| 1. |
- Björnfot, Cecilia, et al.
(författare)
-
Assessing cerebral arterial pulse wave velocity using 4D flow MRI
- 2021
-
Ingår i: Journal of Cerebral Blood Flow and Metabolism. - : Sage Publications. - 0271-678X .- 1559-7016. ; 41:10, s. 2769-2777
-
Tidskriftsartikel (refereegranskat)abstract
- Intracranial arterial stiffening is a potential early marker of emerging cerebrovascular dysfunction and could be mechanistically involved in disease processes detrimental to brain function via several pathways. A prominent consequence of arterial wall stiffening is the increased velocity at which the systolic pressure pulse wave propagates through the vasculature. Previous non-invasive measurements of the pulse wave propagation have been performed on the aorta or extracranial arteries with results linking increased pulse wave velocity to brain pathology. However, there is a lack of intracranial “target-organ” measurements. Here we present a 4D flow MRI method to estimate pulse wave velocity in the intracranial vascular tree. The method utilizes the full detectable branching structure of the cerebral vascular tree in an optimization framework that exploits small temporal shifts that exists between waveforms sampled at varying depths in the vasculature. The method is shown to be stable in an internal consistency test, and of sufficient sensitivity to robustly detect age-related increases in intracranial pulse wave velocity.
|
|
| 2. |
- Björnfot, Cecilia, et al.
(författare)
-
Cerebral arterial stiffness is linked to white matter hyperintensities and perivascular spaces in older adults : a 4D flow MRI study
- 2024
-
Ingår i: Journal of Cerebral Blood Flow and Metabolism. - : Sage Publications. - 0271-678X .- 1559-7016. ; 44:8, s. 1343-1351
-
Tidskriftsartikel (refereegranskat)abstract
- White matter hyperintensities (WMH), perivascular spaces (PVS) and lacunes are common MRI features of small vessel disease (SVD). However, no shared underlying pathological mechanism has been identified. We investigated whether SVD burden, in terms of WMH, PVS and lacune status, was related to changes in the cerebral arterial wall by applying global cerebral pulse wave velocity (gcPWV) measurements, a newly described marker of cerebral vascular stiffness. In a population-based cohort of 190 individuals, 66–85 years old, SVD features were estimated from T1-weighted and FLAIR images while gcPWV was estimated from 4D flow MRI data. Additionally, the gcPWV’s stability to variations in field-of-view was analyzed. The gcPWV was 10.82 (3.94) m/s and displayed a significant correlation to WMH and white matter PVS volume (r = 0.29, p < 0.001; r = 0.21, p = 0.004 respectively from nonparametric tests) that persisted after adjusting for age, blood pressure variables, body mass index, ApoB/A1 ratio, smoking as well as cerebral pulsatility index, a previously suggested early marker of SVD. The gcPWV displayed satisfactory stability to field-of-view variations. Our results suggest that SVD is accompanied by changes in the cerebral arterial wall that can be captured by considering the velocity of the pulse wave transmission through the cerebral arterial network.
|
|
| 3. |
- Vikner, Tomas, et al.
(författare)
-
Blood-brain barrier integrity is linked to cognitive function, but not to cerebral arterial pulsatility, among elderly
- 2024
-
Ingår i: Scientific Reports. - : Springer Nature. - 2045-2322. ; 14:1
-
Tidskriftsartikel (refereegranskat)abstract
- Blood-brain barrier (BBB) disruption may contribute to cognitive decline, but questions remain whether this association is more pronounced for certain brain regions, such as the hippocampus, or represents a whole-brain mechanism. Further, whether human BBB leakage is triggered by excessive vascular pulsatility, as suggested by animal studies, remains unknown. In a prospective cohort (N = 50; 68-84 years), we used contrast-enhanced MRI to estimate the permeability-surface area product (PS) and fractional plasma volume ( formula presented ), and 4D flow MRI to assess cerebral arterial pulsatility. Cognition was assessed by the Montreal Cognitive Assessment (MoCA) score. We hypothesized that high PS would be associated with high arterial pulsatility, and that links to cognition would be specific to hippocampal PS. For 15 brain regions, PS ranged from 0.38 to 0.85 (·10-3 min-1) and formula presented from 0.79 to 1.78%. Cognition was related to PS (·10-3 min-1) in hippocampus (β = - 2.9; p = 0.006), basal ganglia (β = - 2.3; p = 0.04), white matter (β = - 2.6; p = 0.04), whole-brain (β = - 2.7; p = 0.04) and borderline-related for cortex (β = - 2.7; p = 0.076). Pulsatility was unrelated to PS for all regions (p > 0.19). Our findings suggest PS-cognition links mainly reflect a whole-brain phenomenon with only slightly more pronounced links for the hippocampus, and provide no evidence of excessive pulsatility as a trigger of BBB disruption.
