Search

The success of embolisation, a minimally invasive treatment of liver cancer, could be evaluated in the operational room with cone-beam CT by acquiring a dynamic perfusion scan. The reconstruction algorithm must address the issues of low temporal sampling and higher noise levels inherent in cone-beam CT systems, compared to conventional CT. Therefore, a model-based perfusion reconstruction based on the time separation technique (TST) was applied. TST uses basis functions to model time attenuation curves. These functions are either analytical or based on prior knowledge, extracted using singular value decomposition from CT perfusion data. To explore how well the prior knowledge can model perfusion dynamics and what the potential limitations are, the dynamic CBCT perfusion scan was simulated under different noise levels. The TST method was compared to static reconstruction. It was demonstrated that a set consisting of only four basis functions results in perfusion maps that preserve relevant information, denoises the data, and outperforms static reconstruction under higher noise levels. TST with prior knowledge would not only outperform static reconstruction, but also the TST with analytical basis functions. Furthermore, it has been shown that only eight CBCT rotations, unlike previously assumed ten, are sufficient to obtain the perfusion maps comparable to the reference CT perfusion maps. This contributes to saving dose and reconstruction time. The real dynamic CBCT perfusion scan, reconstructed under the same conditions as the simulated scan, shows potential for maintaining the accuracy of the perfusion maps. By visual inspection, the embolised region was matching to that in corresponding CT perfusion maps. Further analysis of a larger cohort of patient data is needed to draw final conclusions regarding the expected advantages of the time separation technique., Comment: 28 pages, 11 figures

Perfusion imaging is a valuable tool for diagnosing and treatment planning for liver tumours. The time separation technique (TST) has been successfully used for modelling C-arm cone-beam computed tomography (CBCT) perfusion data. The reconstruction can be accompanied by the segmentation of the liver - for better visualisation and for generating comprehensive perfusion maps. Recently introduced Turbolift learning has been seen to perform well while working with TST reconstructions, but has not been explored for the time-resolved volumes (TRV) estimated out of TST reconstructions. The segmentation of the TRVs can be useful for tracking the movement of the liver over time. This research explores this possibility by training the multi-scale attention UNet of Turbolift learning at its third stage on the TRVs and shows the robustness of Turbolift learning since it can even work efficiently with the TRVs, resulting in a Dice score of 0.864$\pm$0.004.

Model-based reconstruction employing the time separation technique (TST) was found to improve dynamic perfusion imaging of the liver using C-arm cone-beam computed tomography (CBCT). To apply TST using prior knowledge extracted from CT perfusion data, the liver should be accurately segmented from the CT scans. Reconstructions of primary and model-based CBCT data need to be segmented for proper visualisation and interpretation of perfusion maps. This research proposes Turbolift learning, which trains a modified version of the multi-scale Attention UNet on different liver segmentation tasks serially, following the order of the trainings CT, CBCT, CBCT TST - making the previous trainings act as pre-training stages for the subsequent ones - addressing the problem of limited number of datasets for training. For the final task of liver segmentation from CBCT TST, the proposed method achieved an overall Dice scores of 0.874$\pm$0.031 and 0.905$\pm$0.007 in 6-fold and 4-fold cross-validation experiments, respectively - securing statistically significant improvements over the model, which was trained only for that task. Experiments revealed that Turbolift not only improves the overall performance of the model but also makes it robust against artefacts originating from the embolisation materials and truncation artefacts. Additionally, in-depth analyses confirmed the order of the segmentation tasks. This paper shows the potential of segmenting the liver from CT, CBCT, and CBCT TST, learning from the available limited training data, which can possibly be used in the future for the visualisation and evaluation of the perfusion maps for the treatment evaluation of liver diseases.

