Search NASA⌕ Search

SEARCH · Search NASA

Results for “Image Segmentation”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 505 records · Page 28

Summary Statistics for TA_ST01_130731: Micro-CT Data Acquired at LLNL, Specimen 1 of 3

Using x-ray micro computed tomography (MicroCT), we have characterized the linear attenuation coefficients (LAC), μ, of a sample of a dry powder material, tartaric acid (TA). The specimen was prepared at Lawrence Livermore National Laboratory (LLNL), loaded into a 60mL low density polyethylene (LDPE) bottle. After completed packing, the specimen was scanned following the protocol for MicroCT measurements under Test Plan 79. This particular specimen, TA_ST01_130731, recorded the bulk packing density (mass of sample divided by volume of sample) shown above. Two additional preparations were made and analyzed. We used the computer program IMGREC to reconstruct the CT images. The values of the key parameters used in the x-ray data capture and image reconstruction are given in this report. Additional experimental details may be found in the SOP and a separate document. To characterize the statistical distribution of LAC values in each CT image, we first isolated an ~80% region or segment of volume elements (“voxels”) lying completely within the sample, away from the walls of the container. We then calculated the mean value, standard deviation and entropy for (a) the high and low energy image segments and for (b) their digital gradient images. The statistics of the initial image of LAC values are called “first order statistics;” those of the gradient image, “second order statistics.” See Seetho for details of the analysis used to obtain the numbers reported in this document.

42 ENGINEERING↗

Summary Statistics for TA_SH02_130820: Micro-CT Data Acquired at LLNL, Specimen 2 of 3

Using x-ray micro computed tomography (MicroCT), we have characterized the linear attenuation coefficients (LAC), μ, of a sample of a dry powder material, tartaric acid (TA). The specimen was prepared at Lawrence Livermore National Laboratory (LLNL), loaded into a 60mL low density polyethylene (LDPE) bottle. After completed packing, the specimen was scanned following the protocol for MicroCT measurements under Test Plan 79. This particular specimen, TA_SH02_130820, recorded the bulk packing density (mass of sample divided by volume of sample) shown above. Two additional preparations were made and analyzed. We used the computer program IMGREC to reconstruct the CT images. The values of the key parameters used in the x-ray data capture and image reconstruction are given in this report. Additional experimental details may be found in the SOP and a separate document. To characterize the statistical distribution of LAC values in each CT image, we first isolated an ~80% region or segment of volume elements (“voxels”) lying completely within the sample, away from the walls of the container. We then calculated the mean value, standard deviation and entropy for (a) the high and low energy image segments and for (b) their digital gradient images. The statistics of the initial image of LAC values are called “first order statistics;” those of the gradient image, “second order statistics.” See Seetho for details of the analysis used to obtain the numbers reported in this document.

42 ENGINEERING↗

Summary Statistics for TA_MT03_130730: Micro-CT Data Acquired at LLNL, Specimen 3 of 3

Using x-ray micro computed tomography (MicroCT), we have characterized the linear attenuation coefficients (LAC), μ, of a sample of a dry powder material, tartaric acid (TA). The specimen was prepared at Lawrence Livermore National Laboratory (LLNL), loaded into a 60mL low density polyethylene (LDPE) bottle. After completed packing, the specimen was scanned following the protocol for MicroCT measurements under Test Plan 79. This particular specimen, TA_MT03_130730, recorded the bulk packing density (mass of sample divided by volume of sample) shown above. Two additional preparations were made and analyzed. We used the computer program IMGREC to reconstruct the CT images. The values of the key parameters used in the x-ray data capture and image reconstruction are given in this report. Additional experimental details may be found in the SOP and a separate document. To characterize the statistical distribution of LAC values in each CT image, we first isolated an ~80% region or segment of volume elements (“voxels”) lying completely within the sample, away from the walls of the container. We then calculated the mean value, standard deviation and entropy for (a) the high and low energy image segments and for (b) their digital gradient images. The statistics of the initial image of LAC values are called “first order statistics;” those of the gradient image, “second order statistics.” See Seetho for details of the analysis used to obtain the numbers reported in this document.