|
|
| 4. |
- Adjeiwaah, Mary, 1980-, et al.
(författare)
-
Sensitivity analysis of different quality assurance methods for magnetic resonance imaging in radiotherapy
- 2020
-
Ingår i: Physics and Imaging in Radiation Oncology. - : Elsevier. - 2405-6316. ; 13, s. 21-27
-
Tidskriftsartikel (refereegranskat)abstract
- Background and purpose: There are currently no standard quality assurance (QA) methods for magnetic resonance imaging (MRI) in radiotherapy (RT). This work was aimed at evaluating the ability of two QA protocols to detect common events that affect quality of MR images under RT settings.Materials and methods: The American College of Radiology (ACR) MRI QA phantom was repeatedly scanned using a flexible coil and action limits for key image quality parameters were derived. Using an exploratory survey, issues that reduce MR image quality were identified. The most commonly occurring events were introduced as provocations to produce MR images with degraded quality. From these images, detection sensitivities of the ACR MRI QA protocol and a commercial geometric accuracy phantom were determined.Results: Machine-specific action limits for key image quality parameters set at mean±3σ" role="presentation" style="box-sizing: border-box; margin: 0px; padding: 0px; display: inline-block; line-height: normal; font-size: 16.2px; word-spacing: normal; overflow-wrap: normal; white-space: nowrap; float: none; direction: ltr; max-width: none; max-height: none; min-width: 0px; min-height: 0px; border: 0px; position: relative;">mean±3σ were comparable with the ACR acceptable values. For the geometric accuracy phantom, provocations from uncorrected gradient nonlinearity effects and a piece of metal in the bore of the scanner resulted in worst distortions of 22.2 mm and 3.4 mm, respectively. The ACR phantom was sensitive to uncorrected signal variations, electric interference and a piece of metal in the bore of the scanner but could not adequately detect individual coil element failures.Conclusions: The ACR MRI QA phantom combined with the large field-of-view commercial geometric accuracy phantom were generally sensitive in identifying some common MR image quality issues. The two protocols when combined may provide a tool to monitor the performance of MRI systems in the radiotherapy environment.
