Semi-automatic Segmentation Technique Research Articles

Juvenile idiopathic arthritis—recent advances

Juvenile idiopathic arthritis—recent advances

Intelligent Signal and Image Processing in eHealth

Intelligent Signal and Image Processing in eHealth

DoctorEye: A Clinically Driven Multifunctional Platform, for Accurate Processing of Tumors in Medical Images

This paper presents a novel, open access interactive platform for 3D medical image analysis, simulation and visualization, focusing in oncology images. The platform was developed through constant interaction and feedback from expert clinicians integrating a thorough analysis of their requirements while having an ultimate goal of assisting in accurately delineating tumors. It allows clinicians not only to work with a large number of 3D tomographic datasets but also to efficiently annotate multiple regions of interest in the same session. Manual and semi-automatic segmentation techniques combined with integrated correction tools assist in the quick and refined delineation of tumors while different users can add different components related to oncology such as tumor growth and simulation algorithms for improving therapy planning. The platform has been tested by different users and over large number of heterogeneous tomographic datasets to ensure stability, usability, extensibility and robustness with promising results.AvailabilityThe platform, a manual and tutorial videos are available at: http://biomodeling.ics.forth.gr.It is free to use under the GNU General Public License.

Accuracy and reproducibility of a novel semi-automatic segmentation technique for MR volumetry of the pituitary gland

Although several reports about volumetric determination of the pituitary gland exist, volumetries have been solely performed by indirect measurements or manual tracing on the gland's boundaries. The purpose of this study was to evaluate the accuracy and reproducibility of a novel semi-automatic MR-based segmentation technique. In an initial technical investigation, T1-weighted 3D native magnetised prepared rapid gradient echo sequences (1.5T) with 1mm isotropic voxel size achieved high reliability and were utilised in different in vitro and in vivo studies. The computer-assisted segmentation technique was based on an interactive watershed transform after resampling and gradient computation. Volumetry was performed by three observers with different software and neuroradiologic experiences, evaluating phantoms of known volume (0.3, 0.9 and 1.62ml) and healthy subjects (26 to 38years; overall 135 volumetries). High accuracy of the volumetry was shown by phantom analysis; measurement errors were <4% with a mean error of 2.2%. In vitro, reproducibility was also promising with intra-observer variability of 0.7% for observer 1 and 0.3% for observers 2 and 3; mean inter-observer variability was in vitro 1.2%. In vivo, scan-rescan, intra-observer and inter-observer variability showed mean values of 3.2%, 1.8% and 3.3%, respectively. Unifactorial analysis of variance demonstrated no significant differences between pituitary volumes for various MR scans or software calculations in the healthy study groups (p > 0.05). The analysed semi-automatic MR volumetry of the pituitary gland is a valid, reliable and fast technique. Possible clinical applications are hyperplasia or atrophy of the gland in pathological circumstances either by a single assessment or by monitoring in follow-up studies.

18F-fluorocholine PET-guided target volume delineation techniques for partial prostate re-irradiation in local recurrent prostate cancer

Background and purpose We evaluate the contribution of 18F-choline PET/CT in the delineation of gross tumour volume (GTV) in local recurrent prostate cancer after initial irradiation using various PET image segmentation techniques. Materials and methods Seventeen patients with local-only recurrent prostate cancer (median = 5.7 years) after initial irradiation were included in the study. Rebiopsies were performed in 10 patients that confirmed the local recurrence. Following injection of 300 MBq of 18F-fluorocholine, dynamic PET frames (3 min each) were reconstructed from the list-mode acquisition. Five PET image segmentation techniques were used to delineate the 18F-choline-based GTVs. These included manual delineation of contours (GTV man) by two teams consisting of a radiation oncologist and a nuclear medicine physician each, a fixed threshold of 40% and 50% of the maximum signal intensity (GTV 40% and GTV 50%), signal-to-background ratio-based adaptive thresholding (GTV SBR), and a region growing (GTV RG) algorithm. Geographic mismatches between the GTVs were also assessed using overlap analysis. Results Inter-observer variability for manual delineation of GTVs was high but not statistically significant ( p = 0.459). In addition, the volumes and shapes of GTVs delineated using semi-automated techniques were significantly higher than those of GTVs defined manually. Conclusions Semi-automated segmentation techniques for 18F-choline PET-guided GTV delineation resulted in substantially higher GTVs compared to manual delineation and might replace the latter for determination of recurrent prostate cancer for partial prostate re-irradiation. The selection of the most appropriate segmentation algorithm still needs to be determined.