This paper proposes an extension to the Dual Branch Prior-Net for sparse view interventional CBCT reconstruction incorporating a high quality planning scan. An additional head learns to segment interventional instruments and thus guides the reconstruction task. The prior scans are misaligned by up to +-5deg in-plane during training. Experiments show that the proposed model, Dual Branch Prior-SegNet, significantly outperforms any other evaluated model by >2.8dB PSNR. It also stays robust wrt. rotations of up to +-5.5deg., Comment: 3 pages, 1 figure, accepted short paper submission at MIDL 2022

In this paper, the Primal-Dual UNet for sparse view CT reconstruction is modified to be applicable to cone beam projections and perform reconstructions of entire volumes instead of slices. Experiments show that the PSNR of the proposed method is increased by 10dB compared to the direct FDK reconstruction and almost 3dB compared to the modified original Primal-Dual Network when using only 23 projections. The presented network is not optimized wrt. memory consumption or hyperparameters but merely serves as a proof of concept and is limited to low resolution projections and volumes., Comment: 3 pages, 1 figure, accepted short paper submission at MIDL 2022

Magnetic resonance imaging (MRI) provides high spatial resolution and excellent soft-tissue contrast without using harmful ionising radiation. Dynamic MRI is an essential tool for interventions to visualise movements or changes of the target organ. However, such MRI acquisition with high temporal resolution suffers from limited spatial resolution - also known as the spatio-temporal trade-off of dynamic MRI. Several approaches, including deep learning based super-resolution approaches, have been proposed to mitigate this trade-off. Nevertheless, such an approach typically aims to super-resolve each time-point separately, treating them as individual volumes. This research addresses the problem by creating a deep learning model which attempts to learn both spatial and temporal relationships. A modified 3D UNet model, DDoS-UNet, is proposed - which takes the low-resolution volume of the current time-point along with a prior image volume. Initially, the network is supplied with a static high-resolution planning scan as the prior image along with the low-resolution input to super-resolve the first time-point. Then it continues step-wise by using the super-resolved time-points as the prior image while super-resolving the subsequent time-points. The model performance was tested with 3D dynamic data that was undersampled to different in-plane levels. The proposed network achieved an average SSIM value of 0.951$\pm$0.017 while reconstructing the lowest resolution data (i.e. only 4\% of the k-space acquired) - which could result in a theoretical acceleration factor of 25. The proposed approach can be used to reduce the required scan-time while achieving high spatial resolution.

Computed tomography and magnetic resonance imaging are two widely used clinical imaging modalities for non-invasive diagnosis. However, both of these modalities come with certain problems. CT uses harmful ionising radiation, and MRI suffers from slow acquisition speed. Both problems can be tackled by undersampling, such as sparse sampling. However, such undersampled data leads to lower resolution and introduces artefacts. Several techniques, including deep learning based methods, have been proposed to reconstruct such data. However, the undersampled reconstruction problem for these two modalities was always considered as two different problems and tackled separately by different research works. This paper proposes a unified solution for both sparse CT and undersampled radial MRI reconstruction, achieved by applying Fourier transform-based pre-processing on the radial MRI and then finally reconstructing both modalities using sinogram upsampling combined with filtered back-projection. The Primal-Dual network is a deep learning based method for reconstructing sparsely-sampled CT data. This paper introduces Primal-Dual UNet, which improves the Primal-Dual network in terms of accuracy and reconstruction speed. The proposed method resulted in an average SSIM of 0.932\textpm0.021 while performing sparse CT reconstruction for fan-beam geometry with a sparsity level of 16, achieving a statistically significant improvement over the previous model, which resulted in 0.919\textpm0.016. Furthermore, the proposed model resulted in 0.903\textpm0.019 and 0.957\textpm0.023 average SSIM while reconstructing undersampled brain and abdominal MRI data with an acceleration factor of 16, respectively - statistically significant improvements over the original model, which resulted in 0.867\textpm0.025 and 0.949\textpm0.025.

CT perfusion imaging (CTP) plays an important role in decision making for the treatment of acute ischemic stroke with large vessel occlusion. Since the CT perfusion scan time is approximately one minute, the patient is exposed to a non-negligible dose of ionizing radiation. However, further dose reduction increases the level of noise in the data and the resulting perfusion maps. We present a method for reducing noise in perfusion data based on dimension reduction of time attenuation curves. For dimension reduction, we use either the fit of the first five terms of the trigonometric polynomial or the first five terms of the SVD decomposition of the time attenuation profiles. CTP data from four patients with large vessel occlusion and three control subjects were studied. To compare the noise level in the perfusion maps, we use the wavelet estimation of the noise standard deviation implemented in the scikit-image package. We show that both methods significantly reduce noise in the data while preserving important information about the perfusion deficits. These methods can be used to further reduce the dose in CT perfusion protocols or in perfusion studies using C-arm CT, which are burdened by high noise levels., Comment: 4 pages, 2 figures, submitted and accepted as a contribution at the SPIE Medical Imaging 2022