|
|
| 5. |
- Frankel, Jennifer, 1981-
(författare)
-
Characterization of the MRI patient exposure environment and exposure assessment methods for magnetic fields in MRI scanners
- 2021
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- Magnetic resonance imaging (MRI) has become one of the most common imaging modalities available in modern medicine, and it is an indispensable diagnostic tool thanks to the unparalleled soft-tissue contrast and high image resolution. It is also a unique exposure environment consisting of a complex mix of magnetic fields. During an MRI scan, the patient is simultaneously exposed to a strong static magnetic field, a fast-switching gradient magnetic field, and a pulsed radiofrequency (RF) magnetic field. Transient acute effects, such as nerve excitation and tissue heating, are well known and limited by universal safety guidelines. Long-term health effects related to MRI exposure have, however, not been scientifically established, and no interaction mechanisms have been verified, despite a growing body of research on electromagnetic field exposure. Further epidemiological and experimental research on MRI exposure has been recommended but the lack of a common definition of dose or exposure metric makes evaluation of past research and the design of future experiments difficult.The objectives of this thesis were to characterize the MRI patient exposure environment in terms of the magnetic fields involved, suggest relevant exposure metrics, and introduce exposure assessment methods suitable for epidemiological and experimental research on MRI and long-term health effects.In Paper I, we discussed the MRI exposure environment and its complexity and gave an overview of the current scientific situation. In Paper II, we investigated the exposure variability between different MRI sequences and suggested patient-independent exposure metrics that describe different characteristics of the magnetic field exposure, including mean, peak, and threshold values. In Paper III, we presented three exposure assessment methods, specifically suited to the complex MRI exposure environment: a measurement-based method, a calculation-based method, and a proxy method.Papers I and II showed that MRI exams are not homogenous in terms of exposure, and exposure variability exists between the individual sequences that comprise an exam. Differences in mean exposure between sequences were several-fold, peak exposure differences up to 30-fold, and differences in threshold exposure were in some cases more than 100-fold. Furthermore, within-sequence exposure variability, related to the parameter adjustments that can be made at the scanner console before the start of a scan, gave rise to 5-to-8-fold exposure increases. Paper III showed that magnetic field models could be used to approximate the exposure at arbitrary locations inside the scanner, with slight underestimation of gradient field metrics and large variability in some RF field metrics. With improvements in accuracy and efficiency, the method could become a useful exposure assessment tool for in vitro and in vivo research as well as clinical work on medical implant safety. Our search for suitable exposure metric proxies resulted in a limited selection with low correlation between proxies and their counterpart metrics, but, with further development, the proxy method has the potential to allow for much needed exposure classification relevant to large-scale epidemiological research.The work in this thesis has contributed to increased awareness of the unique MRI exposure environment, the characteristics of the magnetic fields involved, and the inherent exposure variability in MRI exams. The metrics and methods presented are specifically suited to exposure assessment of the unique MRI environment, and may contribute to improved research quality by allowing for meaningful comparisons between study results and for experimental conditions to be easily replicated in future studies.
|
|
| 6. |
- Frankel, Jennifer, et al.
(författare)
-
Measurements of the switched gradient magnetic field in MRI : A closer look at some unintuitive spatial characteristics
- 2023
-
Ingår i: iRADIOLOGY. - : John Wiley & Sons. - 2834-2860 .- 2834-2879. ; 1:4, s. 390-396
-
Tidskriftsartikel (refereegranskat)abstract
- Concomitant fields are the unwanted transverse components that arise when spatial encoding gradients are applied in MRI. We measured the changing gradient magnetic field at multiple locations inside the scanner and examined the internal distribution and linearity of the three vector components of the field. Our results illustrate some not-so-obvious spatial characteristics of the gradient field, which can seem unintuitive at first glance, but are quite reasonable when considering electromagnetic theory and MRI-scanner physics constraints.
|
|
| 7. |
- Hellström, Max, et al.
(författare)
-
Denoising and uncertainty estimation in parameter mapping with approximate Bayesian deep image priors
- 2023
-
Ingår i: Magnetic Resonance in Medicine. - : John Wiley & Sons. - 0740-3194 .- 1522-2594. ; 90:6, s. 2557-2571
-
Tidskriftsartikel (refereegranskat)abstract
- Purpose: To mitigate the problem of noisy parameter maps with high uncertainties by casting parameter mapping as a denoising task based on Deep Image Priors.Methods: We extend the concept of denoising with Deep Image Prior (DIP) into parameter mapping by treating the output of an image-generating network as a parametrization of tissue parameter maps. The method implicitly denoises the parameter mapping process by filtering low-level image features with an untrained convolutional neural network (CNN). Our implementation includes uncertainty estimation from Bernoulli approximate variational inference, implemented with MC dropout, which provides model uncertainty in each voxel of the denoised parameter maps. The method is modular, so the specifics of different applications (e.g., T1 mapping) separate into application-specific signal equation blocks. We evaluate the method on variable flip angle T1 mapping, multi-echo T2 mapping, and apparent diffusion coefficient mapping.Results: We found that deep image prior adapts successfully to several applications in parameter mapping. In all evaluations, the method produces noise-reduced parameter maps with decreased uncertainty compared to conventional methods. The downsides of the proposed method are the long computational time and the introduction of some bias from the denoising prior.Conclusion: DIP successfully denoise the parameter mapping process and applies to several applications with limited hyperparameter tuning. Further, it is easy to implement since DIP methods do not use network training data. Although time-consuming, uncertainty information from MC dropout makes the method more robust and provides useful information when properly calibrated.
|
|
| 8. |
- Löfstedt, Tommy, et al.