42 ENGINEERING↗

Summary Statistics for TA_SH01_130819: Micro-CT Data Acquired at LLNL, Specimen 1 of 3

Using x-ray micro computed tomography (MicroCT), we have characterized the linear attenuation coefficients (LAC), μ, of a sample of a dry powder material, tartaric acid (TA). The specimen was prepared at Lawrence Livermore National Laboratory (LLNL), loaded into a 60mL low density polyethylene (LDPE) bottle. After completed packing, the specimen was scanned following the protocol for MicroCT measurements under Test Plan 79. This particular specimen, TA_SH01_130819, recorded the bulk packing density (mass of sample divided by volume of sample) shown above. Two additional preparations were made and analyzed. We used the computer program IMGREC to reconstruct the CT images. The values of the key parameters used in the x-ray data capture and image reconstruction are given in this report. Additional experimental details may be found in the SOP and a separate document. To characterize the statistical distribution of LAC values in each CT image, we first isolated an ~80% region or segment of volume elements (“voxels”) lying completely within the sample, away from the walls of the container. We then calculated the mean value, standard deviation and entropy for (a) the high and low energy image segments and for (b) their digital gradient images. The statistics of the initial image of LAC values are called “first order statistics;” those of the gradient image, “second order statistics.” See Seetho for details of the analysis used to obtain the numbers reported in this document.

42 ENGINEERING↗

Summary Statistics for TA_MT01_130725: Micro-CT Data Acquired at LLNL, Specimen 1 of 3

Using x-ray micro computed tomography (MicroCT), we have characterized the linear attenuation coefficients (LAC), μ, of a sample of a dry powder material, tartaric acid (TA). The specimen was prepared at Lawrence Livermore National Laboratory (LLNL), loaded into a 60mL low density polyethylene (LDPE) bottle. After completed packing, the specimen was scanned following the protocol for MicroCT measurements under Test Plan 79. This particular specimen, TA_MT01_130725, recorded the bulk packing density (mass of sample divided by volume of sample) shown above. Two additional preparations were made and analyzed. We used the computer program IMGREC to reconstruct the CT images. The values of the key parameters used in the x-ray data capture and image reconstruction are given in this report. Additional experimental details may be found in the SOP and a separate document. To characterize the statistical distribution of LAC values in each CT image, we first isolated an ~80% region or segment of volume elements (“voxels”) lying completely within the sample, away from the walls of the container. We then calculated the mean value, standard deviation and entropy for (a) the high and low energy image segments and for (b) their digital gradient images. The statistics of the initial image of LAC values are called “first order statistics;” those of the gradient image, “second order statistics.” See Seetho for details of the analysis used to obtain the numbers reported in this document.

42 ENGINEERING↗

Summary Statistics for TA_MT02_130729: Micro-CT Data Acquired at LLNL, Specimen 2 of 3

Using x-ray micro computed tomography (MicroCT), we have characterized the linear attenuation coefficients (LAC), μ, of a sample of a dry powder material, tartaric acid (TA). The specimen was prepared at Lawrence Livermore National Laboratory (LLNL), loaded into a 60mL low density polyethylene (LDPE) bottle. After completed packing, the specimen was scanned following the protocol for MicroCT measurements under Test Plan 79. This particular specimen, TA_MT02_130729, recorded the bulk packing density (mass of sample divided by volume of sample) shown above. Two additional preparations were made and analyzed. We used the computer program IMGREC to reconstruct the CT images. The values of the key parameters used in the x-ray data capture and image reconstruction are given in this report. Additional experimental details may be found in the SOP and a separate document. To characterize the statistical distribution of LAC values in each CT image, we first isolated an ~80% region or segment of volume elements (“voxels”) lying completely within the sample, away from the walls of the container. We then calculated the mean value, standard deviation and entropy for (a) the high and low energy image segments and for (b) their digital gradient images. The statistics of the initial image of LAC values are called “first order statistics;” those of the gradient image, “second order statistics.” See Seetho for details of the analysis used to obtain the numbers reported in this document.

