3D Texture Features Research Articles

3D Texture Analysis of Corpus Caliosum Based on MR Images Inpatients with Alzheimer’s Disease and Mild Cognitive Impairment

This study investigated three-dimensional (3D) texture as a possible diagnostic marker of Alzheimers disease (AD). Methods: T1-weighted MRI of 18 AD patients, 18 Mild Cognitive Impairment (MCI) patients and 18 normal controls (NC) were selected.3D Texture parameters of the corpus callosum,including contrast, inverse difference moment , entropy, short run emphasis, long run emphasis, grey level nonuniformity, run length nonuniformity and fraction were extracted from the gray level co-occurrence matrix and run length matrix. Finally statistic significance was tested among three groups, and the correlations between parameters and Mini-Mental State Examination (MMSE) scores were calculated. Results: The results showed that the 3D texture features had significant differences (p<0.05) among three groups except grey level nonuniformity and run length nonuniformity that the difference was not significant (p>0.05) between MCI and NC or AD and MCI , and they were correlated with MMSE scores.Conclusions: 3D texture analysis can reflect the pathological changes of corpus callosum in patients with AD and MCI, and it may be helpful to AD early diagnosis.

3D Texture Features Mining for MRI Brain Tumor Identification

Medical image segmentation is a process to extract region of interest and to divide an image into its individual meaningful, homogeneous components. Actually, these components will have a strong relationship with the objects of interest in an image. For computer-aided diagnosis and therapy process, medical image segmentation is an initial mandatory step. Medical image segmentation is a sophisticated and challenging task because of the sophisticated nature of the medical images. Indeed, successful medical image analysis heavily dependent on the segmentation accuracy. Texture is one of the major features to identify region of interests in an image or to classify an object. 2D textures features yields poor classification results. Hence, this paper represents 3D features extraction using texture analysis and SVM as segmentation technique in the testing methodologies.

3D texture analysis in renal cell carcinoma tissue image grading.

One of the most significant processes in cancer cell and tissue image analysis is the efficient extraction of features for grading purposes. This research applied two types of three-dimensional texture analysis methods to the extraction of feature values from renal cell carcinoma tissue images, and then evaluated the validity of the methods statistically through grade classification. First, we used a confocal laser scanning microscope to obtain image slices of four grades of renal cell carcinoma, which were then reconstructed into 3D volumes. Next, we extracted quantitative values using a 3D gray level cooccurrence matrix (GLCM) and a 3D wavelet based on two types of basis functions. To evaluate their validity, we predefined 6 different statistical classifiers and applied these to the extracted feature sets. In the grade classification results, 3D Haar wavelet texture features combined with principal component analysis showed the best discrimination results. Classification using 3D wavelet texture features was significantly better than 3D GLCM, suggesting that the former has potential for use in a computer-based grading system.

Reliable Fault Classification of Induction Motors Using Texture Feature Extraction and a Multiclass Support Vector Machine

This paper proposes a method for the reliable fault detection and classification of induction motors using two-dimensional (2D) texture features and a multiclass support vector machine (MCSVM). The proposed model first converts time-domain vibration signals to 2D gray images, resulting in texture patterns (or repetitive patterns), and extracts these texture features by generating the dominant neighborhood structure (DNS) map. The principal component analysis (PCA) is then used for the purpose of dimensionality reduction of the high-dimensional feature vector including the extracted texture features due to the fact that the high-dimensional feature vector can degrade classification performance, and this paper configures an effective feature vector including discriminative fault features for diagnosis. Finally, the proposed approach utilizes the one-against-all (OAA) multiclass support vector machines (MCSVMs) to identify induction motor failures. In this study, the Gaussian radial basis function kernel cooperates with OAA MCSVMs to deal with nonlinear fault features. Experimental results demonstrate that the proposed approach outperforms three state-of-the-art fault diagnosis algorithms in terms of fault classification accuracy, yielding an average classification accuracy of 100% even in noisy environments.

