In this paper we report the set-up and results of the Multimodal Brain Tumor Image Segmentation Benchmark (BRATS) organized in conjunction with the MICCAI 2012 and 2013 conferences Twenty state-of-the-art tumor segmentation algorithms were applied to a set of 65 multi-contrast MR scans of low- and high-grade glioma patients—manually annotated by up to four raters—and to 65 comparable scans generated using tumor image simulation software Quantitative evaluations revealed considerable disagreement between the human raters in segmenting various tumor sub-regions (Dice scores in the range 74%–85%), illustrating the difficulty of this task We found that different algorithms worked best for different sub-regions (reaching performance comparable to human inter-rater variability), but that no single algorithm ranked in the top for all sub-regions simultaneously Fusing several good algorithms using a hierarchical majority vote yielded segmentations that consistently ranked above all individual algorithms, indicating remaining opportunities for further methodological improvements The BRATS image data and manual annotations continue to be publicly available through an online evaluation system as an ongoing benchmarking resource

/pdf/the-multimodal-brain-tumor-image-segmentation-benchmark-2gfraj98dk.pdf

The Multimodal Brain Tumor Image Segmentation Benchmark (BRATS)

The Visual Display of Quantitative Information

With an estimated 160,000 deaths in 2018, lung cancer is the most common cause of cancer death in the United States1. Lung cancer screening using low-dose computed tomography has been shown to reduce mortality by 20–43% and is now included in US screening guidelines1–6. Existing challenges include inter-grader variability and high false-positive and false-negative rates7–10. We propose a deep learning algorithm that uses a patient’s current and prior computed tomography volumes to predict the risk of lung cancer. Our model achieves a state-of-the-art performance (94.4% area under the curve) on 6,716 National Lung Cancer Screening Trial cases, and performs similarly on an independent clinical validation set of 1,139 cases. We conducted two reader studies. When prior computed tomography imaging was not available, our model outperformed all six radiologists with absolute reductions of 11% in false positives and 5% in false negatives. Where prior computed tomography imaging was available, the model performance was on-par with the same radiologists. This creates an opportunity to optimize the screening process via computer assistance and automation. While the vast majority of patients remain unscreened, we show the potential for deep learning models to increase the accuracy, consistency and adoption of lung cancer screening worldwide. A convolutional neural network performs automated prediction of malignancy risk of pulmonary nodules in chest CT scan volumes and improves accuracy of lung cancer screening.

End-to-end lung cancer screening with three-dimensional deep learning on low-dose chest computed tomography

Image enhancement plays an important role in image processing and analysis. Among various enhancement algorithms, Retinex-based algorithms can efficiently enhance details and have been widely adopted. Since Retinex-based algorithms regard illumination removal as a default preference and fail to limit the range of reflectance, the naturalness of non-uniform illumination images cannot be effectively preserved. However, naturalness is essential for image enhancement to achieve pleasing perceptual quality. In order to preserve naturalness while enhancing details, we propose an enhancement algorithm for non-uniform illumination images. In general, this paper makes the following three major contributions. First, a lightness-order-error measure is proposed to access naturalness preservation objectively. Second, a bright-pass filter is proposed to decompose an image into reflectance and illumination, which, respectively, determine the details and the naturalness of the image. Third, we propose a bi-log transformation, which is utilized to map the illumination to make a balance between details and naturalness. Experimental results demonstrate that the proposed algorithm can not only enhance the details but also preserve the naturalness for non-uniform illumination images.

Naturalness Preserved Enhancement Algorithm for Non-Uniform Illumination Images

Validation, comparison, and combination of algorithms for automatic detection of pulmonary nodules in computed tomography images: The LUNA16 challenge.

A system is disclosed for quantitative analysis of perfusion images comprising image elements having intensity values associated therewith. The system comprises a frequency distribution computing subsystem (1) for computing a plurality of frequency distributions of the intensity values of at least part of the images. The system comprises a perfusion information extractor (2) for extracting information relating to perfusion from the plurality of frequency distributions. The perfusion information extractor (2) comprises a shift detector (3) for detecting a shift of the intensity values of the frequency distribution. The perfusion information extractor (2) is arranged for extracting the information relating to perfusion, based on the detected shift. A user interface element (8) enables a user to indicate a boundary between the core region and the rim region by a single degree of freedom. A vesselness subsystem (9) associates a vesselness value with an image element.

Quantitative perfusion analysis

A method for reformatting image data includes obtaining volumetric image data indicative of an anatomical structure of interest, identifying a surface of interest of the anatomical structure of interest in the volumetric image data, identifying a thickness for a sub-volume of interest of the volumetric image data, shaping the sub-volume of interest such that at least one of its sides follows the surface of interest, and generating, via a processor, a maximum intensity projection (MIP) or direct volume rendering (DVR) based on the identified surface of interest and the shaped sub-volume of interest.

Image data reformatting

An apparatus produces image space data (35) indicative of the spatially varying strength of the vascular connections between locations in the image space and a lesion or other feature of interest. The data may be presented by way of a maximum intensity projection (MIP) display in which the brightness of the image represents the strength of the vascular connection.

Visualization of vascularization

A system (20) an image aquisition device, an image workstation, computer- readable medium and method for visualizing an anatomical tree structure is disclosed. The anatomical tree structure is segmented from a three-dimensional image set of at least a portion of a body. A planar angular visualization of the anatomical tree structure is then determined. The planar angular visualization of the anatomical tree structure is then displayed. Since the displayed anatomical tree structure has no occlusions or intersections of segments, the anatomical tree is much easier to read which facilitates subsequent automated or manual diagnosis of the tree for anomalies.

Planar angular visualization of the bronchial tree

To compare two-dimensional (2D) and three-dimensional (3D) computed tomography (CT) measurements of ablation zones (AZs) related to the shaft of two different large-volume monopolar multi-tined expandable electrodes. Percutaneous radiofrequency (RF) ablation was performed in 12 pigs (81.6 ± 7.8 kg) using two electrodes (LeVeen 5 cm, Rita XL 5 cm; n = 6 in each group). Contrast-enhanced CT with the electrode shaft in place evaluated the AZ. The largest sphere centred on the electrode shaft within the AZ was calculated (1) based on the 2D axial CT image in the plane of the shaft assuming rotational symmetry of the AZ and (2) using prototype software and the 3D volume data of the AZ measured with CT. The mean largest diameter of a sphere centred on the electrode shaft was always smaller using the 3D data of the AZ than using 2D CT measurements assuming rotational symmetry of the AZ (3D vs 2D): LeVeen 18.2 ± 4.8 mm; 24.5 ± 3.1 mm; p = 0.001; Rita XL 20.0 ± 3.7 mm; 28.8 ± 4.9 mm; p = 0.0002. All AZ showed indentations around the tines. Two-dimensional CT measurements assuming rotational symmetry of the AZ overestimate the largest ablated sphere centred on the electrode shaft compared with 3D CT measurements.

Rafael Wiemker

Papers

Quantitative perfusion analysis

Image data reformatting

Visualization of vascularization

Planar angular visualization of the bronchial tree

Large-volume multi-tined expandable RF ablation in pig livers: comparison of 2D and volumetric measurements of the ablation zone