42 ENGINEERING↗

Summary Statistics for LC_HT02_130812: Micro-CT Data Acquired at LLNL, Specimen 2 of 3

Using x-ray micro computed tomography (MicroCT), we have characterized the linear attenuation coefficients (LAC), μ, of a sample of a dry powder material, lithium carbonate (LC). The specimen was prepared at Lawrence Livermore National Laboratory (LLNL), loaded into a 60mL low density polyethylene (LDPE) bottle. After completed packing, the specimen was scanned following the protocol for MicroCT measurements under Test Plan 79. This particular specimen, LC_HT02_130812, recorded the bulk packing density (mass of sample divided by volume of sample) shown above. Two additional preparations were made and analyzed. We used the computer program IMGREC to reconstruct the CT images. The values of the key parameters used in the x-ray data capture and image reconstruction are given in this report. Additional experimental details may be found in the SOP and a separate document. To characterize the statistical distribution of LAC values in each CT image, we first isolated an ~80% region or segment of volume elements (“voxels”) lying completely within the sample, away from the walls of the container. We then calculated the mean value, standard deviation and entropy for (a) the high and low energy image segments and for (b) their digital gradient images. The statistics of the initial image of LAC values are called “first order statistics;” those of the gradient image, “second order statistics.” See Seetho for details of the analysis used to obtain the numbers reported in this document.

42 ENGINEERING↗

Summary Statistics for LC_HT01_130808: Micro-CT Data Acquired at LLNL, Specimen 1 of 3

Using x-ray micro computed tomography (MicroCT), we have characterized the linear attenuation coefficients (LAC), μ, of a sample of a dry powder material, lithium carbonate (LC). The specimen was prepared at Lawrence Livermore National Laboratory (LLNL), loaded into a 60mL low density polyethylene (LDPE) bottle. After completed packing, the specimen was scanned following the protocol for MicroCT measurements under Test Plan 79. This particular specimen, LC_HT01_130808, recorded the bulk packing density (mass of sample divided by volume of sample) shown above. Two additional preparations were made and analyzed. We used the computer program IMGREC to reconstruct the CT images. The values of the key parameters used in the x-ray data capture and image reconstruction are given in this report. Additional experimental details may be found in the SOP and a separate document. To characterize the statistical distribution of LAC values in each CT image, we first isolated an ~80% region or segment of volume elements (“voxels”) lying completely within the sample, away from the walls of the container. We then calculated the mean value, standard deviation and entropy for (a) the high and low energy image segments and for (b) their digital gradient images. The statistics of the initial image of LAC values are called “first order statistics;” those of the gradient image, “second order statistics.” See Seetho for details of the analysis used to obtain the numbers reported in this document.

42 ENGINEERING↗

Image-based reconstruction of three-dimensional myocardial infarct geometry for patient-specific modeling of cardiac electrophysiology

Purpose: Accurate three-dimensional (3D) reconstruction of myocardial infarct geometry is crucial to patient-specific modeling of the heart aimed at providing therapeutic guidance in ischemic cardiomyopathy. However, myocardial infarct imaging is clinically performed using two-dimensional (2D) late-gadolinium enhanced cardiac magnetic resonance (LGE-CMR) techniques, and a method to build accurate 3D infarct reconstructions from the 2D LGE-CMR images has been lacking. The purpose of this study was to address this need. Methods: The authors developed a novel methodology to reconstruct 3D infarct geometry from segmented low-resolution (Lo-res) clinical LGE-CMR images. Their methodology employed the so-called logarithm of odds (LogOdds) function to implicitly represent the shape of the infarct in segmented image slices as LogOdds maps. These 2D maps were then interpolated into a 3D image, and the result transformed via the inverse of LogOdds to a binary image representing the 3D infarct geometry. To assess the efficacy of this method, the authors utilized 39 high-resolution (Hi-res) LGE-CMR images, including 36 in vivo acquisitions of human subjects with prior myocardial infarction and 3 ex vivo scans of canine hearts following coronary ligation to induce infarction. The infarct was manually segmented by trained experts in each slice of the Hi-res images, and the segmented data were downsampled to typical clinical resolution. The proposed method was then used to reconstruct 3D infarct geometry from the downsampled images, and the resulting reconstructions were compared with the manually segmented data. The method was extensively evaluated using metrics based on geometry as well as results of electrophysiological simulations of cardiac sinus rhythm and ventricular tachycardia in individual hearts. Several alternative reconstruction techniques were also implemented and compared with the proposed method. Results: The accuracy of the LogOdds method in reconstructing 3D infarct geometry, as measured by the Dice similarity coefficient, was 82.10% ± 6.58%, a significantly higher value than those of the alternative reconstruction methods. Among outcomes of electrophysiological simulations with infarct reconstructions generated by various methods, the simulation results corresponding to the LogOdds method showed the smallest deviation from those corresponding to the manual reconstructions, as measured by metrics based on both activation maps and pseudo-ECGs. Conclusions: The authors have developed a novel method for reconstructing 3D infarct geometry from segmented slices of Lo-res clinical 2D LGE-CMR images. This method outperformed alternative approaches in reproducing expert manual 3D reconstructions and in electrophysiological simulations.