(författare)
-
Bayesian non-linear regression with spatial priors for noise reduction and error estimation in quantitative MRI with an application in T1 estimation
- 2020
-
Ingår i: Physics in Medicine and Biology. - : Institute of Physics (IOP). - 0031-9155 .- 1361-6560. ; 65:22
-
Tidskriftsartikel (refereegranskat)abstract
- Purpose. To develop a method that can reduce and estimate uncertainty in quantitative MR parameter maps without the need for hand-tuning of any hyperparameters.Methods. We present an estimation method where uncertainties are reduced by incorporating information on spatial correlations between neighbouring voxels. The method is based on a Bayesian hierarchical non-linear regression model, where the parameters of interest are sampled, using Markov chain Monte Carlo (MCMC), from a high-dimensional posterior distribution with a spatial prior. The degree to which the prior affects the model is determined by an automatic hyperparameter search using an information criterion and is, therefore, free from manual user-dependent tuning. The samples obtained further provide a convenient means to obtain uncertainties in both voxels and regions. The developed method was evaluated on T1 estimations based on the variable flip angle method.Results. The proposed method delivers noise-reduced T1 parameter maps with associated error estimates by combining MCMC sampling, the widely applicable information criterion, and total variation-based denoising. The proposed method results in an overall decrease in estimation error when compared to conventional voxel-wise maximum likelihood estimation. However, this comes with an increased bias in some regions, predominately at tissue interfaces, as well as an increase in computational time.Conclusions. This study provides a method that generates more precise estimates compared to the conventional method, without incorporating user subjectivity, and with the added benefit of uncertainty estimation.
|
|
| 9. |
- Simkó, Attila, et al.
(författare)
-
Changing the Contrast of Magnetic Resonance Imaging Signals using Deep Learning
- 2021
-
Ingår i: Proceedings of the Fourth Conference on Medical Imaging with Deep Learning, PMLR. - : Lübeck University; Hamburg University of Technology. ; , s. 713-727
-
Konferensbidrag (refereegranskat)abstract
- The contrast settings to select before acquiring magnetic resonance imaging (MRI) signal depend heavily on the subsequent tasks. As each contrast highlights different tissues, automated segmentation tools for example might be optimized for a certain contrast. While for radiotherapy, multiple scans of the same region with different contrasts can achieve a better accuracy for delineating tumours and organs at risk. Unfortunately, the optimal contrast for the subsequent automated methods might not be known during the time of signal acquisition, and performing multiple scans with different contrasts increases the total examination time and registering the sequences introduces extra work and potential errors. Building on the recent achievements of deep learning in medical applications, the presented work describes a novel approach for transferring any contrast to any other. The novel model architecture incorporates the signal equation for spin echo sequences, and hence the model inherently learns the unknown quantitative maps for proton density, ?1 and ?2 relaxation times (??, ?1 and ?2, respectively). This grants the model the ability to retrospectively reconstruct spin echo sequences by changing the contrast settings Echo and Repetition Time (?? and ??, respectively). The model learns to identify the contrast of pelvic MR images, therefore no paired data of the same anatomy from different contrasts is required for training. This means that the experiments are easily reproducible with other contrasts or other patient anatomies. Despite the contrast of the input image, the model achieves accurate results for reconstructing signal with contrasts available for evaluation. For the same anatomy, the quantitative maps are consistent for a range of contrasts of input images. Realized in practice, the proposed method would greatly simplify the modern radiotherapy pipeline. The trained model is made public together with a tool for testing the model on example images.
