| 1. |
- Nysjö, Fredrik, 1985-
(författare)
-
Modeling and Visualization for Virtual Interaction with Medical Image Data
- 2020
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- Interactive systems for exploring and analysing medical three dimensional (3D) volume image data using techniques such as stereoscopic rendering and haptics can lead to new workflows for virtual surgery planning. This includes the design of patient-specific surgical guides and plates for additive manufacturing (3D printing). Our applications, medical visualization and cranio-maxillofacial surgery planning, involve large volume data such as computed tomo\-graphy (CT) images with millions of data points. This motivates the development of fast and efficient methods for visualization and haptic rendering, as well as the development of efficient modeling techniques for simplifying the design of 3D printable parts. In this thesis, we develop methods for visualization and haptic rendering of isosurfaces in volume image data, and show applications of these methods to medical visualization and virtual surgery planning. We further develop methods for modeling surgical guides and plates for cranio-maxillofacial surgery, and integrate them into our system for haptics-assisted surgery planning called HASP. This system is now installed at the department of surgical sciences, Uppsala University, and is being evaluated for use in clinical research.
|
|
| 2. |
- Bengtsson Bernander, Karl
(författare)
-
Equivariant Neural Networks for Biomedical Image Analysis
- 2024
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- While artificial intelligence and deep learning have revolutionized many fields in the last decade, one of the key drivers has been access to data. This is especially true in biomedical image analysis where expert annotated data is hard to come by. The combination of Convolutional Neural Networks (CNNs) with data augmentation has proven successful in increasing the amount of training data at the cost of overfitting. In this thesis, equivariant neural networks have been used to extend the equivariant properties of CNNs to more transformations than translations. The networks have been trained and evaluated on biomedical image datasets, including bright-field microscopy images of cytological samples indicating oral cancer, and transmission electron microscopy images of virus samples. By designing the networks to be equivariant to e.g. rotations, it is shown that the need for data augmentation is reduced, that less overfitting occurs, and that convergence during training is faster. Furthermore, equivariant neural networks are more data efficient than CNNs, as demonstrated by scaling laws. These benefits are not present in all problem settings and which benefits will occur is somewhat unpredictable. We have identified that the results to some extent depend on architectures, hyperparameters and datasets. Further research may broaden the performed studies to explain how the results occur with new theory.
|
|
| 3. |
- Nedelcu, Robert
(författare)
-
In Vivo Accuracy and Precision in Prosthodontics
- 2020
-
Doktorsavhandling (övrigt vetenskapligt/konstnärligt)abstract
- Background: There has been a dramatic increase in commercially available intraoral scanners (IOS) in the last decade, offering to replace indirect digitization of models (MOD) fabricated from impressions (IMPR). IOS has benefits of less patient discomfort and a faster workflow to fabricate fixed dental prosthesis (FDP), and implant-supported prostheses (IFD). However, in vivo evidence is lacking not only for IOS, but also for MOD, FDP and IFD fit.Aims: Paper I: to evaluate in vitro finish line distinction and accuracy in seven IOS and one MOD. To assess parameters of resolution, tessellation, topography, and color. Paper II: to evaluate a method of acquiring an in vivo reference measurement in dentate subjects and analyse accuracy and precision of IOS and MOD. Paper III: to evaluate an in vivo reference-measurement method in fully edentulous maxillae with full-arch implant treatments and to analyse accuracy of MOD and fit of existing IFD. Paper IV: to analyse precision and accuracy of IOS using different acquisition protocols compared to the reference-measurement in Paper III.Material and Methods: Paper I: A model with a crown preparation was reference-scanned with an industrial scanner, (ATOS), scanned with seven IOS and the MOD of an IMPR was digitized. Best-fit Alignment and 3D Compare Analysis was followed by descriptive analysis. Paper II: A reference-scan was acquired with ATOS. Subjects were scanned with IOS and one MOD of an IMPR was digitized. Accuracy and precision were evaluated after Best-Fit Alignment and 3D Compare Analysis. Paper III: A reference-measurement of implant positions was acquired with ATOS. MOD from IMPR was digitized and IFD scanned. Datum and Relative Point System Alignment was followed by accuracy and precision analysis. Paper IV: Subjects in Paper III were scanned with IOS using three different protocols, followed by accuracy and precision analysis.Results: Paper I: There were considerable differences between IOS depiction of finish line and finish line accuracy. Paper II: IOS presented varying results for impressions in up to ten units. No differences were found for MOD. Paper III: IFD was significantly less accurate than MOD. Paper IV: Differences were found between scanning protocols. Compared to Paper III, IFD was less accurate. No differences were found for MOD.Conclusion: There are relevant differences between IOS when scanning subgingival preparations. Some IOS are better suited for long-span scans. Some IOS can be used for full-arch impressions for IFD in the maxilla, however, adequate soft-tissue management is crucial.
|
|
| 4. |
|
|