3D Texture Analysis Reveals Imperceptible MRI Textural Alterations in the Thalamus and Putamen in Progressive Myoclonic Epilepsy Type 1, EPM1

Progressive myoclonic epilepsy type 1 (EPM1) is an autosomal recessively inherited neurodegenerative disorder characterized by young onset age, myoclonus and tonic-clonic epileptic seizures. At the time of diagnosis, the visual assessment of the brain MRI is usually normal, with no major changes found later. Therefore, we utilized texture analysis (TA) to characterize and classify the underlying properties of the affected brain tissue by means of 3D texture features. Sixteen genetically verified patients with EPM1 and 16 healthy controls were included in the study. TA was performed upon 3D volumes of interest that were placed bilaterally in the thalamus, amygdala, hippocampus, caudate nucleus and putamen. Compared to the healthy controls, EPM1 patients had significant textural differences especially in the thalamus and right putamen. The most significantly differing texture features included parameters that measure the complexity and heterogeneity of the tissue, such as the co-occurrence matrix-based entropy and angular second moment, and also the run-length matrix-based parameters of gray-level non-uniformity, short run emphasis and long run emphasis. This study demonstrates the usability of 3D TA for extracting additional information from MR images. Textural alterations which suggest complex, coarse and heterogeneous appearance were found bilaterally in the thalamus, supporting the previous literature on thalamic pathology in EPM1. The observed putamenal involvement is a novel finding. Our results encourage further studies on the clinical applications, feasibility, reproducibility and reliability of 3D TA.

Computational identification and quantification of trabecular microarchitecture classes by 3-D texture analysis-based clustering

High resolution peripheral quantitative computed tomography (HR-pQCT) permits the non-invasive assessment of cortical and trabecular bone density, geometry, and microarchitecture. Although researchers have developed various post-processing algorithms to quantify HR-pQCT image properties, few of these techniques capture image features beyond global structure-based metrics. While 3D-texture analysis is a key approach in computer vision, it has been utilized only infrequently in HR-pQCT research. Motivated by high isotropic spatial resolution and the information density provided by HR-pQCT scans, we have developed and evaluated a post-processing algorithm that quantifies microarchitecture characteristics via texture features in HR-pQCT scans. During a training phase in which clustering was applied to texture features extracted from each voxel of trabecular bone, three distinct clusters, or trabecular microarchitecture classes (TMACs) were identified. These TMACs represent trabecular bone regions with common texture characteristics. The TMACs were then used to automatically segment the voxels of new data into three regions corresponding to the trained cluster features. Regional trabecular bone texture was described by the histogram of relative trabecular bone volume covered by each cluster. We evaluated the intra-scanner and inter-scanner reproducibility by assessing the precision errors (PE), intra class correlation coefficients (ICC) and Dice coefficients (DC) of the method on 14 ultradistal radius samples scanned on two HR-pQCT systems. DC showed good reproducibility in intra-scanner set-up with a mean of 0.870±0.027 (no unit). Even in the inter-scanner set-up the ICC showed high reproducibility, ranging from 0.814 to 0.964. In a preliminary clinical test application, the TMAC histograms appear to be a good indicator, when differentiating between postmenopausal women with (n=18) and without (n=18) prevalent fragility fractures. In conclusion, we could demonstrate that 3D-texture analysis and feature clustering seems to be a promising new HR-pQCT post-processing tool with good reproducibility, even between two different scanners.

SU-E-J-186: CT Textures Can be Predictive for Tumor Shrinkage.

To determine if NSCLC GTV CT textures can be predictive for tumor shrinkage after proton therapy with concurrent chemotherapy. Simulation and weekly 4DCTs were obtained from 25 patients with locally advanced NSCLC treated by proton therapy at a 74 Gy dose level with concurrent chemotherapy. In-house deformable image registration software based on the demons algorithm propagated the physician-generated planning GTV onto each weekly image set. GTV soft tissue volumes were then extracted from CT images by in-house software which pruned out voxels below a cutoff HU threshold. Patients whose normalized end-treatment pruned GTV was less than the group median were classified as 'strong responders.'The remaining patients were classified as 'weak responders.' For each pruned GTV, the MaZda texture analysis software generated quantitative 3D texture features based on the image histogram, absolute gradient, co-occurrence matrix, and run-length matrix. Using the Fisher coefficient, a subset of these texture features that were relevant to distinguishing the two response groups was obtained. In-house software and MaZda were then used to perform neural network data classification with leave-one-out cross-validation based on this texture feature subset. The true positive rate (TPR), false positive rate (FPR), classification accuracy, and one-sided p-values were recorded. Six classification models were tested. Each model had a unique combination of feature extraction and analysis parameters. Taking 'strong responders' to be the 'positive' class, the average TPR was 0.66 (0.60 - 0.75), the average FPR was 0.36 (0.31 - 0.40), the average classification accuracy was 64.8% (60.0% - 68.0%), and the average one-sided p-value was 0.118 (0.055 - 0.215). There is some evidence showing that NSCLC GTV CT textures can be predictive for tumor shrinkage after proton therapy. Further work should be done to include more patients, texture metrics, and data analysis techniques. The presenting author receives funding support from a Hertz Foundation Applied Science Fellowship.