|
|
| 10. |
- Simkó, Attila, 1995-
(författare)
-
Contributions to deep learning for imaging in radiotherapy
- 2023
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- Purpose: The increasing importance of medical imaging in cancer treatment, combined with the growing popularity of deep learning gave relevance to the presented contributions to deep learning solutions with applications in medical imaging.Relevance: The projects aim to improve the efficiency of MRI for automated tasks related to radiotherapy, building on recent advancements in the field of deep learning.Approach: Our implementations are built on recently developed deep learning methodologies, while introducing novel approaches in the main aspects of deep learning, with regards to physics-informed augmentations and network architectures, and implicit loss functions. To make future comparisons easier, we often evaluated our methods on public datasets, and made all solutions publicly available.Results: The results of the collected projects include the development of robust models for MRI bias field correction, artefact removal, contrast transfer and sCT generation. Furthermore, the projects stress the importance of reproducibility in deep learning research and offer guidelines for creating transparent and usable code repositories.Conclusions: Our results collectively build the position of deep learning in the field of medical imaging. The projects offer solutions that are both novel and aim to be highly applicable, while emphasizing generalization towards a wide variety of data and the transparency of the results.
|
|
| 11. |
- Simkó, Attila, et al.
(författare)
-
Improving MR image quality with a multi-task model, using convolutional losses
- 2023
-
Ingår i: BMC Medical Imaging. - : BioMed Central (BMC). - 1471-2342. ; 23:1
-
Tidskriftsartikel (refereegranskat)abstract
- PURPOSE: During the acquisition of MRI data, patient-, sequence-, or hardware-related factors can introduce artefacts that degrade image quality. Four of the most significant tasks for improving MRI image quality have been bias field correction, super-resolution, motion-, and noise correction. Machine learning has achieved outstanding results in improving MR image quality for these tasks individually, yet multi-task methods are rarely explored.METHODS: In this study, we developed a model to simultaneously correct for all four aforementioned artefacts using multi-task learning. Two different datasets were collected, one consisting of brain scans while the other pelvic scans, which were used to train separate models, implementing their corresponding artefact augmentations. Additionally, we explored a novel loss function that does not only aim to reconstruct the individual pixel values, but also the image gradients, to produce sharper, more realistic results. The difference between the evaluated methods was tested for significance using a Friedman test of equivalence followed by a Nemenyi post-hoc test.RESULTS: Our proposed model generally outperformed other commonly-used correction methods for individual artefacts, consistently achieving equal or superior results in at least one of the evaluation metrics. For images with multiple simultaneous artefacts, we show that the performance of using a combination of models, trained to correct individual artefacts depends heavily on the order that they were applied. This is not an issue for our proposed multi-task model. The model trained using our novel convolutional loss function always outperformed the model trained with a mean squared error loss, when evaluated using Visual Information Fidelity, a quality metric connected to perceptual quality.CONCLUSION: We trained two models for multi-task MRI artefact correction of brain, and pelvic scans. We used a novel loss function that significantly improves the image quality of the outputs over using mean squared error. The approach performs well on real world data, and it provides insight into which artefacts it detects and corrects for. Our proposed model and source code were made publicly available.
|
|
| 12. |
- Simkó, Attila, et al.
(författare)
-
MRI bias field correction with an implicitly trained CNN
- 2022
-
Ingår i: Proceedings of the 5th international conference on medical imaging with deep learning. - : ML Research Press. ; , s. 1125-1138
-
Konferensbidrag (refereegranskat)abstract
- In magnetic resonance imaging (MRI), bias fields are difficult to correct since they are inherently unknown. They cause intra-volume intensity inhomogeneities which limit the performance of subsequent automatic medical imaging tasks, \eg, tissue-based segmentation. Since the ground truth is unavailable, training a supervised machine learning solution requires approximating the bias fields, which limits the resulting method. We introduce implicit training which sidesteps the inherent lack of data and allows the training of machine learning solutions without ground truth. We describe how training a model implicitly for bias field correction allows using non-medical data for training, achieving a highly generalized model. The implicit approach was compared to a more traditional training based on medical data. Both models were compared to an optimized N4ITK method, with evaluations on six datasets. The implicitly trained model improved the homogeneity of all encountered medical data, and it generalized better for a range of anatomies, than the model trained traditionally. The model achieves a significant speed-up over an optimized N4ITK method—by a factor of 100, and after training, it also requires no parameters to tune. For tasks such as bias field correction - where ground truth is generally not available, but the characteristics of the corruption are known - implicit training promises to be a fruitful alternative for highly generalized solutions.
|
|
| 13. |
- Simkó, Attila, et al.
