Dice Similarity Coefficient Index Research Articles

Contouring aid tools in radiotherapy. Smoothing: the false friend.

Contouring accuracy is critical in modern radiotherapy. Several tools are available to assist clinicians in this task. This study aims to evaluate the performance of the smoothing tool in the ARIA system to obtain more consistent volumes. Eleven different geometric shapes were delineated in ARIA v15.6 (Sphere, Cube, Square Prism, Six-Pointed Star Prism, Arrow Prism, And Cylinder and the respective volumes at 45° of axis deviation (_45)) in 1, 3, 5, 7, and 10cm side or diameter each. Post-processing drawing tools to smooth those first-generated volumes were applied in different options (2D-ALL vs 3D) and grades (1, 3, 5, 10, 15, and 20). These volumetric transformations were analyzed by comparing different parameters: volume changes, center of mass, and DICE similarity coefficient index. Then we studied how smoothing affected two different volumes in a head and neck cancer patient: a single rounded node and the volume delineating cervical nodal areas. No changes in data were found between 2D-ALL or 3D smoothing. Minimum deviations were found (range from 0 to 0.45cm) in the center of mass. Volumes and the DICE index decreased as the degree of smoothing increased. Some discrepancies were found, especially in figures with cleft and spikes that behave differently. In the clinical case, smoothing should be applied only once throughout the target delineation process, preferably in the largest volume (PTV) to minimize errors. Smoothing is a good tool to reduce artifacts due to the manual delineation of radiotherapy volumes. The resulting volumes must be always carefully reviewed.

A super-voxel-based method for generating surrogate lung ventilation images from CT.

Purpose: This study aimed to develop and evaluate , a super-voxel-based method for surrogate computed tomography ventilation imaging (CTVI). Methods and Materials: The study used four-dimensional CT (4DCT) and single-photon emission computed tomography (SPECT) images and corresponding lung masks from 21 patients with lung cancer obtained from the Ventilation And Medical Pulmonary Image Registration Evaluation dataset. The lung volume of the exhale CT for each patient was segmented into hundreds of super-voxels using the Simple Linear Iterative Clustering (SLIC) method. These super-voxel segments were applied to the CT and SPECT images to calculate the mean density values (D mean) and mean ventilation values (Vent mean), respectively. The final CT-derived ventilation images were generated by interpolation from the D mean values to yield . For the performance evaluation, the voxel- and region-wise differences between and SPECT were compared using Spearman's correlation and the Dice similarity coefficient index. Additionally, images were generated using two deformable image registration (DIR)-based methods, and , and compared with the SPECT images. Results: The correlation between the D mean and Vent mean of the super-voxel was 0.59 ± 0.09, representing a moderate-to-high correlation at the super-voxel level. In the voxel-wise evaluation, the method achieved a stronger average correlation (0.62 ± 0.10) with SPECT, which was significantly better than the correlations achieved with the (0.33 ± 0.14, p < 0.05) and (0.23 ± 0.11, p < 0.05) methods. For the region-wise evaluation, the Dice similarity coefficient of the high functional region for (0.63 ± 0.07) was significantly higher than the corresponding values for the (0.43 ± 0.08, p < 0.05) and (0.42 ± 0.05, p < 0.05) methods. Conclusion: The strong correlation between and SPECT demonstrates the potential usefulness of this novel method of ventilation estimation for surrogate ventilation imaging.

Highly-efficient technique for automatic segmentation of X-ray bone images based on fuzzy logic and an edge detection technique

Development of medical image segmentation techniques has become one of the most important challenges in many applications that employ computers-based medical image analysis techniques. However, most of current-existing medical image segmentation techniques still have poor efficiency and complexity in calculations. A new technique for X-ray bone image segmentation has been presented in this paper. The proposed technique is designed to generate a highly-efficient quality of bone extraction from X-ray bone image at low calculation burdens. The proposed approach begins with obtaining the inverse of the original X-ray bone image then applying bounded-product operator and fuzzy controller to control the contrast of the inverse image. At the same time, an adaptive thresholding of the gradient magnitude of the original X-ray bone image is achieved to detect the edges of the bone regions. Accordingly, the bone edges are integrated with the contrasted image to give the bone segmented image. To ensure the efficient quality of the proposed algorithm, more than sixty-one of X-ray bone images were tested using much vision and statistical investigations. Then, evaluations employed several measures such as Dice similarity coefficient index, Confusion Matrix, Accuracy, Precision, Sensitivity, Specificity, and processing speed. Furthermore, results obtained using the proposed were compared to those of conventional image segmentation techniques (such as Watershed segmentation, Otsu-thresholding, K-means, and fuzzy C-means). Comparison results demonstrated the superiority of the proposed technique over other conventional techniques in both of quality and processing speed. All obtained results were obtained using MATLAB R20014a over Windows XP with processing speed 2 GHz. The high efficiency and processing speed of the proposed technique makes it such a promising solution to be implemented in many real applications.

