The Deformable Feature Map — Adaptive Plasticity for Function Approximation

Wismüller, Axel; Vietze, Frank; Dersch, Dominik R.; Hahn, Klaus; Ritter, Helge

doi:10.1007/978-1-4471-1599-1_14

Cited by 4 publications

(9 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A common choice would be the ordering of the r j 's on a regular low-dimensional grid (k ∈ {2, 3}). However, this is not mandatory (see, e.g., [11]). By choosing the metric of the cooperativity function a j according to the neuron positions r j in the target space, we obtain Kohonen's self-organizing map (SOM) algorithm [12,13].…”

Section: Optimization Of the W J : Vector Quantizationmentioning

confidence: 97%

See 1 more Smart Citation

Segmentation with Neural Networks

Wismueller

2009

Handbook of Medical Image Processing and Analysis

View full text Add to dashboard Cite

Section: Optimization Of the W J : Vector Quantizationmentioning

confidence: 97%

“…In [39,40], the author presented an algorithm that provides a reverse, alternative approach to adaptive function…”

Section: The Deformable Feature Mapmentioning

confidence: 99%

Segmentation with Neural Networks

Wismueller

2009

Handbook of Medical Image Processing and Analysis

View full text Add to dashboard Cite

“…In [61,62], the author has presented an algorithm that provides a reverse, alternative approach to adaptive function approximation: the Deformable Feature Map. Here, the output weights of a GRBF network are kept constant, whereas the re-training is performed at the VQ level using an unsupervised method.…”

Section: The Deformable Feature Mapmentioning

confidence: 99%

“…The segmentation of Y (c) served as a reference data set for a fully automatic segmentation of X, shown in (d). From [61] ing vectors, representing reference points for the definition of a mapping S : X → Y, x → y. Between these reference points, S has to be determined by some appropriate interpolation or approximation procedure.…”

Section: The Deformable Feature Mapmentioning

confidence: 99%

“…In addition, its application to the real-world problem of automatic nonlinear multispectral image registration, employing magnetic resonance imaging data sets of the human brain has been described in this paper, whereas the publications [61,[64][65][66] refer to the application of the DM algorithm to multispectral image segmentation, see Fig. 9.…”

Section: The Deformable Feature Mapmentioning

confidence: 99%

See 1 more Smart Citation

Computational intelligence in biomedical imaging: multidimensional analysis of spatio-temporal patterns

Wismüller

2010

Comput Sci Res Dev

Self Cite

View full text Add to dashboard Cite

Technical innovations in radiology, such as advanced cross-sectional imaging methods, have opened up new vistas for the exploration of structure and function of the human body enabling both high spatial and temporal resolution. However, these techniques have led to vast amounts of data whose precise and reliable visual analysis by radiologists requires a considerable amount of human intervention and expertise, thus resulting in a cost factor of substantial economic relevance. Hence, the computer-assisted analysis of biomedical image data has moved into the focus of interest as an issue of high priority research efforts. In this context, innovative approaches to exploratory analysis of huge complex spatio-temporal patterns play a key role to improve computer-assisted signal and image processing in radiology. Examples of such approaches are various unsupervised vector quantization methods or supervised function approximation techniques, such as Generalized RadialBasis-Functions-(GRBF-) neural networks. Recent developments motivated by concepts of computational intelligence are the 'Deformable Feature Map' (DM) as an algorithm for self-organized model adaptation, the 'Mutual Connectivity Analysis' (MCA) as an instrument for the analysis of large time-series ensembles and the 'Exploratory Observation Machine' (XOM) as a novel general framework for learning by self-organization-three methods that the author has invented and applied to biomedical real-world applications. This contribution covers both conceptual foundations A. Wismüller ( )

show abstract

Introducing anisotropic Minkowski functionals and quantitative anisotropy measures for local structure analysis in biomedical imaging

Wismüller

Lochmüller

et al. 2013

SPIE Proceedings

Self Cite

View full text Add to dashboard Cite

The ability of Minkowski Functionals to characterize local structure in different biological tissue types has been demonstrated in a variety of medical image processing tasks. We introduce anisotropic Minkowski Functionals (AMFs) as a novel variant that captures the inherent anisotropy of the underlying gray-level structures. To quantify the anisotropy characterized by our approach, we further introduce a method to compute a quantitative measure motivated by a technique utilized in MR diffusion tensor imaging, namely fractional anisotropy. We showcase the applicability of our method in the research context of characterizing the local structure properties of trabecular bone micro-architecture in the proximal femur as visualized on multi-detector CT. To this end, AMFs were computed locally for each pixel of ROIs extracted from the head, neck and trochanter regions. Fractional anisotropy was then used to quantify the local anisotropy of the trabecular structures found in these ROIs and to compare its distribution in different anatomical regions. Our results suggest a significantly greater concentration of anisotropic trabecular structures in the head and neck regions when compared to the trochanter region (p < 10−4). We also evaluated the ability of such AMFs to predict bone strength in the femoral head of proximal femur specimens obtained from 50 donors. Our results suggest that such AMFs, when used in conjunction with multi-regression models, can outperform more conventional features such as BMD in predicting failure load. We conclude that such anisotropic Minkowski Functionals can capture valuable information regarding directional attributes of local structure, which may be useful in a wide scope of biomedical imaging applications.

show abstract

The Deformable Feature Map — Adaptive Plasticity for Function Approximation

Cited by 4 publications

References 5 publications

Segmentation with Neural Networks

Segmentation with Neural Networks

Computational intelligence in biomedical imaging: multidimensional analysis of spatio-temporal patterns

Introducing anisotropic Minkowski functionals and quantitative anisotropy measures for local structure analysis in biomedical imaging

Contact Info

Product

Resources

About