Ivo Rössling

Fast and flexible distance measures for treatment planning

PurposeDistance measures are required for diagnoses, therapy decision and documentation. With today’s high-resolution CT and MR imaging techniques, high- quality images have become possible. Yet, manual measurement can be tedious. We present a method for automatically determining different distance-based measures on segmented anatomic structures, like shortest distances, diameters, and wall thicknesses.MethodsOur method is inspired from computational geometry and based on a surface mesh representation. The computation takes all primitives (points, edges, faces) into account and organizes them efficiently in a spatial tree structure. We followed the generic design paradigm in order to achieve maximum flexibility.ResultsThe generic approach allows for a variety of intervention-relevant distance measures to be computed, using only a single type of data structure. For shortest distance, our approach in empirical tests turned out to be more efficient than previous methods from medical application literature. Besides the numerical value, also its defining geometric primitives are determined.ConclusionsThe presented technique is both, fast and flexible. It can be used to interactively derive automatic distance measures for arbitrary mesh-based segmentations. Due to the geometrically exact measurements, it is possible to reliably estimate safety margins, assess possible infiltrations and other clinically relevant measures. To exploit this benefit, the method requires precise segmentations as input data.

Model-based segmentation of pathological lymph nodes in CT data

For the computer-aided diagnosis of tumor diseases knowledge about the position, size and type of the lymph nodes is needed to compute the tumor classification (TNM). For the computer-aided planning of subsequent surgeries like the Neck Dissection spatial information about the lymph nodes is also important. Thus, an efficient and exact segmentation method for lymph nodes in CT data is necessary, especially pathological altered lymph nodes play an important role here. Based on prior work, in this paper we present a noticeably enhanced model-based segmentation method for lymph nodes in CT data, which now can be used also for enlarged and mostly well separated necrotic lymph nodes. Furthermore, the kind of pathological variation can be determined automatically during segmentation, which is important for the automatic TNM classification. Our technique was tested on 21 lymph nodes from 5 CT datasets, among several enlarged and necrotic ones. The results lie in the range of the inter-personal variance of human experts and improve the results of former work again. Bigger problems were only noticed for pathological lymph nodes with vague boundaries due to infiltrated neighbor tissue.

On Design and Implementation of a Generic Number Type for Real Algebraic Number Computations Based on Expression Dags

We report on the design and implementation of a number type called Real_algebraic. This number type allows us to compute the sign of arithmetic expressions involving the operations

Distance-aware smoothing of surface meshes for surgical planning

{+,-,\cdot,/, \sqrt[d]{}}

Effiziente automatische Bestimmung interventionsrelevanter Entfernungsmaße

. The sign computation is always correct and, in this sense, not subject to rounding errors. We focus on modularity and use generic programming techniques to make key parts of the implementation exchangeable. Thus, our design allows for easily performing experiments with different implementations or thereby tailoring the number type for specific tasks. For many problems in computational geometry, instantiations of our number type Real_algebraic are a user-friendly alternative for implementing the exact geometric computation paradigm in order to abandon numerical robustness problems.

Erzeugung und Simulation eines dynamischen 3D-Modells der Kopf-Hals-Region aus CT-Daten

Besides other main criterias like size, infiltration and anatomical district, the sideness of tumor and local metastases (i. e., malignant lymph nodes) is very crucial for head and neck tumor assessment. An automatic sideness determination may speed up tumor staging noticeably. Inspired by preliminary work of others we present a modified approach for the estimation of the midsagittal plane based on surface meshes. The suitability of the computed result being used in the course of TNM classification was tested in a case study, in terms of an according automatic determination of the sideness for segmented lymph nodes and tumors.

Interaktive Visualisierung von Abständen und Ausdehnungen anatomischer Strukturen für die Interventionsplanung

The evaluation of spatial relationships between anatomic structures is a major task in surgical planning. Surface models generated from medical image data (intensity, binary) are often used for visualization and 3D measurement of extents and distances between neighboring structures. In applications for intervention or radiation treatment planning, the surface models need to exhibit a natural look (referring to smoothness of the surface), but also to be accurate. Smoothing algorithms allow to reduce artifacts from mesh generation, but the result is always a tradeoff between smoothness and accuracy. Required features will be removed and distances between adjacent structures get changed. Thus, we present a modification to common mesh smoothing algorithms, which allows to generate smooth surfaces models while distances of neighboring structures are preserved. We compared our distance-aware approach to conventional uniform smoothing methods and evaluated the resulting surface models regarding smoothness and accuracy for their application within the context of surgical planning.

Complete fully automatic model-based segmentation of normal and pathological lymph nodes in CT data

In the cycle of Algorithm Engineering, the design phase opens after the modeling phase. We may assume that the algorithmic task to be performed is well understood, i. e., that the desired input-output relation is specified, and an agreement has been reached as to what makes a solution to the problem a good solution. These questions must be settled in cooperation with representatives from fields of application.