60 APPLIED LIFE SCIENCES↗

Advanced Real Time Optical Imaging: Part II

The first segment of this special topic, Advanced Real-time Optical Imaging: Part I (July 2018), featured a variety of studies based on techniques including hot thermocouple, particle image velocimetry, high-speed camera, infrared imaging, and environmental confocal scanning laser microscopy. Here, the papers contributed by experts in each field re-emphasized the significance of real-time optical imaging in materials science and processing. Emerging, traditional, and mixed imaging techniques were presented describing creative and innovative methods of real-time materials investigations. Part II follows the same goal as the first segment: that advanced real-time optical imaging techniques are assembled in a single venue to emphasize the diversity and strength of these approaches.

47 OTHER INSTRUMENTATION↗

Data for Clumping Index Estimation With 30°-tilted Cameras in Row Crops: Evaluation of Methods and Segment Size Effects

The clumping index (CI) quantifies the spatial distribution of foliage elements and is essential for accurately estimating the plant area index (PAI), canopy radiative transfer, and photosynthesis. Traditionally, the finite-length averaging method (LX), the gap size distribution method (CC), and a combined approach of CC and LX (CLX) have been applied to instruments like TRAC and digital hemispherical photography to estimate CI. However, a comprehensive evaluation of these methods in row crops remains limited, especially regarding the influence of segment size on CI. Meanwhile, digital cameras offer a cost-effective and user-friendly solution for canopy measurements in row crops, yet their application in this context remains underexplored. In this study, we employed a new approach using a 30°-tilted digital camera to estimate CI in corn and soybean fields, applying the LX, CC, and CLX methods. We systematically assessed the performance of these three methods by combining field measurements in real-world fields with simulations using the LESS 3D radiative transfer model. Our results showed that CLX applied to the whole image and 45° segment offered accurate estimation of CI (bias within ±0.1, RMSE < 0.2) and PAI (bias within ±0.4, RMSE < 1) in real-world fields and LESS simulations. The accuracy of the LX method was highly sensitive to segment size, with the best performance observed at the 15° segment (PAI bias within ±0.4). In contrast, the CC method remained stable across different segment sizes, and its performance was generally comparable to that of LX, except at the 15° segment. Across view zenith angles, CI derived from CC generally showed a continuous increase, while those from LX and CLX followed a rising trend at small zenith angles but began to decline at 68°, likely due to an increasing proportion of no-gap segments. Seasonally, LX tended to show decreasing CI during early growth stages but increased as the canopy matured, whereas CC and CLX showed gradually increasing CI before plateauing at peak PAI. The 30°-tilted camera effectively captured CI variations across different angles and growth stages, making it a practical and robust instrument for row crop canopy structure analysis. Applying these CI methods to digital cameras offers a low-cost and accessible CI estimation alternative, improving canopy structure monitoring accuracy in row crops.

Modeling↗

Theoretical investigation of the design and performance of a dual energy (kV and MV) radiotherapy imager

