Deep-learning data processing for photon-counting CT
Title: Deep-learning data processing for photon-counting CT
DNr: Berzelius-2024-61
Project Type: LiU Berzelius
Principal Investigator: Mats Persson <>
Affiliation: Kungliga Tekniska högskolan
Duration: 2024-02-13 – 2024-09-01
Classification: 20603


X-ray computed tomography (CT) imaging is a widely used medical imaging modality with around 100 million examinations per year in the US alone, providing healthcare with an important tool for diagnosis and treatment planning for a wide range of diseases such as stroke, cancer, and cardiovascular disease. The latest improvement of this technology is photon-counting CT, which is based on a new, energy-resolving x-ray detector type that promises lower dose, higher spatial resolution, and improved tissue characterization. A key problem is to develop data processing algorithms for photon-counting CT that make optimal use of the measured data while being sufficiently fast to be clinically adopted. In recent years, deep-learning-based image-processing methods have been shown to be highly effective for improving image quality for conventional CT imaging while being computationally inexpensive to apply once the training process is complete. In our work we have also seen drastic improvements in image quality for photon-counting CT. Our goal is to develop deep-learning-based data processing methods for reduction of artifacts and noise that can be integrated into the image reconstruction process and lead to improved photon-counting CT imaging. We will train convolutional neural networks on simulated CT scans of numerical phantoms generated from anonymized CT image datasets available online, using images corrupted by artifacts and noise as input and corresponding high-fidelity images as labels. Since the images are anonymized, the data will not contain any protected health information. We will evaluate performance by reconstructing images of test objects resembling human patients imaged with a prototype photon-counting CT scanner, based on technology developed in our research group, that is in operation at MedTechLabs at Karolinska University Hospital. In our previous allocations, we found that Berzelius is well suited for this project since the default allocation allows us to train efficiently on datasets up to 40 GB and try out multi-GPU implementations for larger datasets. We obtained very promising results for image denoising, ring artifact reduction and material decomposition. In this continuation project, we will further develop our neural-network models in two main research directions: 1) We will develop an x-ray CT imaging technique that performs well for images acquired at very low radiation dose, which is important for applications such as lung-cancer screening and pediatric imaging. 2) We will further develop our method using the energy-resolved information to perform three-material decomposition into soft tissue, calcium and iodinated contrast agent, an important task for assessing cardiovascular disease, which has so far only been evaluated on a limited dataset. The expected outcome is a proof of concept demonstrating that incorporating deep-learning image processing steps in the CT image reconstruction can play an important role in improving the diagnostic quality of photon-counting CT. This will be an important step towards our long-term goal of introducing a novel highly accurate photon-counting CT imaging technique in clinical practice, with large potential benefits to human health.