Perfusion imaging is an interesting new modality for evaluation and assessment of the liver cancer treatment. C-Arm CT provides a possibility to perform perfusion imaging scans intra-operatively for even faster evaluation. The slow speed of the C-Arm CT rotation and the presence of the noise, however, have an impact on the reconstruction and therefore model based approaches have to be applied. In this work we apply the Time separation technique (TST), to denoise data, speed up reconstruction and improve resulting perfusion images. We show on animal experiment data that Dynamic C-Arm CT Liver Perfusion Imaging together with the processing of the data based on the TST provides comparable results to standard CT liver perfusion imaging., Comment: 5 pages, 8 figures, published in Proceedings of the 16th Virtual International Meeting on Fully 3D Image Reconstruction in Radiology and Nuclear Medicine 2021 arXiv:2110.04143

Krylov subspace methods are considered a standard tool to solve large systems of linear algebraic equations in many scientific disciplines such as image restoration or solving partial differential equations in mechanics of continuum. In the context of computer tomography however, the mostly used algebraic reconstruction techniques are based on classical iterative schemes. In this work we present software package that implements fully 3D cone beam projection operator and uses Krylov subspace methods, namely CGLS and LSQR to solve related tomographic reconstruction problems. It also implements basic preconditioning strategies. On the example of the cone beam CT reconstruction of 3D Shepp-Logan phantom we show that the speed of convergence of the CGLS clearly outperforms PSIRT algorithm. Therefore Krylov subspace methods present an interesting option for the reconstruction of large 3D cone beam CT problems., Comment: 5 pages, 3 figures, published in Proceedings of the 16th Virtual International Meeting on Fully 3D Image Reconstruction in Radiology and Nuclear Medicine 2021 arXiv:2110.04143

In this paper, we introduce a novel class of projectors for 3D cone beam tomographic reconstruction. Analytical formulas are derived to compute the relationship between the volume of a voxel projected onto a detector pixel and its contribution to the line integral of attenuation recorded by that pixel. Based on these formulas, we construct a near-exact projector and backprojector, particularly suited for algebraic reconstruction techniques and hierarchical reconstruction approaches with nonuniform voxel grids. Unlike traditional projectors, which assume a uniform grid with fixed voxel sizes, our method enables local refinement of voxels, allowing for adaptive grid resolution and improved reconstruction quality in regions of interest. We have implemented this cutting voxel projector along with a relaxed, speed-optimized version and compared them to two established projectors: a ray-tracing projector based on Siddon's algorithm and a TT footprint projector. Our results demonstrate that the cutting voxel projector achieves higher accuracy than the TT projector, especially for large cone beam angles. Furthermore, the relaxed version of the cutting voxel projector offers a significant speed advantage over current footprint projector implementations, while maintaining comparable accuracy. In contrast, Siddon's algorithm, when achieving similar accuracy, is considerably slower than the cutting voxel projector. All algorithms are implemented in an open-source framework for algebraic reconstruction using OpenCL and C++, optimized for efficient GPU computation. GitHub repository of the project https://github.com/kulvait/KCT_cbct., Comment: 18 pages, 4 figures

CT perfusion imaging (CTP) is used in the diagnostic workup of acute ischemic stroke (AIS). CTP may be performed within the angio suite using flat detector CT (FDCT) to help reduce patient management time. In order to significantly improve FDCT perfusion (FDCTP) imaging, data-processing algorithms need to be able to compensate for the higher levels of noise, slow rotation speed, and a lower frame rate of current FDCT devices. We performed a realistic simulation of FDCTP acquisition based on CTP data from seven subjects. We used the time separation technique (TST) as a model-based approach for FDCTP data processing. We propose a novel dimension reduction in which we approximate the time attenuation curves by a linear combination of trigonometric functions. Our goal was to show that the TST can be used even without prior assumptions on the shape of the attenuation profiles. We first demonstrated that a trigonometric basis is suitable for dimension reduction of perfusion data. Using simulated FDCTP data, we have shown that a trigonometric basis in the TST provided better results than the classical straightforward processing even with additional noise. Average correlation coefficients of perfusion maps were improved for cerebral blood flow (CBF), cerebral blood volume, mean transit time (MTT) maps. In a moderate noise scenario, the average Pearson's coefficient for the CBF map was improved using the TST from 0.76 to 0.81. For the MTT map, it was improved from 0.37 to 0.45. Furthermore, we achieved a total processing time from the reconstruction of FDCTP data to the generation of perfusion maps of under 5 min. In our study cohort, perfusion maps created from FDCTP data using the TST with a trigonometric basis showed equivalent perfusion deficits to classic CT perfusion maps. It follows, that this novel FDCTP technique has potential to provide fast and accurate FDCTP imaging for AIS patients., Comment: 14 pages, 5 figures, accepted in Medical Physics, due to the licensing and paid open access I upload official version of published article in Early view status, it is to be included into the issue in Medical Physics so that full journal reference is subject to change