Trabeculae carneae as models of the ventricular walls: implications for the delivery of oxygen.

Trabeculae carneae are the smallest naturally arising collections of linearly arranged myocytes in the heart. They are the preparation of choice for studies of function of intact myocardium in vitro. In vivo, trabeculae are unique in receiving oxygen from two independent sources: the coronary circulation and the surrounding ventricular blood. Because oxygen partial pressure (PO2) in the coronary arterioles is identical in specimens from both ventricles, whereas that of ventricular blood is 2.5-fold higher in the left ventricle than in the right ventricle, trabeculae represent a “natural laboratory” in which to examine the influence of “extravascular” PO2 on the extent of capillarization of myocardial tissue. We exploit this advantage to test four hypotheses. (1) In trabeculae from either ventricle, a peripheral annulus of cells is devoid of capillaries. (2) Hence, sufficiently small trabeculae from either ventricle are totally devoid of capillaries. (3) The capillary-to-myocyte ratios in specimens from either ventricle are identical to those of their respective walls. (4) Capillary-to-myocyte ratios are comparable in specimens from either ventricle, reflecting equivalent energy demands in vivo, driven by identical contractile frequencies and comparable wall stresses. We applied confocal fluorescent imaging to trabeculae in cross section, subsequently using semi-automated segmentation techniques to distinguish capillaries from myocytes. We quantified the capillary-to-myocyte ratios of trabeculae from both ventricles and compared them to those determined for the ventricular free walls and septum. Quantitative interpretation was furthered by mathematical modeling, using both the classical solution to the diffusion equation for elliptical cross sections, and a novel approach applicable to cross sections of arbitrary shape containing arbitrary disposition of capillaries and non-respiring collagen cords.

Semiautomatic tractography: motor pathway segmentation in patients with intracranial vascular malformations

Object The visualization of white matter tracts using tractography has previously been achieved by displaying streamlines that pass between regions of interest (ROIs). These techniques require a significant amount of user interaction, and their results are entirely dependent on the positioning of the ROIs. Furthermore, in patients with intracerebral hemorrhage secondary to intracranial vascular malformation, there is often significant cerebral edema and susceptibility artifact from the hematoma, which degrade the reliability of tractography. In this paper, the authors' objectives were to visualize the motor pathways of patients with hemorrhagic and nonhemorrhagic vascular malformations by using a novel semiautomated technique that functions without the need for multiple ROIs. Methods The authors investigated the tractography appearance of the descending motor pathways in 6 patients with intracranial vascular malformations. Of these patients 4 presented with a spontaneous intracranial hemorrhage, 2 of whom were clinically hemiparetic. Diffusion tensor imaging was performed using a 1.5-T clinical MR imaging system, and whole-brain tractography was performed after reconstruction of the data. A fractional anisotropy threshold of 0.05 was used to terminate the tractography. The semiautomatic motor pathway segmentation technique required definition of a single voxel within the corticospinal tract of the medulla from which the descending motor pathways were automatically defined by grouping together all streamlines within the entire image with a geometry similar to that of the single streamline generated from this initial voxel. The results of this segmentation were then visually assessed and compared with the patient's motor function. Results The authors' semiautomatic algorithm consistently visualized the location of the descending motor pathways in patients with nonhemorrhagic and hemorrhagic vascular malformations. In 1 patient whose complete right hemiplegia (complete paralysis) was caused by a large left frontal hematoma that bisected the descending motor pathways, the authors were unable to reconstruct the motor pathways due to severe tract degeneration. However, in all cases in which motor function was intact or only mildly impaired, the technique clearly delineated the motor pathways, even in the presence of large anatomical displacement by the vascular abnormality or associated hemorrhage. Conclusions Semiautomatic tractography allows consistent and rapid demonstration of the descending motor pathways in patients with hemorrhagic and nonhemorrhagic intracranial vascular malformations. The technique allows the use of a comparatively low fractional anisotropy threshold and does not require the definition of multiple ROIs. These techniques may help to improve the clinical feasibility and potentially the reliability of tractography for the evaluation of patients with intracranial vascular malformations as well as other space-occupying lesions of the brain.

