DL-based inpainting for metal artifact reduction for cone beam CT using metal path length information

BACKGROUND

Metallic implants, which are inserted into the patient's body during trauma interventions, are the main cause of heavy artifacts in 3D X-ray acquisitions. These artifacts then hinder the evaluation of the correct implant's positioning, thus leading to a disturbed patient's healing process and increased revision rates.

PURPOSE

This problem is tackled by so-called metal artifact reduction (MAR) methods. This paper examines possible advances in the inpainting process of such MAR methods to decrease disruptive artifacts while simultaneously preserving important anatomical structures adjacent to the inserted implants.

METHODS

In this paper, a learning-based inpainting method for cone-beam computed tomography is proposed that couples a convolutional neural network (CNN) with an estimated metal path length as prior knowledge. Further, the proposed method is solely trained and evaluated on real measured data.

RESULTS

The proposed inpainting approach shows advantages over the inpainting method used by the currently clinically approved frequency split metal artifact reduction (fsMAR) method as well as the learning-based state-of-the-art (SOTA) method PConv-Net. The major improvement of the proposed inpainting method lies in the ability to correctly preserve important anatomical structures in those regions adjacent to the metal implants. Especially these regions are highly important for a correct implant's positioning in an intraoperative setup. Using the proposed inpainting, the corresponding MAR volumes reach a mean structural similarity index measure (SSIM) score of 0.9974 and outperform the other methods by up to 6 dB on single slices regarding the peak signal-to-noise ratio (PSNR) score. Furthermore, it can be shown that the proposed method can generalize to clinical cases at hand.

CONCLUSIONS

In this paper, a learning-based inpainting network is proposed that leverages prior knowledge about the metal path length of the inserted implant. Evaluations on real measured data reveal an increased overall MAR performance, especially regarding the preservation of anatomical structures adjacent to the inserted implants. Further evaluations suggest the ability of the proposed approach to generalize to clinical cases.

Diffusion-weighted MRI for differentiation of neuroblastoma and ganglioneuroblastoma/ganglioneuroma

PURPOSE

The purpose of this study was to assess the apparent diffusion coefficient (ADC) of neuroblastic tumours and to evaluate if the ADC can enable differentiation of neuroblastoma and ganglioneuroma/ganglioneuroblastoma.

PATIENTS AND METHODS

16 histologically classified tumours (10 neuroblastomas and 6 ganglioneuroma/ganglioneuroblastoma) were investigated in 15 children. Diffusion-weighted echo-planar imaging was performed with a b-value of 800s/mm². The contrast of tumour tissue depicted with T2-weighted images and diffusion-weighted images was evaluated by means of region-of-interest (ROI) measurements and a calculation of the ADC by a software tool. The ADC of the psoas-muscle was measured to establish an internal standard, too.

RESULTS

The mean ADC of the 10 neuroblastomas was 0.81×10⁻³mm²/s (SD 0.29×10⁻³mm²/s, range 0.39-1.47×10⁻³)mm²/s). The mean ADC of the four ganglioneuroma and two ganglioneuroblastoma was 1.6×10⁻³mm²/s (SD 0.340×10⁻³mm²/s, range 1.13-1.99)×10⁻³mm²/s. The difference was significant in the t-test (p=0.01). We found no ganglioneuroma or ganglioneuroblastoma with an ADC below 1.1×10⁻³mm²/s.

DISCUSSION

There is a significant difference of the ADC of neuroblastoma compared to the ADC of ganglioneuroma/ganglioneuroblastoma. These first results suggest that the diffusion-weighted imaging could differentiate neuroblastoma and ganglioneuroma/ganglioneuroblastoma by calculating the ADC.

Reduced interhemispheric structural connectivity between anterior cingulate cortices in borderline personality disorder

Functional and structural alterations of the anterior cingulate cortex (ACC), a key region for emotional and cognitive processing, are associated with borderline personality disorder (BPD). However, the interhemispheric structural connectivity between the left and right ACC and between other prefrontal regions in this condition is unknown. We acquired diffusion-tensor imaging data from 20 healthy women and 19 women with BPD and comorbid attention-deficit hyperactivity disorder (ADHD). Interhemispheric structural connectivity between both sides of the ACC, dorsolateral prefrontal cortices and medial orbitofrontal cortices was assessed by a novel probabilistic diffusion tensor-based fiber tracking method. In the BPD group as compared with healthy controls, we found decreased interhemispheric structural connectivity between both ACCs in fiber tracts that pass through the anterior corpus callosum and connect dorsal areas of the ACCs. Decreased interhemispheric structural connectivity between both ACCs may be a structural correlate of BPD.