Hyperpolarized 3He Magnetic Resonance Imaging: Comparison with Four-dimensional X-ray Computed Tomography Imaging in Lung Cancer

Pulmonary functional imaging using four-dimensional x-ray computed tomographic (4DCT) imaging and hyperpolarized (3)He magnetic resonance imaging (MRI) provides regional lung function estimates in patients with lung cancer in whom pulmonary function measurements are typically dominated by tumor burden. The aim of this study was to evaluate the quantitative spatial relationship between 4DCT and hyperpolarized (3)He MRI ventilation maps. Eleven patients with lung cancer provided written informed consent to 4DCT imaging and MRI performed within 11 ± 14 days. Hyperpolarized (3)He MRI was acquired in breath-hold after inhalation from functional residual capacity of 1 L hyperpolarized (3)He, whereas 4DCT imaging was acquired over a single tidal breath of room air. For hyperpolarized (3)He MRI, the percentage ventilated volume was generated using semiautomated segmentation; for 4DCT imaging, pulmonary function maps were generated using the correspondence between identical tissue elements at inspiratory and expiratory phases to generate percentage ventilated volume. After accounting for differences in image acquisition lung volumes ((3)He MRI: 1.9 ± 0.5 L ipsilateral, 2.3 ± 0.7 L contralateral; 4DCT imaging: 1.2 ± 0.3 L ipsilateral, 1.3 ± 0.4 L contralateral), there was no significant difference in percentage ventilated volume between hyperpolarized (3)He MRI (72 ± 11% ipsilateral, 79 ± 12% contralateral) and 4DCT imaging (74 ± 3% ipsilateral, 75 ± 4% contralateral). Spatial correspondence between 4DCT and (3)He MRI ventilation was evaluated using the Dice similarity coefficient index (ipsilateral, 86± 12%; contralateral, 88 ± 12%). Despite rather large differences in image acquisition breathing maneuvers, good spatial and significant quantitative agreement was observed for ventilation maps on hyperpolarized (3)He MRI and 4DCT imaging, suggesting that pulmonary regions with good lung function are similar between modalities in this small group of patients with lung cancer.

Systematic evaluation of three different commercial software solutions for automatic segmentation for adaptive therapy in head-and-neck, prostate and pleural cancer

PurposeTo validate, in the context of adaptive radiotherapy, three commercial software solutions for atlas-based segmentation.Methods and materialsFifteen patients, five for each group, with cancer of the Head&Neck, pleura, and prostate were enrolled in the study. In addition to the treatment planning CT (pCT) images, one replanning CT (rCT) image set was acquired for each patient during the RT course. Three experienced physicians outlined on the pCT and rCT all the volumes of interest (VOIs). We used three software solutions (VelocityAI 2.6.2 (V), MIM 5.1.1 (M) by MIMVista and ABAS 2.0 (A) by CMS-Elekta) to generate the automatic contouring on the repeated CT. All the VOIs obtained with automatic contouring (AC) were successively corrected manually. We recorded the time needed for: 1) ex novo ROIs definition on rCT; 2) generation of AC by the three software solutions; 3) manual correction of AC.To compare the quality of the volumes obtained automatically by the software and manually corrected with those drawn from scratch on rCT, we used the following indexes: overlap coefficient (DICE), sensitivity, inclusiveness index, difference in volume, and displacement differences on three axes (x, y, z) from the isocenter.ResultsThe time saved by the three software solutions for all the sites, compared to the manual contouring from scratch, is statistically significant and similar for all the three software solutions. The time saved for each site are as follows: about an hour for Head&Neck, about 40 minutes for prostate, and about 20 minutes for mesothelioma. The best DICE similarity coefficient index was obtained with the manual correction for: A (contours for prostate), A and M (contours for H&N), and M (contours for mesothelioma).ConclusionsFrom a clinical point of view, the automated contouring workflow was shown to be significantly shorter than the manual contouring process, even though manual correction of the VOIs is always needed.

Automatic Delineation of On-Line Head-And-Neck Computed Tomography Images: Toward On-Line Adaptive Radiotherapy

To develop and validate a fully automatic region-of-interest (ROI) delineation method for on-line adaptive radiotherapy. On-line adaptive radiotherapy requires a robust and automatic image segmentation method to delineate ROIs in on-line volumetric images. We have implemented an atlas-based image segmentation method to automatically delineate ROIs of head-and-neck helical computed tomography images. A total of 32 daily computed tomography images from 7 head-and-neck patients were delineated using this automatic image segmentation method. Manually drawn contours on the daily images were used as references in the evaluation of automatically delineated ROIs. Two methods were used in quantitative validation: (1) the dice similarity coefficient index, which indicates the overlapping ratio between the manually and automatically delineated ROIs; and (2) the distance transformation, which yields the distances between the manually and automatically delineated ROI surfaces. Automatic segmentation showed agreement with manual contouring. For most ROIs, the dice similarity coefficient indexes were approximately 0.8. Similarly, the distance transformation evaluation results showed that the distances between the manually and automatically delineated ROI surfaces were mostly within 3 mm. The distances between two surfaces had a mean of 1 mm and standard deviation of <2 mm in most ROIs. With atlas-based image segmentation, it is feasible to automatically delineate ROIs on the head-and-neck helical computed tomography images in on-line adaptive treatments.