(författare)
-
Reproducibility of the methods in medical imaging with deep learning
- 2023
-
Ingår i: Medical imaging with deep learning 2023. - : ML Research Press. ; , s. 95-106
-
Konferensbidrag (refereegranskat)abstract
- Concerns about the reproducibility of deep learning research are more prominent than ever, with no clear solution in sight. The Medical Imaging with Deep Learning (MIDL) conference has made advancements in employing empirical rigor with regards to reproducibility by advocating open access, and recently also recommending authors to make their code public—both aspects being adopted by the majority of the conference submissions. We have evaluated all accepted full paper submissions to MIDL between 2018 and 2022 using established, but adjusted guidelines addressing the reproducibility and quality of the public repositories. The evaluations show that publishing repositories and using public datasets are becoming more popular, which helps traceability, but the quality of the repositories shows room for improvement in every aspect. Merely 22% of all submissions contain a repository that was deemed repeatable using our evaluations. From the commonly encountered issues during the evaluations, we propose a set of guidelines for machine learning-related research for medical imaging applications, adjusted specifically for future submissions to MIDL. We presented our results to future MIDL authors who were eager to continue an open discussion on the topic of code reproducibility.
|
|
| 14. |
- Simkó, Attila, et al.
(författare)
-
Towards MR contrast independent synthetic CT generation
- 2024
-
Ingår i: Zeitschrift für Medizinische Physik. - : Elsevier. - 0939-3889 .- 1876-4436. ; 34:2, s. 270-277
-
Tidskriftsartikel (refereegranskat)abstract
- The use of synthetic CT (sCT) in the radiotherapy workflow would reduce costs and scan time while removing the uncertainties around working with both MR and CT modalities. The performance of deep learning (DL) solutions for sCT generation is steadily increasing, however most proposed methods were trained and validated on private datasets of a single contrast from a single scanner. Such solutions might not perform equally well on other datasets, limiting their general usability and therefore value. Additionally, functional evaluations of sCTs such as dosimetric comparisons with CT-based dose calculations better show the impact of the methods, but the evaluations are more labor intensive than pixel-wise metrics.To improve the generalization of an sCT model, we propose to incorporate a pre-trained DL model to pre-process the input MR images by generating artificial proton density, T1 and T2 maps (i.e. contrast-independent quantitative maps), which are then used for sCT generation. Using a dataset of only T2w MR images, the robustness towards input MR contrasts of this approach is compared to a model that was trained using the MR images directly. We evaluate the generated sCTs using pixel-wise metrics and calculating mean radiological depths, as an approximation of the mean delivered dose. On T2w images acquired with the same settings as the training dataset, there was no significant difference between the performance of the models. However, when evaluated on T1w images, and a wide range of other contrasts and scanners from both public and private datasets, our approach outperforms the baseline model. Using a dataset of T2w MR images, our proposed model implements synthetic quantitative maps to generate sCT images, improving the generalization towards other contrasts. Our code and trained models are publicly available.
