No abstract

We present a novel method to extract iso-surfaces from distance volumes. It generates high quality semi-regular multiresolution meshes of arbitrary topology. Our technique proceeds in two stages. First, a very coarse mesh with guaranteed topology is extracted. Subsequently an iterative multi-scale force-based solver refines the initial mesh into a semi-regular mesh with geometrically adaptive sampling rate and good aspect ratio triangles. The coarse mesh extraction is performed using a new approach we call surface wavefront propagation. A set of discrete iso-distance ribbons are rapidly built and connected while respecting the topology of the iso-surface implied by the data. Subsequent multi-scale refinement is driven by a simple force-based solver designed to combine good iso-surface fit and high quality sampling through reparameterization. In contrast to the Marching Cubes technique our output meshes adapt gracefully to the iso-surface geometry, have a natural multiresolution structure and good aspect ratio triangles, as demonstrated with a number of examples.

We introduce a new algorithm for fitting a Catmull-Clark subdivision surface to a given shape within a prescribed tolerance, based on the method of quasi-interpolation. The fitting algorithm is fast, local and scales well since it does not require the solution of linear systems. Its convergence rate is optimal for regular meshes and our experiments show that it behaves very well for irregular meshes. We demonstrate the power and versatility of our method with examples from interactive modeling, surface fitting, and scientific visualization.

We evaluated the clinical significance of 18F-FDG PET to detect malignant ovarian neoplasms and tumor spread. 40 patients (median age: 57.5 years) underwent laparotomy because of clinical suspicion of malignant ovarian tumors or recurrent disease. The results of the preoperatively performed PET were correlated with the postoperative histologic diagnosis and the intraoperatively assessed tumor spread. In 10 of 40 patients benign tumors were found, among which a tubo-ovarian abscess was the only one diagnosed as false positive. 4/30 malignant neoplasms did not originate from the coelomic epithelium, but all were correctly recognized as malignant tumors by PET, as was recurrent ovarian cancer in 12 patients. Out of 14 primary ovarian carcinomas, 2 borderline tumors and 1 well-differentiated adenocarcinoma FIGO stage I were not correctly identified. Considering the tumor type, sensitivity, and specificity were 90%, calculating for the positive and negative predictive value 96% and 75%, respectively, and 90% for the diagnostic accuracy. Those statistical parameters were slightly lower for PET detection of lymph node metastasis and peritoneal carcinomatosis. Although its diagnostic accuracy may vary depending on the clinical application, 18F-FDG PET is basically a suitable method for detecting ovarian malignancies, particularly in patients with relapsed ovarian carcinoma.

Figure 1: Collision detection in a thin-shell simulation. Given a control mesh of a bent tube, our algorithm identifies interfering regions on the limit surface up to a user-specified precision. Above, we have resolved interference with progressively increasing precision. AbstractSubdivision surfaces are an attractive representation when modeling arbitrary topology free-form surfaces and show great promise for applications in engineering design [5,6] and computer animation [10]. Interference detection is a critical tool in many of these applications. In this paper we derive normal bounds for subdivision surfaces and use these to develop an efficient algorithm for (self-) interference detection.