MRI is an inherently slow process, which leads to long scan time for high-resolution imaging. The speed of acquisition can be increased by ignoring parts of the data (undersampling). Consequently, this leads to the degradation of image quality, such as loss of resolution or introduction of image artefacts. This work aims to reconstruct highly undersampled Cartesian or radial MR acquisitions, with better resolution and with less to no artefact compared to conventional techniques like compressed sensing. In recent times, deep learning has emerged as a very important area of research and has shown immense potential in solving inverse problems, e.g. MR image reconstruction. In this paper, a deep learning based MR image reconstruction framework is proposed, which includes a modified regularised version of ResNet as the network backbone to remove artefacts from the undersampled image, followed by data consistency steps that fusions the network output with the data already available from undersampled k-space in order to further improve reconstruction quality. The performance of this framework for various undersampling patterns has also been tested, and it has been observed that the framework is robust to deal with various sampling patterns, even when mixed together while training, and results in very high quality reconstruction, in terms of high SSIM (highest being 0.990$\pm$0.006 for acceleration factor of 3.5), while being compared with the fully sampled reconstruction. It has been shown that the proposed framework can successfully reconstruct even for an acceleration factor of 20 for Cartesian (0.968$\pm$0.005) and 17 for radially (0.962$\pm$0.012) sampled data. Furthermore, it has been shown that the framework preserves brain pathology during reconstruction while being trained on healthy subjects.

Dynamic imaging is a beneficial tool for interventions to assess physiological changes. Nonetheless during dynamic MRI, while achieving a high temporal resolution, the spatial resolution is compromised. To overcome this spatio-temporal trade-off, this research presents a super-resolution (SR) MRI reconstruction with prior knowledge based fine-tuning to maximise spatial information while reducing the required scan-time for dynamic MRIs. An U-Net based network with perceptual loss is trained on a benchmark dataset and fine-tuned using one subject-specific static high resolution MRI as prior knowledge to obtain high resolution dynamic images during the inference stage. 3D dynamic data for three subjects were acquired with different parameters to test the generalisation capabilities of the network. The method was tested for different levels of in-plane undersampling for dynamic MRI. The reconstructed dynamic SR results after fine-tuning showed higher similarity with the high resolution ground-truth, while quantitatively achieving statistically significant improvement. The average SSIM of the lowest resolution experimented during this research (6.25~\% of the k-space) before and after fine-tuning were 0.939 $\pm$ 0.008 and 0.957 $\pm$ 0.006 respectively. This could theoretically result in an acceleration factor of 16, which can potentially be acquired in less than half a second. The proposed approach shows that the super-resolution MRI reconstruction with prior-information can alleviate the spatio-temporal trade-off in dynamic MRI, even for high acceleration factors.

The outbreak of COVID-19 has shocked the entire world with its fairly rapid spread and has challenged different sectors. One of the most effective ways to limit its spread is the early and accurate diagnosing infected patients. Medical imaging, such as X-ray and Computed Tomography (CT), combined with the potential of Artificial Intelligence (AI), plays an essential role in supporting medical personnel in the diagnosis process. Thus, in this article five different deep learning models (ResNet18, ResNet34, InceptionV3, InceptionResNetV2 and DenseNet161) and their ensemble, using majority voting have been used to classify COVID-19, pneumoni{\ae} and healthy subjects using chest X-ray images. Multilabel classification was performed to predict multiple pathologies for each patient, if present. Firstly, the interpretability of each of the networks was thoroughly studied using local interpretability methods - occlusion, saliency, input X gradient, guided backpropagation, integrated gradients, and DeepLIFT, and using a global technique - neuron activation profiles. The mean Micro-F1 score of the models for COVID-19 classifications ranges from 0.66 to 0.875, and is 0.89 for the ensemble of the network models. The qualitative results showed that the ResNets were the most interpretable models. This research demonstrates the importance of using interpretability methods to compare different models before making a decision regarding the best performing model.