Purpose: In modern radiotherapy treatment rooms, megavoltage (MV) portal imaging and kilovoltage (kV) cone-beam CT (CBCT) imaging are performed using various active matrix flat-panel imager (AMFPI) designs. To expand the clinical utility of MV and kV imaging, MV AMFPIs incorporating thick, segmented scintillators and, separately, kV imaging using a beam’s eye view geometry have been investigated by a number of groups. Motivated by these previous studies, it is of interest to explore to what extent it is possible to preserve the benefits of kV and MV imaging using a single AMFPI design, given the considerably different x ray energy spectra used for kV and MV imaging. In this paper, considerations for the design of such a dual energy imager are explored through examination of the performance of a variety of hypothetical AMFPIs based on x ray converters employing segmented scintillators. Methods: Contrast, noise, and contrast-to-noise ratio performances were characterized through simulation modeling of CBCT imaging, while modulation transfer function, Swank factor, and signal performance were characterized through simulation modeling of planar imaging. The simulations were based on a previously reported hybrid modeling technique (accounting for both radiation and optical effects), augmented through modeling of electronic additive noise. All designs employed BGO scintillator material with thicknesses ranging from 0.25 to 4 cm and element-to-element pitches ranging from 0.508 to 1.016 mm. A series of studies were performed under both kV and MV imaging conditions to determine the most advantageous imager configuration (involving front or rear x ray illumination and use of a mirror or black reflector), converter design (pitch and thickness), and operating mode (pitch-binning combination). Results: Under the assumptions of the present study, the most advantageous imager design was found to employ rear illumination of the converter in combination with a black reflector, incorporate a BGO converter with a 0.508 mm pitch and a 2 cm thickness, and operate at full resolution for kV imaging and 2 × 2 binning mode for MV imaging. Such a dual energy imager design should provide soft tissue visualization at low, clinically practical doses under MV conditions, while helping to preserve the high spatial resolution and high contrast offered by kV imaging. Conclusions: The authors’ theoretical investigation suggests that a dual energy imager capable of largely preserving the desirable characteristics of both kV and MV imaging is feasible. Such an imager, when coupled to a dual energy radiation source, could facilitate simplification of current treatment room imaging systems (as well as their associated quality assurance), and facilitate more precise integration of kV and MV imaging information by virtue of reduced geometric uncertainties.

07 ISOTOPE AND RADIATION SOURCES↗

Automatic segmentation of the glenohumeral cartilages from magnetic resonance images

Purpose: Magnetic resonance (MR) imaging plays a key role in investigating early degenerative disorders and traumatic injuries of the glenohumeral cartilages. Subtle morphometric and biochemical changes of potential relevance to clinical diagnosis, treatment planning, and evaluation can be assessed from measurements derived from in vivo MR segmentation of the cartilages. However, segmentation of the glenohumeral cartilages, using approaches spanning manual to automated methods, is technically challenging, due to their thin, curved structure and overlapping intensities of surrounding tissues. Automatic segmentation of the glenohumeral cartilages from MR imaging is not at the same level compared to the weight-bearing knee and hip joint cartilages despite the potential applications with respect to clinical investigation of shoulder disorders. In this work, the authors present a fully automated segmentation method for the glenohumeral cartilages using MR images of healthy shoulders. Methods: The method involves automated segmentation of the humerus and scapula bones using 3D active shape models, the extraction of the expected bone–cartilage interface, and cartilage segmentation using a graph-based method. The cartilage segmentation uses localization, patient specific tissue estimation, and a model of the cartilage thickness variation. The accuracy of this method was experimentally validated using a leave-one-out scheme on a database of MR images acquired from 44 asymptomatic subjects with a true fast imaging with steady state precession sequence on a 3 T scanner (Siemens Trio) using a dedicated shoulder coil. The automated results were compared to manual segmentations from two experts (an experienced radiographer and an experienced musculoskeletal anatomist) using the Dice similarity coefficient (DSC) and mean absolute surface distance (MASD) metrics. Results: Accurate and precise bone segmentations were achieved with mean DSC of 0.98 and 0.93 for the humeral head and glenoid fossa, respectively. Mean DSC scores of 0.74 and 0.72 were obtained for the humeral and glenoid cartilage volumes, respectively. The manual interobserver reliability evaluated by DSC was 0.80 ± 0.03 and 0.76 ± 0.04 for the two cartilages, implying that the automated results were within an acceptable 10% difference. The MASD between the automatic and the corresponding manual cartilage segmentations was less than 0.4 mm (previous studies reported mean cartilage thickness of 1.3 mm). Conclusions: This work shows the feasibility of volumetric segmentation and separation of the glenohumeral cartilages from MR images. To their knowledge, this is the first fully automated algorithm for volumetric segmentation of the individual glenohumeral cartilages from MR images. The approach was validated against manual segmentations from experienced analysts. In future work, the approach will be validated on imaging datasets acquired with various MR contrasts in patients.

