Zein Salah

Live-Wire Revisited

The live-wire approach is an interactive, contour-based segmentation technique. Generally, the contour of a targeted object (anatomical structure) is built by interactively selecting control points and finding minimal-cost paths between them. By its very nature, this method is applicable only to 2D images. For the segmentation of 3D datasets (volumes), the interactive generation of live-wire contours has to be applied to each slice of the volume. This process can be quite tedious, due to the sometimes intensive user interaction. In this contribution, we propose adaptive propagation as an alternative to individually processing all image slices or shape-based interpolation of live-wire contours.

Preoperative planning of a complete mastoidectomy: semiautomatic segmentation and evaluation

AbstractObjective Preoperative planning of complete mastoidectomy based on postprocessing, interactive processing, and visualization of CT datasets was implemented and tested. The synthesized images are used to define an access path for the operative treatment of diseases at the lateral skull base. Alternatively, the mastoid may be resected to allocate space for implants such as hearing aids. Methods For the preoperative planning of such surgery, accurate segmentation of the mastoid is required. A semiautomatic pipeline for segmentation of the mastoid was developed. The approach requires minimal user interaction, using a method based on generating a fast 3D-oriented primary segmentation, which (if necessary) can be refined interactively, both in 3D- and slice-based manner. Results To verify the usability of the segmentation and visualization tools for surgery planning, an evaluation study was conducted with datasets of 16 operated patients. For each patient, at least on preoperative CT scan of the skull, with focus on the region of the mastoid to be operated, was available. Qualitative and quantitative evaluation of the images was performed, and the results were tabulated. Conclusion Semiautomated mastoid segmentation provides a feasible means for preoperative planning, based on postprocessing analysis and visualization of CT datasets.

A Fast and Accurate Approach for the Segmentation of the Paranasal Sinus

Recently, functional endoscopic sinus surgery became state of the art in the surgical treatment of endonasal pathology. For a more accurate access planning including 3D measures of the cavity, especially in complex cases of tumor diseases, segmentation of the paranasal sinus can be very useful. Unfortunately, this structure is quite complicated and difficult to segment. In our contribution, we propose a semi-automatic segmentation pipeline that significantly reduces the total processing time and the required interaction.

Data reconstruction and visualization techniques for forensic pathology

Forensic pathology is largely concerned with the determination of the cause and manner of deaths after accidents, or other circumstances in criminal investigations. A major task in that process is the documentation of surface injuries, which is traditionally done by drawing sketches, photography, or more recently by photogrammetry to generate a three-dimensional digital lesion cartography of the body surface. In this paper, we describe a semi-automatic processing pipeline how data from 3D photogrammetry is combined and used to generate a visual surface representation of accident victims. In that course, a number of steps are performed to provide a high-quality interactive, point-based visualization of the acquired data, which can be used in a more routine way than previous forensic surface methods.

Segmenting the Mastoid: Allocating Space in the Head for a Hearing Aid Implantation

We present a procedure for the segmentation of the mastoid — a sponge-like bone structure behind the middle ear — from CT Images. This segmentation is a first pre-operative step for a robot-supported hearing aid implantation. Up to now, the mastoid is milled out by an ENT surgeon to allocate space for the hearing aid. This tedious milling process of the mastoid is intended to be performed by a robot to reduce the required time and improve the accuracy.

Anwendung eines semi-automatischen Algorithmus zur Segmentierung des Mastoid für die OP-Planung an der lateralen Schädelbasis

Bildgestutzte 3-dimensionale Operationsplanungen fur die laterale Schadelbasis beruhen derzeit auf einer manuellen Bearbeitung von CT-Schichtdaten. Diese ist sehr zeitaufwandig und erfordert eine hohe Anwenderexpertise. In der vorgestellten Studie habe wir einen neu entwickelten semiautomatischen Algorithmus zur Segmentierung einer Felsenbeinaushohlung in CT-Datensatzen in einer ersten klinischen Anwendung auf Robustheit, Ergebnisqualitat und Zeiteffizienz untersucht und fur einen Einsatz in der Klinik optimiert. Alle Datensatze waren mit dem Algorithmus problemlos zu bearbeiten. Der Zeitaufwand fur ein qualitativ hochwertiges Ergebnis konnte im Vergleich zum bisherigen Vorgehen auf ein Drittel reduziert werden. Der semiautomatische Algorithmus scheint damit gut geeignet, eine Vereinfachung und Beschleunigung des OP-Planungsprozesses zu ermoglichen.