Subscribe to the gold package and get unlimited access to Shamra Academy
Register a new userSpectral Decomposition in Deep Networks for Segmentation of Dynamic Medical Images
120
0
0.0
(
0
)
Ask ChatGPT about the research
No Arabic abstract
Dynamic contrast-enhanced magnetic resonance imaging (DCE- MRI) is a widely used multi-phase technique routinely used in clinical practice. DCE and similar datasets of dynamic medical data tend to contain redundant information on the spatial and temporal components that may not be relevant for detection of the object of interest and result in unnecessarily complex computer models with long training times that may also under-perform at test time due to the abundance of noisy heterogeneous data. This work attempts to increase the training efficacy and performance of deep networks by determining redundant information in the spatial and spectral components and show that the performance of segmentation accuracy can be maintained and potentially improved. Reported experiments include the evaluation of training/testing efficacy on a heterogeneous dataset composed of abdominal images of pediatric DCE patients, showing that drastic data reduction (higher than 80%) can preserve the dynamic information and performance of the segmentation model, while effectively suppressing noise and unwanted portion of the images.
rate research
Read More
Delineating the lesion area is an important task in image-based diagnosis. Pixel-wise classification is a popular approach to segmenting the region of interest. However, at fuzzy boundaries such methods usually result in glitches, discontinuity, or disconnection, inconsistent with the fact that lesions are solid and smooth. To overcome these undesirable artifacts, we propose the BezierSeg model which outputs bezier curves encompassing the region of interest. Directly modelling the contour with analytic equations ensures that the segmentation is connected, continuous, and the boundary is smooth. In addition, it offers sub-pixel accuracy. Without loss of accuracy, the bezier contour can be resampled and overlaid with images of any resolution. Moreover, a doctor can conveniently adjust the curves control points to refine the result. Our experiments show that the proposed method runs in real time and achieves accuracy competitive with pixel-wise segmentation models.
Medical ultrasound scanners are typically calibrated to the soft tissue average of 1540 m s$^{-1}$. In regions of different sound speed, for example, organs and tumours, the $B$-scan image then becomes a distortion of the true tissue cross-section, due to the misrepresentation of length and refraction. To quantify this distortion we develop a general geometric ray model for an object with an atypical speed of sound embedded in an ambient medium. We analyse the ensuing area distortion for circular and elliptical objects, mapping it out as a function of the key parameters, including the speed of sound mismatch, the object size and its elongation. We find that the area distortion can become significant, even for small-scale speed of sound mismatches. Our findings are verified by ultrasound imaging of a test object.
$bf{Purpose:}$ The goal of this study was (i) to use artificial intelligence to automate the traditionally labor-intensive process of manual segmentation of tumor regions in pathology slides performed by a pathologist and (ii) to validate the use of a well-known and readily available deep learning architecture. Automation will reduce the human error involved in manual delineation, increase efficiency, and result in accurate and reproducible segmentation. This advancement will alleviate the bottleneck in the workflow in clinical and research applications due to a lack of pathologist time. Our application is patient-specific microdosimetry and radiobiological modeling, which builds on the contoured pathology slides. $bf{Methods:}$ A U-Net architecture was used to segment tumor regions in pathology core biopsies of lung tissue with adenocarcinoma stained using hematoxylin and eosin. A pathologist manually contoured the tumor regions in 56 images with binary masks for training. Overlapping patch extraction with various patch sizes and image downsampling were investigated individually. Data augmentation and 8-fold cross-validation were used. $bf{Results:}$ The U-Net achieved accuracy of 0.91$pm$0.06, specificity of 0.90$pm$0.08, sensitivity of 0.92$pm$0.07, and precision of 0.8$pm$0.1. The F1/DICE score was 0.85$pm$0.07, with a segmentation time of 3.24$pm$0.03 seconds per image, achieving a 370$pm$3 times increased efficiency over manual segmentation. In some cases, the U-Net correctly delineated the tumors stroma from its epithelial component in regions that were classified as tumor by the pathologist. $bf{Conclusion:}$ The U-Net architecture can segment images with a level of efficiency and accuracy that makes it suitable for tumor segmentation of histopathological images in fields such as radiotherapy dosimetry, specifically in the subfields of microdosimetry.
The emergence of multi-parametric magnetic resonance imaging (mpMRI) has had a profound impact on the diagnosis of prostate cancers (PCa), which is the most prevalent malignancy in males in the western world, enabling a better selection of patients for confirmation biopsy. However, analyzing these images is complex even for experts, hence opening an opportunity for computer-aided diagnosis systems to seize. This paper proposes a fully automatic system based on Deep Learning that takes a prostate mpMRI from a PCa-suspect patient and, by leveraging the Retina U-Net detection framework, locates PCa lesions, segments them, and predicts their most likely Gleason grade group (GGG). It uses 490 mpMRIs for training/validation, and 75 patients for testing from two different datasets: ProstateX and IVO (Valencia Oncology Institute Foundation). In the test set, it achieves an excellent lesion-level AUC/sensitivity/specificity for the GGG$geq$2 significance criterion of 0.96/1.00/0.79 for the ProstateX dataset, and 0.95/1.00/0.80 for the IVO dataset. Evaluated at a patient level, the results are 0.87/1.00/0.375 in ProstateX, and 0.91/1.00/0.762 in IVO. Furthermore, on the online ProstateX grand challenge, the model obtained an AUC of 0.85 (0.87 when trained only on the ProstateX data, tying up with the original winner of the challenge). For expert comparison, IVO radiologists PI-RADS 4 sensitivity/specificity were 0.88/0.56 at a lesion level, and 0.85/0.58 at a patient level. Additional subsystems for automatic prostate zonal segmentation and mpMRI non-rigid sequence registration were also employed to produce the final fully automated system. The code for the ProstateX-trained system has been made openly available at https://github.com/OscarPellicer/prostate_lesion_detection. We hope that this will represent a landmark for future research to use, compare and improve upon.
Recently, state-of-the-art results have been achieved in semantic segmentation using fully convolutional networks (FCNs). Most of these networks employ encoder-decoder style architecture similar to U-Net and are trained with images and the corresponding segmentation maps as a pixel-wise classification task. Such frameworks only exploit class information by using the ground truth segmentation maps. In this paper, we propose a multi-task learning framework with the main aim of exploiting structural and spatial information along with the class information. We modify the decoder part of the FCN to exploit class information and the structural information as well. We intend to do this while also keeping the parameters of the network as low as possible. We obtain the structural information using either of the two ways: i) using the contour map and ii) using the distance map, both of which can be obtained from ground truth segmentation maps with no additional annotation costs. We also explore different ways in which distance maps can be computed and study the effects of different distance maps on the segmentation performance. We also experiment extensively on two different medical image segmentation applications: i.e i) using color fundus images for optic disc and cup segmentation and ii) using endoscopic images for polyp segmentation. Through our experiments, we report results comparable to, and in some cases performing better than the current state-of-the-art architectures and with an order of 2x reduction in the number of parameters.
suggested questions
Log in to be able to interact and post comments
comments
Fetching comments
Sign in to be able to follow your search criteria