60 APPLIED LIFE SCIENCES↗

SU-C-BRA-06: Automatic Brain Tumor Segmentation for Stereotactic Radiosurgery Applications

Purpose: Stereotactic radiosurgery (SRS), which delivers a potent dose of highly conformal radiation to the target in a single fraction, requires accurate tumor delineation for treatment planning. We present an automatic segmentation strategy, that synergizes intensity histogram thresholding, super-voxel clustering, and level-set based contour evolving methods to efficiently and accurately delineate SRS brain tumors on contrast-enhance T1-weighted (T1c) Magnetic Resonance Images (MRI). Methods: The developed auto-segmentation strategy consists of three major steps. Firstly, tumor sites are localized through 2D slice intensity histogram scanning. Then, super voxels are obtained through clustering the corresponding voxels in 3D with reference to the similarity metrics composited from spatial distance and intensity difference. The combination of the above two could generate the initial contour surface. Finally, a localized region active contour model is utilized to evolve the surface to achieve the accurate delineation of the tumors. The developed method was evaluated on numerical phantom data, synthetic BRATS (Multimodal Brain Tumor Image Segmentation challenge) data, and clinical patients’ data. The auto-segmentation results were quantitatively evaluated by comparing to ground truths with both volume and surface similarity metrics. Results: DICE coefficient (DC) was performed as a quantitative metric to evaluate the auto-segmentation in the numerical phantom with 8 tumors. DCs are 0.999±0.001 without noise, 0.969±0.065 with Rician noise and 0.976±0.038 with Gaussian noise. DC, NMI (Normalized Mutual Information), SSIM (Structural Similarity) and Hausdorff distance (HD) were calculated as the metrics for the BRATS and patients’ data. Assessment of BRATS data across 25 tumor segmentation yield DC 0.886±0.078, NMI 0.817±0.108, SSIM 0.997±0.002, and HD 6.483±4.079mm. Evaluation on 8 patients with total 14 tumor sites yield DC 0.872±0.070, NMI 0.824±0.078, SSIM 0.999±0.001, and HD 5.926±6.141mm. Conclusion: The developed automatic segmentation strategy, which yields accurate brain tumor delineation in evaluation cases, is promising for its application in SRS treatment planning.

60 APPLIED LIFE SCIENCES↗

Automated Grain Boundary (GB) Segmentation and Microstructural Analysis in 347H Stainless Steel Using Deep Learning and Multimodal Microscopy

Austenitic 347H stainless steel offers superior mechanical properties and corrosion resistance required for extreme operating conditions such as high temperature. The change in microstructure due to composition and process variations is expected to impact material properties. Identifying microstructural features such as grain boundaries thus becomes an important task in the process-microstructure-properties loop. Applying convolutional neural network (CNN)-based deep learning models is a powerful technique to detect features from material micrographs in an automated manner. In contrast to microstructural classification, supervised CNN models for segmentation tasks require pixel-wise annotation labels. However, manual labeling of the images for the segmentation task poses a major bottleneck for generating training data and labels in a reliable and reproducible way within a reasonable timeframe. Microstructural characterization especially needs to be expedited for faster material discovery by changing alloy compositions. Here, in this study, we attempt to overcome such limitations by utilizing multimodal microscopy to generate labels directly instead of manual labeling. We combine scanning electron microscopy images of 347H stainless steel as training data and electron backscatter diffraction micrographs as pixel-wise labels for grain boundary detection as a semantic segmentation task. The viability of our method is evaluated by considering a set of deep CNN architectures. We demonstrate that despite producing instrumentation drift during data collection between two modes of microscopy, this method performs comparably to similar segmentation tasks that used manual labeling. Additionally, we find that naïve pixel-wise segmentation results in small gaps and missing boundaries in the predicted grain boundary map. By incorporating topological information during model training, the connectivity of the grain boundary network and segmentation performance is improved. Finally, our approach is validated by accurate computation on downstream tasks of predicting the underlying grain morphology distributions which are the ultimate quantities of interest for microstructural characterization.