3D Texture Feature Extraction Based on a Cubic Uniformity Interpolation Method for Emphysema in Chest CT Images

The application of computer-aided diagnosis system for early diagnosis of lung diseases to achieve the purpose of early treatment is the best solution. In this paper, we propose a Cubic Uniformity Interpolation Method (CUIM) for three dimensional local brightness and structure to detect the pathological change in the chest CT images. we propose an extension of rotation invariant cubic local binary pattern (CLBP), extract texture features in three dimensional directions to represent the brightness uniformity pattern of the image.The 3D optimized texture parameter CLBP allows to discriminate among the three groups of tissue specimens(normal, emphysema ,panlobular emphysema) by quantitative assessment of the underlying texture properties.

Cell cycle phase classification in 3D in vivo microscopy of Drosophila embryogenesis.

BackgroundCell divisions play critical roles in disease and development. The analysis of cell division phenotypes in high content image-based screening and time-lapse microscopy relies on automated nuclear segmentation and classification of cell cycle phases. Automated identification of the cell cycle phase helps biologists quantify the effect of genetic perturbations and drug treatments. Most existing studies have dealt with 2D images of cultured cells. Few, if any, studies have addressed the problem of cell cycle classification in 3D image stacks of intact tissues.ResultsWe developed a workflow for the automated cell cycle phase classification in 3D time-series image datasets of live Drosophila embryos expressing the chromatin marker histone-GFP. Upon image acquisition by laser scanning confocal microscopy and 3D nuclear segmentation, we extracted 3D intensity, shape and texture features from interphase nuclei and mitotic chromosomes. We trained different classifiers, including support vector machines (SVM) and neural networks, to distinguish between 5 cell cycles phases (Interphase and 4 mitotic phases) and achieved over 90% accuracy. As the different phases occur at different frequencies (58% of samples correspond to interphase), we devised a strategy to improve the identification of classes with low representation. To investigate which features are required for accurate classification, we performed feature reduction and selection. We were able to reduce the feature set from 42 to 9 without affecting classifier performance. We observed a dramatic decrease of classification performance when the training and testing samples were derived from two different developmental stages, the nuclear divisions of the syncytial blastoderm and the cell divisions during gastrulation. Combining samples from both developmental stages produced a more robust and accurate classifier.ConclusionsOur study demonstrates that automated cell cycle phase classification, besides 2D images of cultured cells, can also be applied to 3D images of live tissues. We could reduce the initial 3D feature set from 42 to 9 without compromising performance. Robust classifiers of intact animals need to be trained with samples from different developmental stages and cell types. Cell cycle classification in live animals can be used for automated phenotyping and to improve the performance of automated cell tracking.

3D texture analysis on MRI images of Alzheimer’s disease

This study investigated three-dimensional (3D) texture as a possible diagnostic marker of Alzheimer's disease (AD). T1-weighted magnetic resonance (MR) images were obtained from 17AD patients and 17 age and gender-matched healthy controls. 3D texture features were extracted from the circular 3D ROIs placed using a semi-automated technique in the hippocampus and entorhinal cortex. We found that classification accuracies based on texture analysis of the ROIs varied from 64.3% to 96.4% due to different ROI selection, feature extraction and selection options, and that most 3D texture features selected were correlated with the mini-mental state examination (MMSE) scores. The results indicated that 3D texture could detect the subtle texture differences between tissues in AD patients and normal controls, and texture features of MR images in the hippocampus and entorhinal cortex might be related to the severity of AD cognitive impairment. These results suggest that 3D texture might be a useful aid in AD diagnosis.