Analysis of electromagnetic compatibility in biological tissue for a novel artificial anal sphincter

The electromagnetic compatibility analysis of biological tissue surrounding a transcutaneous transformer for an artificial anal sphincter (AAS) is discussed. The electromagnetic compatibility in the biological tissue surrounding the transformer was analysed by the finite integration technique method. The specific absorption rate (SAR) and current density as a function of frequency (100 k-1 MHz) for an emission current of 10 A were calculated. A high-resolution realistic human model consisting of 56 kinds of tissues or organs was constructed based on visible human project data set through a semi-automatic image segmentation technique. As a result, the SAR in the vicinity of the transformer is 10 W/kg and sufficiently small, which is divided by the basic restriction of International Commission on Non-Ionising Radiation Protection (ICNIRP). On the other hand, the current density is also far smaller than the ICNIRP s basic restrictions as the frequency rises. The result shows that the analysis of electromagnetic compatibility is very important for developing the transcutaneous transformer for an AAS system.

Liver segmentation from computed tomography scans: A survey and a new algorithm

In the recent years liver segmentation from computed tomography scans has gained a lot of importance in the field of medical image processing since it is the first and fundamental step of any automated technique for the automatic liver disease diagnosis, liver volume measurement, and 3D liver volume rendering. In this paper we report a review study about the semi-automatic and automatic liver segmentation techniques, and we describe our fully automatized method. The survey reveals that automatic liver segmentation is still an open problem since various weaknesses and drawbacks of the proposed works must still be addressed. Our gray-level based liver segmentation method has been developed to tackle all these problems; when tested on 40 patients it achieves satisfactory results, comparable to the mean intra- and inter-observer variation. We believe that our technique outperforms those presented in the literature; nevertheless, a common test set with its gold standard traced by experts, and a generally accepted performance measure are required to demonstrate it.

Motion-guided segmentation for cine DENSE MRI.

Defining myocardial contours is often the most time-consuming portion of dynamic cardiac MRI image analysis. Displacement encoding with stimulated echoes (DENSE) is a quantitative MRI technique that encodes tissue displacement into the phase of the complex MRI images. Cine DENSE provides a time series of these images, thus facilitating the non-invasive study of myocardial kinematics. Epicardial and endocardial contours need to be defined at each frame on cine DENSE images for the quantification of regional displacement and strain as a function of time. This work presents a reliable and effective two-dimensional semi-automated segmentation technique that uses the encoded motion to project a manually-defined region of interest through time. Contours can then easily be extracted for each cardiac phase. This method boasts several advantages, including, (1) parameters are based on practical physiological limits, (2) contours are calculated for the first few cardiac phases, where it is difficult to visually distinguish blood from myocardium, and (3) the method is independent of the shape of the tissue delineated and can be applied to short- or long-axis views, and on arbitrary regions of interest. Motion-guided contours were compared to manual contours for six conventional and six slice-followed mid-ventricular short-axis cine DENSE datasets. Using an area measure of segmentation error, the accuracy of the segmentation algorithm was shown to be similar to inter-observer variability. In addition, a radial segmentation error metric was introduced for short-axis data. The average radial epicardial segmentation error was 0.36+/-0.08 and 0.40+/-0.10 pixels for slice-followed and conventional cine DENSE, respectively, and the average radial endocardial segmentation error was 0.46+/-0.12 and 0.46+/-0.16 pixels for slice following and conventional cine DENSE, respectively. Motion-guided segmentation employs the displacement-encoded phase shifts intrinsic to DENSE MRI to accurately propagate a single set of pre-defined contours throughout the remaining cardiac phases.