Dynamic MRI is a technique of acquiring a series of images continuously to follow the physiological changes over time. However, such fast imaging results in low resolution images. In this work, abdominal deformation model computed from dynamic low resolution images have been applied to high resolution image, acquired previously, to generate dynamic high resolution MRI. Dynamic low resolution images were simulated into different breathing phases (inhale and exhale). Then, the image registration between breathing time points was performed using the B-spline SyN deformable model and using cross-correlation as a similarity metric. The deformation model between different breathing phases were estimated from highly undersampled data. This deformation model was then applied to the high resolution images to obtain high resolution images of different breathing phases. The results indicated that the deformation model could be computed from relatively very low resolution images., Comment: 2019 41st Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC)

Granular material flowing through bottlenecks like the openings of silos tend to clog and to inhibit further flow. We study this phenomenon in a three-dimensional hopper for spherical and shape-anisotropic particles by means of X-ray tomography. The X-ray tomograms provide information on the bulk of the granular filling, and allows to determine the particle positions and orientations inside the silo. In addition, it allows to calculate local packing densities in different parts of the container. We find that in the flowing zone of the silo particles show a preferred orientation and thereby a higher order. Similarly to simple shear flows, the average orientation of the particles is not parallel to the streamlines but encloses a certain angle with it. In most parts of the hopper, the angular distribution of the particles did not reach the one corresponding to stationary shear flow, thus the average orientation angle in the hopper deviates more from the streamlines than in stationary shear flows. In the flowing parts of the silo shear induced dilation is observed, which is more pronounced for elongated grains than for nearly spherical particles. The clogged state is characterized by a dome, i. e. the geometry of the layer of grains blocking the outflow. The shape of the dome depends on the particle shape., Comment: 10 pages, 6 figures

The evolution of wide shear zones (or shear bands) was investigated experimentally and numerically for quasistatic dry granular flows in split bottom shear cells. We compare the behavior of materials consisting of beads, irregular grains (e.g. sand) and elongated particles. Shearing an initially random sample, the zone width was found to significantly decrease in the first stage of the process. The characteristic shear strain associated with this decrease is about unity and it is systematically increasing with shape anisotropy, i.e. when the grain shape changes from spherical to irregular (e.g. sand) and becomes elongated (pegs). The strongly decreasing tendency of the zone width is followed by a slight increase which is more pronounced for rod like particles than for grains with smaller shape anisotropy (beads or irregular particles). The evolution of the zone width is connected to shear induced density change and for nonspherical particles it also involves grain reorientation effects. The final zone width is significantly smaller for irregular grains than for spherical beads., Comment: 11 pages, 12 figures, submitted to Phys. Rev. E

Granular material exposed to shear shows a variety of unique phenomena: Reynolds dilatancy, positional order and orientational order effects may compete in the shear zone. We study granular packings consisting of macroscopic prolate, oblate and spherical grains and compare their behaviour. X-ray tomography is used to determine the particle positions and orientations in a cylindrical split bottom shear cell. Packing densities and the arrangements of individual particles in the shear zone are evaluated. For anisometric particles, we observe the competition of two opposite effects. One the one hand, the sheared granulate is dilated, but on the other hand the particles reorient and align with respect to the streamlines. Even though aligned cylinders in principle may achieve higher packing densities, this alignment compensates for the effect of dilatancy only partially. The complex rearrangements lead to a depression of the surface above the well oriented region while neigbouring parts still show the effect of dilation in the form of heaps. For grains with isotropic shapes, the surface remains rather flat. Perfect monodisperse spheres crystallize in the shear zone, whereby positional order partially overcompensates dilatancy effects. However, already slight deviations from the ideal monodisperse sphere shape inhibit crystallization., Comment: 12 pages, 13 figures, accepted in Soft Matter