|
|
| 15. |
- Vu, Minh Hoang, 1988-
(författare)
-
Resource efficient automatic segmentation of medical images
- 2023
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- Cancer is one of the leading causes of death worldwide. In 2020, there were around 10 million cancer deaths and nearly 20 million new cancer cases in the world. Radiation therapy is essential in cancer treatments because half of the cancer patients receive radiation therapy at some point. During a radiotherapy treatment planning (RTP), an oncologist must manually outline two types of areas of the patient’s body: target, which will be treated, and organs-at-risks (OARs), which are essential to avoid. This step is called delineation. The purpose of the delineation is to generate a sufficient dose plan that can provide adequate radiation dose to a tumor and limit the radiation exposure to healthy tissue. Therefore, accurate delineations are essential to achieve this goal.Delineation is tedious and demanding for oncologists because it requires hours of concentrating work doing a repeated job. This is a RTP bottleneck which is often time- and resource-intensive. Current software, such as atlasbased techniques, can assist with this procedure by registering the patient’s anatomy to a predetermined anatomical map. However, the atlas-based methods are often slowed down and erroneous for patients with abnormal anatomies.In recent years, deep learning (DL) methods, particularly convolutional neural networks (CNNs), have led to breakthroughs in numerous medical imaging applications. The core benefits of CNNs are weight sharing and that they can automatically detect important visual features. A typical application of CNNs for medical images is to automatically segment tumors, organs, and structures, which is assumed to save radiation oncologists much time when delineating. This thesis contributes to resource efficient automatic segmentation and covers different aspects of resource efficiency.In Paper I, we proposed a novel end-to-end cascaded network for semantic segmentation in brain tumors in the multi-modal magnetic resonance imaging challenge in 2019. The proposed method used the hierarchical structure of the tumor sub-regions and was one of the top-ranking teams in the task of quantification of uncertainty in segmentation. A follow-up work to this paper was ranked second in the same task in the same challenge a year later.We systematically assessed the segmentation performance and computational costs of the technique called pseudo-3D as a function of the number of input slices in Paper II. We compared the results to typical two-dimensional (2D) and three-dimensional (3D) CNNs and a method called triplanar orthogonal 2D. The typical pseudo-3D approach considers adjacent slices to be several image input channels. We discovered that a substantial benefit from employing multiple input slices was apparent for a specific input size.We introduced a novel loss function in Paper III to address diverse issues, including imbalanced datasets, partially labeled data, and incremental learning. The proposed loss function adjusts to the given data to use all accessible data, even if some lack annotations. We show that the suggested loss function also performs well in an incremental learning context, where an existing model can be modified to incorporate the delineations of newly appearing organs semi-automatically.In Paper IV, we proposed a novel method for compressing high-dimensional activation maps, which are the primary source of memory use in modern systems. We examined three distinct compression methods for the activation maps to accomplishing this. We demonstrated that the proposed method induces a regularization effect that acts on the layer weight gradients. By employing the proposed technique, we reduced activation map memory usage by up to 95%.We investigated the use of generative adversarial networks (GANs) to enlarge a small dataset by generating synthetic images in Paper V. We use the real and generated data during training CNNs for the downstream segmentation tasks. Inspired by an existing GAN, we proposed a conditional version to generate high-dimensional and high-quality medical images of different modalities and their corresponding label maps. We evaluated the quality of the generated medical images and the effect of this augmentation on the performance of the segmentation task on six datasets.
|
|
| 16. |
- Wang, Jianfeng, 1984-, et al.
(författare)
-
Contrast Agent Quantification by Using Spatial Information in Dynamic Contrast Enhanced MRI
- 2021
-
Ingår i: Frontiers in Signal Processing. - : Frontiers Media S.A.. - 2673-8198. ; 1
-
Tidskriftsartikel (refereegranskat)abstract
- The purpose of this work is to investigate spatial statistical modelling approaches to improve contrast agent quantification in dynamic contrast enhanced MRI, by utilising the spatial dependence among image voxels. Bayesian hierarchical models (BHMs), such as Besag model and Leroux model, were studied using simulated MRI data. The models were built on smaller images where spatial dependence can be incorporated, and then extended to larger images using the maximum a posteriori (MAP) method. Notable improvements on contrast agent concentration estimation were obtained for both smaller and larger images. For smaller images: the BHMs provided substantial improved estimates in terms of the root mean squared error (rMSE), compared to the estimates from the existing method for a noise level equivalent of a 12-channel head coil at 3T. Moreover, Leroux model outperformed Besag models with two different dependence structures. Specifically, the Besag models increased the estimation precision by 27% around the peak of the dynamic curve, while the Leroux model improved the estimation by 40% at the peak, compared with the existing estimation method. For larger images: the proposed MAP estimators showed clear improvements on rMSE for vessels, tumor rim and white matter.
|
|