A new method for measurement of left atrial volumes using 64-slice spiral computed tomography: Comparison with two-dimensional echocardiographic techniques

Background Left atrial (LA) volume, is related to cardiovascular morbidity. LA enlargement is usually assessed using trans-thoracic echocardiography (TTE). The association of modern multislice computed tomography (MSCT) imaging and new 3D reconstruction software, allows direct cardiac chamber volume measurement without geometrical assumptions. This study was designed to evaluate the maximal (LAmax) and minimal (LAmin) LA volumes during the cardiac cycle using MSCT and TTE approaches. Methods We screened 26 consecutive patients referred for coronary imaging using a 64-MSCT scanner and a TTE within 12 h. Contiguous multiphase images were generated from axial MSCT data and semi-automated 3D segmentation technique was applied to generate LA volumes. Using TTE, LA volumes and LA ejection fraction (LAEF) were obtained using five assumptions methods: cubing equation, diameter-length formula, area-length formula, ellipsoidal formula and biplane Simpson rule. Results Five patients were excluded for inadequate TTE visualization and one for ectopic beats during MSCT. The sample consisted in 20 patients (11 men, age: 56 ± 14 years). Using MSCT, LA volumes indexed to body surface area were: LAmax = 74 ± 27 ml/m 2, LAmin = 49 ± 26 ml/m 2, with close correlations with TTE measurements and a significant underestimation by all TTE approaches. A close correlation was observed between LAEF using MSCT and TTE Simpson's method: 36 ± 14% vs. 37 ± 14%, r = 0.99, p < 0.0001. Conclusion Theses results suggest that the assessment of LA volumes and ejection fraction was reliable using 64-MSCT in patients referred for coronary computed tomography imaging.

Axon tracing in the adult rat optic nerve and tract after intravitreal injection of MnDPDP using a semiautomatic segmentation technique

To develop and validate an objective technique for 3D segmentation of manganese-enhanced MR images of the optic nerve/tract (ON) in adult rats to improve contrast-to-noise (CNR) calculations and use the technique to ascertain if manganese dipyridoxyl diphosphate (MnDPDP) gives sufficient Mn(2+) enhancement compared to MnCl(2) when used for functional imaging of the visual pathway. Intravitreous injection of the manganese-releasing MR contrast agent MnDPDP (30 nmol Mn(2+)) was performed to trace the ON in adult rats (n = 4). A positive control group of rats (n = 5) received a standard preparation of MnCl(2) (200 nmol Mn(2+)), while gadodiamide (1500 nmol Gd(3+)) was administered in negative control rats (n = 2). An objective technique for 3D segmentation of the enhanced ON was developed. CNR profiles along the ON were calculated by resampling the 3D image-volume in 2D planes perpendicular to the Mn(2+) enhanced ON in 0.2 mm steps, 4 mm along the segmented ON measured from the lamina cribrosa. The ON was successfully segmented and CNR calculated accurately within 2 minutes in a representative 3D MR image volume. Intravitreal MnDPDP injection resulted in significant MRI contrast enhancement of the retina and ON after 12-24 hours similar to that of MnCl(2) injection. 3D semiautomated image segmentation and the use of MnDPDP can improve in vivo axon tracing based on MRI. Mn(2+) was found to be released from MnDPDP after intravitreal injection in sufficient amounts to obtain functional tracing of the adult rat primary visual pathway.