36 MATERIALS SCIENCE↗

Interactive volumetric segmentation for textile micro-tomography data using wavelets and nonlocal means

This work addresses segmentation of volumetric images of woven carbon fiber textiles from micro-tomography data. We propose a semi-supervised algorithm to classify carbon fibers that requires sparse input as opposed to completely labeled images. The main contributions are: (a) design of effective discriminative classifiers, for three-dimensional textile samples, trained on wavelet features for segmentation; (b) coupling of previous step with nonlocal means as simple, efficient alternative to the Potts model; and (c) demonstration of reuse of classifier to diverse samples containing similar content. We evaluate our work by curating test sets of voxels in the absence of a complete ground truth mask. The algorithm obtains an average 0.95 F1 score on test sets and average F1 score of 0.93 on new samples. Finally, we conclude with discussion of failure cases and propose future directions toward analysis of spatiotemporal high-resolution micro-tomography images.

3D image processing↗

TU-AB-202-11: Tumor Segmentation by Fusion of Multi-Tracer PET Images Using Copula Based Statistical Methods

Purpose: Multi-tracer PET imaging is getting more attention in radiotherapy by providing additional tumor volume information such as glucose and oxygenation. However, automatic PET-based tumor segmentation is still a very challenging problem. We propose a statistical fusion approach to joint segment the sub-area of tumors from the two tracers FDG and FMISO PET images. Methods: Non-standardized Gamma distributions are convenient to model intensity distributions in PET. As a serious correlation exists in multi-tracer PET images, we proposed a new fusion method based on copula which is capable to represent dependency between different tracers. The Hidden Markov Field (HMF) model is used to represent spatial relationship between PET image voxels and statistical dynamics of intensities for each modality. Real PET images of five patients with FDG and FMISO are used to evaluate quantitatively and qualitatively our method. A comparison between individual and multi-tracer segmentations was conducted to show advantages of the proposed fusion method. Results: The segmentation results show that fusion with Gaussian copula can receive high Dice coefficient of 0.84 compared to that of 0.54 and 0.3 of monomodal segmentation results based on individual segmentation of FDG and FMISO PET images. In addition, high correlation coefficients (0.75 to 0.91) for the Gaussian copula for all five testing patients indicates the dependency between tumor regions in the multi-tracer PET images. Conclusion: This study shows that using multi-tracer PET imaging can efficiently improve the segmentation of tumor region where hypoxia and glucidic consumption are present at the same time. Introduction of copulas for modeling the dependency between two tracers can simultaneously take into account information from both tracers and deal with two pathological phenomena. Future work will be to consider other families of copula such as spherical and archimedian copulas, and to eliminate partial volume effect by considering dependency between neighboring voxels.

60 APPLIED LIFE SCIENCES↗

Interactive lung segmentation in abnormal human and animal chest CT scans

Purpose: Many medical image analysis systems require segmentation of the structures of interest as a first step. For scans with gross pathology, automatic segmentation methods may fail. The authors’ aim is to develop a versatile, fast, and reliable interactive system to segment anatomical structures. In this study, this system was used for segmenting lungs in challenging thoracic computed tomography (CT) scans. Methods: In volumetric thoracic CT scans, the chest is segmented and divided into 3D volumes of interest (VOIs), containing voxels with similar densities. These VOIs are automatically labeled as either lung tissue or nonlung tissue. The automatic labeling results can be corrected using an interactive or a supervised interactive approach. When using the supervised interactive system, the user is shown the classification results per slice, whereupon he/she can adjust incorrect labels. The system is retrained continuously, taking the corrections and approvals of the user into account. In this way, the system learns to make a better distinction between lung tissue and nonlung tissue. When using the interactive framework without supervised learning, the user corrects all incorrectly labeled VOIs manually. Both interactive segmentation tools were tested on 32 volumetric CT scans of pigs, mice and humans, containing pulmonary abnormalities. Results: On average, supervised interactive lung segmentation took under 9 min of user interaction. Algorithm computing time was 2 min on average, but can easily be reduced. On average, 2.0% of all VOIs in a scan had to be relabeled. Lung segmentation using the interactive segmentation method took on average 13 min and involved relabeling 3.0% of all VOIs on average. The resulting segmentations correspond well to manual delineations of eight axial slices per scan, with an average Dice similarity coefficient of 0.933. Conclusions: The authors have developed two fast and reliable methods for interactive lung segmentation in challenging chest CT images. Both systems do not require prior knowledge of the scans under consideration and work on a variety of scans.

60 APPLIED LIFE SCIENCES↗