This paper presents a comparison study between 10 automatic and six interactive methods for liver segmentation from contrast-enhanced CT images. It is based on results from the "MICCAI 2007 Grand Challenge" workshop, where 16 teams evaluated their algorithms on a common database. A collection of 20 clinical images with reference segmentations was provided to train and tune algorithms in advance. Participants were also allowed to use additional proprietary training data for that purpose. All teams then had to apply their methods to 10 test datasets and submit the obtained results. Employed algorithms include statistical shape models, atlas registration, level-sets, graph-cuts and rule-based systems. All results were compared to reference segmentations five error measures that highlight different aspects of segmentation accuracy. All measures were combined according to a specific scoring system relating the obtained values to human expert variability. In general, interactive methods reached higher average scores than automatic approaches and featured a better consistency of segmentation quality. However, the best automatic methods (mainly based on statistical shape models with some additional free deformation) could compete well on the majority of test images. The study provides an insight in performance of different segmentation approaches under real-world conditions and highlights achievements and limitations of current image analysis techniques.

No abstract

IntroductionWhile Cumulus – a semi-automated method for measuring breast density – is utilised extensively in research, it is labour-intensive and unsuitable for screening programmes that require an efficient and valid measure on which to base screening recommendations. We develop an automated method to measure breast density (AutoDensity) and compare it to Cumulus in terms of association with breast cancer risk and breast cancer screening outcomes.MethodsAutoDensity automatically identifies the breast area in the mammogram and classifies breast density in a similar way to Cumulus, through a fast, stand-alone Windows or Linux program. Our sample comprised 985 women with screen-detected cancers, 367 women with interval cancers and 4,975 controls (women who did not have cancer), sampled from first and subsequent screening rounds of a film mammography screening programme. To test the validity of AutoDensity, we compared the effect estimates using AutoDensity with those using Cumulus from logistic regression models that tested the association between breast density and breast cancer risk, risk of small and large screen-detected cancers and interval cancers, and screening programme sensitivity (the proportion of cancers that are screen-detected). As a secondary analysis, we report on correlation between AutoDensity and Cumulus measures.ResultsAutoDensity performed similarly to Cumulus in all associations tested. For example, using AutoDensity, the odds ratios for women in the highest decile of breast density compared to women in the lowest quintile for invasive breast cancer, interval cancers, large and small screen-detected cancers were 3.2 (95% CI 2.5 to 4.1), 4.7 (95% CI 3.0 to 7.4), 6.4 (95% CI 3.7 to 11.1) and 2.2 (95% CI 1.6 to 3.0) respectively. For Cumulus the corresponding odds ratios were: 2.4 (95% CI 1.9 to 3.1), 4.1 (95% CI 2.6 to 6.3), 6.6 (95% CI 3.7 to 11.7) and 1.3 (95% CI 0.9 to 1.8). Correlation between Cumulus and AutoDensity measures was 0.63 (P < 0.001).ConclusionsBased on the similarity of the effect estimates for AutoDensity and Cumulus in models of breast density and breast cancer risk and screening outcomes, we conclude that AutoDensity is a valid automated method for measuring breast density from digitised film mammograms.

A computer-aided detection (CAD) system is presented for the localization of interstitial lesions in chest radiographs. The system analyzes the complete lung fields using a two-class supervised pattern classification approach to distinguish between normal texture and texture affected by interstitial lung disease. Analysis is done pixel-wise and produces a probability map for an image where each pixel in the lung fields is assigned a probability of being abnormal. Interstitial lesions are often subtle and ill defined on x-rays and hence difficult to detect, even for expert radiologists. Therefore a new, semiautomatic method is proposed for setting a reference standard for training and evaluating the CAD system. The proposed method employs the fact that interstitial lesions are more distinct on a computed tomography (CT) scan than on a radiograph. Lesion outlines, manually drawn on coronal slices of a CT scan of the same patient, are automatically transformed to corresponding outlines on the chest x-ray, using manually indicated correspondences for a small set of anatomical landmarks. For the texture analysis, local structures are described by means of the multiscale Gaussian filter bank. The system performance is evaluated with ROC analysis on a database of digital chest radiographs containing 44 abnormal and 8 normal cases. The best performance is achieved for the linear discriminant and support vector machine classifiers, with an area under the ROC curve (A(z)) of 0.78. Separate ROC curves are built for classification of abnormalities of different degrees of subtlety versus normal class. Here the best performance in terms of A(z) is 0.90 for differentiation between obviously abnormal and normal pixels. The system is compared with two human observers, an expert chest radiologist and a chest radiologist in training, on evaluation of regions. Each lung field is divided in four regions, and the reference standard and the probability maps are converted into region scores. The system performance does not significantly differ from that of the observers, when the perihilar regions are excluded from evaluation, and reaches A(z) = 0.85 for the system, with A(z) = 0.88 for both observers.

The version in the Kent Academic Repository may differ from the final published version. Users are advised to check http://kar.kent.ac.uk for the status of the paper. Users should always cite the published version of record.