Validation of Semiautomatic Measurement of the Extent of Breast Tumors Using Contrast-Enhanced Magnetic Resonance Imaging

The aim of this study was to assess whether the semiautomatic measurement of breast-tumor extent using contrast-enhanced magnetic resonance imaging (CE-MRI) decreases measurement variation compared with manual evaluation and increases precision with respect to tumor extent measured at histopathology. Forty-three patients who underwent breast-conserving therapy for 46 tumors were consecutively included. Extended histopathology analysis was performed on the excision specimens. Two experienced breast-MR radiologists independently manually assessed the largest diameter of the tumors at preoperative CE-MRI. Two observers used a semiautomatic segmentation technique to independently obtain the largest diameter and the volume of the tumors. Semiautomatic volumetric measurement of tumor volume was more accurately correlated with histopathology (adjusted R-square 0.84 and 0.81) compared with conventional measurement (adjusted R-square 0.56 and 0.70) and semiautomatic measurement (adjusted R-square 0.63 and 0.60) of the largest diameter of the tumor. The interobserver variability between the volumetric measurements was significantly smaller than the interobserver variability between the largest-diameter measurements of tumor extent obtained by the radiologists (P < 0.001). Semiautomatic volumetric measurement of breast-tumor extent at CE-MRI provides a more accurate assessment of tumor extent with respect to histopathology and reduces measurement variation compared with manual assessment of the largest diameter. Automated volumetric measurement thus provides potential for increased precision in establishing response to treatment.

Detector characteristics of the micro crystal element scanner (MiCES)

This work reports on the performance characteristics of detector modules for the micro crystal element scanner (MiCES). MiCES is a small animal PET scanner that is designed for imaging mice. It will consist of four rings of 18 detector modules. The crystal array for each detector is a 22 /spl times/ 22 grid of 0.8 /spl times/ 0.8 /spl times/ 10 mm mixed lutetium silicate (MLS) crystals set at 0.9 mm centers and framed for rigidity to compose a 19.94 mm square. A technique to mount multi-element crystal arrays to position sensitive photomultiplier tubes (PSPMTs) is described. The crystal arrays are slightly larger than the dimensions of the anode strips of the PSPMT. Therefore the detector module assembly requires precise alignment of the crystals with the PMT optics for optimum decoding. The crystal arrays are glued to the PSPMT using Histomount (National Diagnostics) glue. It has excellent transmittance at MLS' peak emission wavelength. In addition, tapping down the crystals while the glue was curing was found to significantly improve the amount of light collected. We introduce a semi-automatic boundary segmentation technique to generate look-up-tables used for decoding the crystal of interaction from the collected PSPMT signals. The algorithm is based on valley detection followed by line segment tracking. The software tool provides a more convenient and precise way to create the crystal map for each detector module. Performance characteristics of each of the first 18 detector modules that were assembled and processed with the described techniques are presented. The overall results are an average energy resolution of 23.8% and an average peak-to-valley ratio of 7.6. These averages, however, include modules that had the crystal arrays removed and re-glued multiple times as part of our learning curve of how to optimize the gluing and positioning procedures.

Liver and Spleen Volumetry with Quantitative MR Imaging and Dual-Space Clustering Segmentation

The purpose of this HIPAA-compliant, institutional review board-approved study was to assess the liver and spleen volumes calculated by using a semiautomated dual-space clustering segmentation technique, as compared with the volumes calculated by using the manual contour-tracing method. The quantitative magnetic resonance (MR) imaging data used as input were computed from images acquired by using a mixed fast spin-echo pulse sequence that was implemented with respiratory triggering. Linear regression analysis was used to assess agreement regarding the volumes calculated by using both segmentation techniques. There was strong agreement regarding the regression parameters for the liver (r = 0.98, P < .001) and the spleen (r = 0.99, P < .001) and the mean percentage volume differences for the liver (1.2%) and the spleen (0.9%). The mean segmentation time per patient was significantly shorter with use of the dual-space clustering method (P < .001).

Voxel-based computational models of real human anatomy: a review.

