Optimizing the automatic segmentation of the left ventricle in magnetic resonance images.

Automatic segmentation of the left ventricular (LV) myocardial borders in cardiovascular MR (CMR) images allows a significant speed-up of the procedure of quantifying LV function, and improves its reproducibility. The automated boundary delineation is usually based on a set of parameters that define the algorithms. Since the automatic segmentation algorithms are usually sensitive to the image quality and frequently depend heavily on the acquisition protocol, optimizing the parameters of the algorithm for such different protocols may be necessary to obtain optimal results. In other words, using a default set of parameters may be far from optimal for different scanners or protocols. For the MASS-software, for example, this means that a total of 14 parameters need to be optimized. This optimization is a difficult and labor-intensive process. To be able to more consistently and rapidly tune the parameters, an automated optimization system would be extremely desirable. In this paper we propose such an approach, which is based on genetic algorithms (GAs). The GA is an unsupervised iterative tool that generates new sets of parameters and converges toward an optimal set. We implemented and compared two different types of the genetic algorithms: a simple GA (SGA) and a steady state GA (2SGA). The difference between these two algorithms lies in the characteristics of the generated populations: "nonoverlapping populations" and "overlapping populations," respectively "nonoverlapping" population means that the two populations are disjoint, and "overlapping" means that the best parameters found in the previous generation are included in the present population. The performance of both algorithms was evaluated on twenty routinely obtained short-axis examinations (eleven examinations acquired with a steady-state free precession pulse sequence, and nine examinations with a gradient echo pulse sequence). The optimal parameters obtained with the GAs were used for the LV myocardial border delineation. Finally, the automatically outlined contours were compared to the gold standard--manually drawn contours by experts. The result of the comparison was expressed as a degree of similarity after a processing time of less than 72 h to a 59.5% of degree of similarity for SGA and a 66.7% of degree of similarity for 2SGA. In conclusion, genetic algorithms are very suitable to automatically tune the parameters of a border detection algorithm. Based on our data, the 2SGA was more suitable than the SGA method. This approach can be generalized to other optimization problems in medical image processing.

Quantitative analysis of vascular images, in particular of abdominal aorta aneurysms from 3D CTA data sets.

This article presents a combination of well known image processing techniques to automatically segment CTA images of the Abdominal Aortic Aneurysm. Current results are that about 80% of the contours need no manual corrections. The remaining 20% fail due to calcified plaque close to the lumen border. After correction a 3D surface model is created from the 2D contours which is used as input for flow simulations and for parameter extraction of the AAA by clinicians for selecting the proper size and shape endograft, and to plan the placement procedure of this endograft in the patient.

Automated segmentation and analysis of vascular structures in magnetic resonance angiographic images.

The accurate assessment of the presence and extent of vascular disease, and planning of vascular interventions based on MRA requires the determination of vessel dimensions. The current standard is based on measuring vessel diameters on maximum intensity projections (MIPs) using calipers. In order to increase the accuracy and reproducibility of the method, automated analysis of the 3D MR data is required. A novel method for automatically determining the trajectory of the vessel of interest, the luminal boundaries, and subsequent the vessel dimensions is presented. The automated segmentation in 3D uses deformable models, combined with knowledge of the acquisition protocol. The trajectory determination was tested on 20 in vivo studies of the abdomen and legs. In 93% the detected trajectory followed the vessel. The luminal boundary detection was validated on contrast-enhanced (CE) MRA images of five stenotic phantoms. The results from the automated analysis correlated very well with the true diameters of the phantoms used in the in vitro study (r = 0.999, P < 0.001). MRA and x-ray angiography (XA) of the phantoms also correlated well (r = 0.895, P < 0.001). The average unsigned difference between the MRA and XA measurements was 0.08 +/- 0.05 mm. In conclusion, the automated approach allows the accurate assessment of vessel dimensions in MRA images.