Discriminative generalized Hough transform for object localization in medical images.

PURPOSE: This paper proposes the discriminative generalized Hough transform (DGHT) as an efficient and reliable means for object localization in medical images. It is meant to give a deeper insight into the underlying theory and a comprehensive overview of the methodology and the scope of applications. METHODS: The DGHT combines the generalized Hough transform (GHT) with a discriminative training technique for the GHT models to obtain more efficient and robust localization results. To this end, the model points are equipped with individual weights, which are trained discriminatively with respect to a minimal localization error. Through this weighting, the models become more robust since the training focuses on common features of the target object over a set of training images. Unlike other weighting strategies, our training algorithm focuses on the error rate and allows for negative weights, which can be employed to encode rivaling structures into the model. The basic algorithm is presented here in conjunction with several extensions for fully automatic and faster processing. These include: (1) the automatic generation of models from training images and their iterative refinement, (2) the training of joint models for similar objects, and (3) a multi-level approach. RESULTS: The algorithm is tested successfully for the knee in long-leg radiographs (97.6 % success rate), the vertebrae in C-arm CT (95.5 % success rate), and the femoral head in whole-body MR (100 % success rate). In addition, it is compared to Hough forests (Gall et al. in IEEE Trans Pattern Anal Mach Intell 33(11):2188-2202, 2011) for the task of knee localization (97.8 % success rate). Conclusion The DGHT has proven to be a general procedure, which can be easily applied to various tasks with high success rates.

Tensor grid based image registration with application to ventilation estimation on 4D CT lung data.

PURPOSE: For many image registration tasks, the information contained in the original resolution of the image data is crucial for a subsequent medical analysis, e.g. accurate assessment of local pulmonary ventilation. However, the complexity of a non-parametric registration scheme is directly connected to the resolution of the images. Therefore, the registration is often performed on a downsampled version in order to meet runtime demands and thereby producing suboptimal results. To enable the application of the highest resolution at least in regions of high clinical importance, an approach is presented replacing the usually taken equidistant grids by tensor grids for image representation. METHODS: We employ a non-parametric approach for the registration of a respiratory-gated 4D CT thorax scan. Tensor grids are introduced for the registration setting and compared to equidistant grids. For ventilation assessment, the Jacobian metric is explored. RESULTS: The application of the tensor grid approach makes the local usage of the original resolution feasible; thereby a smaller registration error is achieved in regions of higher resolution using the tensor grids, while the two types of grids perform similar in regions of equal resolution. Concerning the ventilation assessment, the Jacobian metric yields reasonable results, showing more detail using the tensor grids due to the higher resolution. CONCLUSIONS: The proposed approach using tensor grids preserves registration accuracy, while reducing computational demands. The application of the Jacobian metric for ventilation assessment in conjunction with tensor grids is promising; however, due to a missing ground-truth the medical relevance could not be established for the ventilation estimation so far.