Computational models of human anatomy are mathematical representations of human anatomy designed to be used in dosimetry calculations. They have been used in dosimetry calculations for radiography, radiotherapy, nuclear medicine, radiation protection and to investigate the effects of low frequency electromagnetic fields. Tomographic medical imaging techniques have allowed the construction of digital three-dimensional computational models based on the actual anatomy of individual humans. These are called voxel models, tomographic models or phantoms. Their usefulness lies in their faithful representation of human anatomy and the flexibility they afford by being able to be scaled in size to match the required human dimensions. Segmenting medical images in order to make voxel models is very time-consuming so semi-automatic segmentation techniques are being developed. Some 21 whole or partial body models currently exist and more are being prepared. These models are listed and discussed.

Evaluation of a semi-automated segmentation technique using partial differential equations

Evaluation of a semi-automated segmentation technique using partial differential equations

Computer assessment of neurodegeneration in Parkinson disease using data fusion techniques with MR images

Rationale and Objectives. Recently developed MR imaging techniques using inversion recovery are a sensitive tool to identify and quantify morphologic changes in the substantia nigra due to neurodegeneration. Using a semi-automated computer segmentation technique to isolate the substantia nigra pars compacta (SN C), we propose a colored image fusion technique to visually assess the sites of damage in the SN C and integrate the information obtained from two implemented inversion-recovery sequences. Patients and Methods. Six patients and six age-matched control subjects were scanned using a combination of two MR imaging inversion-recovery (IR) pulse sequences. A subgroup of them was used to develop our technique. Images were blended together into a final (RGBA) image, where A stands for the α channel describing transparency. Results. Abnormalities in the SN C can be accurately assessed in location, shape, and variations of signal intensities within the segmented SN C by varying the transparency (α) channel of the color fusion image. Several previous findings such as the lateral–medial gradient of signal change and a ventral–dorsal broadening of the pars compacta are accompanied by an overall mild-to-severe heterogeneity of neurodegeneration patterns. Conclusion. Color fusion techniques revealed subtle changes in the neurodegeneration of the substantia nigra in Parkinson disease, which can be helpful for an objective and hence effective visual assessment of disease progression.

Investigation of time-dependency of intracranial brain shift and its relation to the extent of tumor removal using intra-operative MRI

The object of the paper is to investigate intra-operative brainshift and its relation to the extent of tumor removal. Repeated T1w 3D datasets were acquired at different time points intra-operatively (T0; T1; T2... Tx) using a vertical open 0.5T MR scanner in six patients with intracranial tumor. An offline analysis with initial linear registration, intensity adjustment and finally nonlinear registration of the first versus subsequent time points (T0/T1; T0/T2... T0/Tx) was performed, yielding a 3D displacement vector field that describes the brainshift. Brainshift was analysed qualitatively and quantitatively. A semi-automatic segmentation technique was used for calculation of the tumor size and the size of tumor remnants. Semi-automatic segmentation was reliable in all but two cases. Segmentation was difficult and unreliable in astrocytomas grade II. The shift basically followed gravity. The major shift reached levels up to 25 mm. Significant shift was observed at the first time point (T0). Intra-operative brainshift can be analysed qualitatively and also captured quantitatively. Neuronavigation that is based on pre-operatively acquired datasets is associated with a significant risk of surgical morbidity at a very early time point. Parallelisation on a workstation cluster may reduce computation time so that information about the displacement can facilitate updated navigation.

Automating the segmentation of medical images for the production of voxel tomographic computational models.

Radiation dosimetry for the diagnostic medical imaging procedures performed on humans requires anatomically accurate, computational models. These may be constructed from medical images as voxel-based tomographic models. However, they are time consuming to produce and as a consequence, there are few available. This paper discusses the emergence of semi-automatic segmentation techniques and describes an application (iRAD) written in Microsoft Visual Basic that allows the bitmap of a medical image to be segmented interactively and semi-automatically while displayed in Microsoft Excel. iRAD will decrease the time required to construct voxel models.