| 1. |
|
|
| 2. |
|
|
| 3. |
|
|
| 4. |
- Ravikumar, Sadhana, et al.
(författare)
-
Improved Segmentation of Deep Sulci in Cortical Gray Matter Using a Deep Learning Framework Incorporating Laplace’s Equation
- 2023
-
Ingår i: Information Processing in Medical Imaging - 28th International Conference, IPMI 2023, Proceedings. - 0302-9743 .- 1611-3349. - 9783031340475 ; 13939 LNCS, s. 692-704
-
Konferensbidrag (refereegranskat)abstract
- When developing tools for automated cortical segmentation, the ability to produce topologically correct segmentations is important in order to compute geometrically valid morphometry measures. In practice, accurate cortical segmentation is challenged by image artifacts and the highly convoluted anatomy of the cortex itself. To address this, we propose a novel deep learning-based cortical segmentation method in which prior knowledge about the geometry of the cortex is incorporated into the network during the training process. We design a loss function which uses the theory of Laplace’s equation applied to the cortex to locally penalize unresolved boundaries between tightly folded sulci. Using an ex vivo MRI dataset of human medial temporal lobe specimens, we demonstrate that our approach outperforms baseline segmentation networks, both quantitatively and qualitatively.
|
|
| 5. |
- Klein, T., et al.
(författare)
-
Modeling of Multi-View 3D Freehand Radio Frequency Ultrasound
- 2012
-
Ingår i: Medical Image Computing and Computer-assisted Intervention - Miccai 2012, Pt I. - Berlin, Heidelberg : Springer. - 9783642334153 - 9783642334146 ; , s. 422-429
-
Konferensbidrag (refereegranskat)abstract
- Nowadays ultrasound (US) examinations are typically performed with conventional machines providing two dimensional imagery. However, there exist a multitude of applications where doctors could benefit from three dimensional ultrasound providing better judgment, due to the extended spatial view. 3D freehand US allows acquisition of images by means of a tracking device attached to the ultrasound transducer. Unfortunately, view dependency makes the 3D representation of ultrasound a non-trivial task. To address this we model speckle statistics, in envelope-detected radio frequency (RF) data, using a finite mixture model (FMM), assuming a parametric representation of data, in which the multiple views are treated as components of the FMM. The proposed model is show-cased with registration, using an ultrasound specific distribution based pseudo-distance, and reconstruction tasks, performed on the manifold of Gamma model parameters. Example field of application is neurology using transcranial US, as this domain requires high accuracy and data systematically features low SNR, making intensity based registration difficult. In particular, 3D US can be specifically used to improve differential diagnosis of Parkinson's disease (PD) compared to conventional approaches and is therefore of high relevance for future application.
|
|
| 6. |
- Klein, T., et al.
(författare)
-
Spatial statistics based feature descriptor for RF ultrasound data
- 2011
-
Ingår i: 2011 IEEE International Symposium on Biomedical Imaging: From Nano to Macro. - : Institute of Electrical and Electronics Engineers (IEEE). - 9781424441273
-
Konferensbidrag (refereegranskat)abstract
- In this paper we present a feature descriptor, based on a Markov random field (MRF) texture model, for radio-frequency (RF) ultrasound data. The proposed approach combines global data statistics in terms of a maximum-likelihood-estimated (MLE) distribution with local pattern characteristics employing MRF interaction parameters. This combining approach facilitates the encoding of the underlying nature of the ultrasound envelope data and therefore represents a powerful feature descriptor. Applicability and performance is showcased on RF data from a human neck.
|
|
