs PDF

Inverse problems span across diverse fields. In medical contexts, computed tomography (CT) plays a crucial role in reconstructing a patient’s internal structure, presenting challenges due to artifacts caused by inherently ill-posed inverse problems. Previous research advanced image quality via post-processing and deep unrolling algorithms but faces challenges, such as extended convergence times with ultra-sparse data. Despite enhancements, resulting images often show significant artifacts, limiting their effectiveness for real-world diagnostic applications. We aim to explore deep second-order unrolling algorithms for solving imaging inverse problems, emphasizing their faster convergence and lower time complexity compared to common first-order methods like gradient descent. In this paper, we introduce QN-Mixer, an algorithm based on the quasi-Newton approach. We use learned parameters through the BFGS algorithm and introduce Incept-Mixer, an efficient neural architecture that serves as a non-local regularization term, capturing long-range dependencies within images. To address the computational demands typically associated with quasi-Newton algorithms that require full Hessian matrix computations, we present a memory-efficient alternative. Our approach intelligently downsamples gradient information, significantly reducing computational requirements while maintaining performance. The approach is validated through experiments on the sparse-view CT problem, involving various datasets and scanning protocols, and is compared with post-processing and deep unrolling stateof-the-art approaches. Our method outperforms existing approaches and achieves state-of-the-art performance in terms of SSIM and PSNR, all while reducing the number of unrolling iterations required.

2022

Tomographic image reconstruction from incomplete data via a hybrid GAN

Ishak Ayad, Cécilia Tarpau, and Mai K. Nguyen

In IEEE NSS MIC RTSD, 2022

Abs PDF

Limited data tomographic reconstruction has been widely used in medical imaging to reduce the radiation dose or shorten the data acquisition time. This is achieved by truncating the scanning angular range or increasing the angular sampling rate. However, the use of classical reconstruction methods with such incomplete data involves severe streaking artifacts, blurred edges, distorted boundaries, contrast loss, and decreased intensities in the reconstructed images. In this work, we propose a generative adversarial network made of a V-net generator with modified skip connections that simulates the algebraic reconstruction and a discriminator combining both information from the image and projection domains which only penalizes structure at the scale of image patches. The proposed method obtains promising results for data coming from both limitedangle and sparse-view projections, that is, data acquired from an angular range of 90 degrees and an angular sampling step of 10 degrees.

2021

Deep morphological network-based artifact suppression for limited-angle tomography

Ishak Ayad, Cécilia Tarpau, Mai K Nguyen, and Son Vu

In Proceedings of the International Conference on Image Processing and Computer Vision (IPCV), 2021

Abs PDF

Computed tomography has been widely used in biomedical and industrial applications. The well-known filtered back-projection algorithm, probably the most used reconstruction technique, fails when the angular range used for data acquisition is not sufficient. As a consequence, reconstructions exhibit artifacts. In order to eliminate these artifacts, we propose in this article a new deep learning approach based on a U-net architecture which includes a morphological operation. This operation of mathematical morphology allows us to capture better some non-linear properties of the object to reconstruct. The proposed method provides good results for angular ranges of 170, 150, 130 and even 110 degrees. To the best of our knowledge, it is the first time a limited-angle artifact suppression method works with 110 projections.