Ventral and dorsal pathways for language

Built on an analogy between the visual and auditory systems, the following dual stream model for language processing was suggested recently: a dorsal stream is involved in mapping sound to articulation, and a ventral stream in mapping sound to meaning. The goal of the study presented here was to test the neuroanatomical basis of this model. Combining functional magnetic resonance imaging (fMRI) with a novel diffusion tensor imaging (DTI)-based tractography method we were able to identify the most probable anatomical pathways connecting brain regions activated during two prototypical language tasks. Sublexical repetition of speech is subserved by a dorsal pathway, connecting the superior temporal lobe and premotor cortices in the frontal lobe via the arcuate and superior longitudinal fascicle. In contrast, higher-level language comprehension is mediated by a ventral pathway connecting the middle temporal lobe and the ventrolateral prefrontal cortex via the extreme capsule. Thus, according to our findings, the function of the dorsal route, traditionally considered to be the major language pathway, is mainly restricted to sensory-motor mapping of sound to articulation, whereas linguistic processing of sound to meaning requires temporofrontal interaction transmitted via the ventral route.

Connecting and merging fibres: pathway extraction by combining probability maps

Probability mapping of connectivity is a powerful tool to determine the fibre structure of white matter in the brain. Probability maps are related to the degree of connectivity to a chosen seed area. In many applications, however, it is necessary to isolate a fibre bundle that connects two areas. A frequently suggested solution is to select curves, which pass only through two or more areas. This is very inefficient, especially for long-distance pathways and small areas. In this paper, a novel probability-based method is presented that is capable of extracting neuronal pathways defined by two seed points. A Monte Carlo simulation based tracking method, similar to the Probabilistic Index of Connectivity (PICo) approach, was extended to preserve the directional information of the main fibre bundles passing a voxel. By combining two of these extended visiting maps arising from different seed points, two independent parameters are determined for each voxel: the first quantifies the uncertainty that a voxel is connected to both seed points; the second represents the directional information and estimates the proportion of fibres running in the direction of the other seed point (connecting fibre) or face a third area (merging fibre). Both parameters are used to calculate the probability that a voxel is part of the bundle connecting both seed points. The performance and limitations of this DTI-based method are demonstrated using simulations as well as in vivo measurements.

Visualization of tissue velocity data from cardiac wall motion measurements with myocardial fiber tracking: principles and implications for cardiac fiber structures

OBJECTIVE

The spatial arrangement of myocardial fiber structure affects the mechanical and electrical properties of the heart. Therefore, information on the structure and dynamics of the orientation of the muscle fibers in the human heart might provide significant insight into principles of the mechanics of normal ventricular contraction and electrical propagation and may subsequently aid pre- and postsurgical evaluation of patients. Fiber orientation is inherently linked to cardiac wall motion, which can be measured with phase contrast magnetic resonance imaging (MRI), also termed tissue phase mapping (TPM).

METHODS

This study provides initial results of the visualization of velocity data with fiber tracking algorithms and discusses implications for the fiber orientations. In order to generate datasets with sufficient volume coverage and resolution TPM measurements with three-dimensional (3D) velocity encoding were executed during breath-hold periods and free breathing. Subsequent postprocessing evaluation with a tracking algorithm for acceleration fields derived from the velocity data was performed.

RESULTS

Myocardial acceleration tracking illustrated the dynamics of fiber structure during four different phases of left ventricular performance, that include isovolumetric contraction (IVC), mid-systole, isovolumetric relaxation (IVR), and mid-diastole. Exact reconstruction of the myocardial fiber structure from velocity data requires mathematical modeling of spatiotemporal evolution of the velocity fields.

CONCLUSIONS

'Acceleration fibers' were reconstructed at these four phases during the cardiac cycle, and these findings may become (a) surrogate parameters in the normal ventricle, (b) baseline markers for subsequent clinical studies of abnormal hearts with altered architecture, and (c) may help to explain and illustrate functional features of cardiac performance in structural models like the helical ventricular myocardial band.

Multitensor approach for analysis and tracking of complex fiber configurations

A multidiffusion-tensor model (MDT) is presented containing two anisotropic and one isotropic diffusion tensors. This approach has the ability to detect areas of fiber crossings and resolve the direction of crossing fibers. The mean diffusivity and the ratio of the tensor compartments were merged to one independent parameter by fitting MDT to the diffusion-weighted intensities of a two-point data acquisition scheme. By an F-test between the errors of the standard single diffusion tensor and the more complex MDT, fiber crossings were detected and the more accurate model was chosen voxel by voxel. The performance of crossing detection was compared with the spherical harmonics approach in simulations as well as in vivo. Similar results were found in both methods. The MDT model, however, did not only detect crossings but also yielded the single fiber directions. The FACT algorithm and a probabilistic connectivity algorithm were extended to support the MDT model. For example, a mean angular error smaller than 10 degrees was found for the MDT model in a simulated fiber crossing with an SNR of 80. By tracking the corticospinal tract the MDT-based tracks reached a significantly greater area of the gyrus precentralis.