자궁경부암 진단을 위한 3차원 세포핵 질감 특성값 유의성 평가에 관한 연구

본 연구의 목적은 세포핵의 3차원 염색질 질감 특성값이 암의 진행정도를 인식하는데 있어 유용한 특성값인지 평가하는데 있다. 특히, 제안한 방법이 악성이라고 진단된 세포진 도말 표본에서 정상으로 보이는 세포의 염색질 패턴에서의 미세한 차이를 인식할 수 있는지 살펴보고자 한다. 분류등급 정상(Normal), 저등급 편평 상피내 병변(LSIL, Low grade Squamous Intraepithelial Lesion), 고등급 편평 상피내 병변(HSIL, High grade Squamous Intraepithelial Lesion)에서 각각 100개씩의 세포 볼륨데이터로부터 3차원 GLCM(Gray Level Co occurrence Matrix)에 기반한 질감 특성값과 3차원 Wavelet 변환에 기반한 질감 특성값을 추출하고 분류기를 생성한 후 각 분류기에 대한 분류정확도를 비교하였으며, 2차원 세포진 영상에서의 세포핵 질감 특성값과 비교하기 위해 동일한 실험 볼륨데이터의 투영된 2차원 영상을 이용하여 같은 방법으로 2차원 세포핵 질감 특성값을 추출하고 분류기를 생성한 후 분류정확도를 비교하였다. 2차원 세포핵 질감 특성값과의 비교연구에서 3차원 세포핵 질감 특성값이 등급별 분류에 있어 보다 효율적인 것을 확인 할 수 있었으며 이는 3차원 염색질 질감 특성값이 자궁경부 세포의 정량화에 대한 정확성과 재현성을 개선할 수 있음을 의미한다. The aim of this study is to evaluate whether 3D nuclear chromatin texture features are significant in recognizing the progression of cervical cancer. In particular, we assessed that our method could detect subtle differences in the chromatin pattern of seemingly normal cells on specimens with malignancy. We extracted nuclear texture features based on 3D GLCM(Gray Level Co occurrence Matrix) and 3D Wavelet transform from 100 cell volume data for each group (Normal, LSIL and HSIL). To evaluate the feasibility of 3D chromatin texture analysis, we compared the correct classification rate for each of the classifiers using them. In addition to this, we compared the correct classification rates for the classifiers using the proposed 3D nuclear texture features and the 2D nuclear texture features which were extracted in the same way. The results showed that the classifier using the 3D nuclear texture features provided better results. This means our method could improve the accuracy and reproducibility of quantification of cervical cell.

TU‐G‐211‐03: Automatic Segmentation of Non‐Small Cell Lung Carcinoma Using 3D Texture Features in Co‐Registered FDG PET/CT Images

Purpose: To evaluate the usage of a combination of FDG‐PET/CT features to improve automated segmentation of the gross tumor volume (GTV) in the thorax in order to reduce target definition uncertainty in radiotherapy. Methods: Features of co‐registered FDG‐PET/CT images of patients with non small cell lung carcinoma (NSCLC) were investigated using spatial gray‐level dependence matrices, neighborhood gray tone difference matrices, Tamura textures, first order statistics and structural characteristics. A training data set of PET and CT scans from 21 patients diagnosed with NSCLC was used. Feature samples were taken from regions of interest that included GTV, positive nodes and healthy structures found in the thorax. A decision tree incorporating KNN classifiers as nodes was trained to segment GTVs using an exhaustive search for the optimal combination of features by area under the curve (AUC) at each node. A validation set of 10 patients deemed difficult to contour was used and a probabilistic ground truth was derived from a combination of three observer contours using simultaneous truth and performance level estimation (STAPLE).Results: The concordance index of the three observers was found to average 0.370. CT skewness and PET coarseness were found to be the most useful discriminators when evaluated independently with AUCs of 0.705 and 0.972 respectively. Evaluation of the segmentation results using Dice coefficients found the resulting DTKNN outperformed a variety of thresholds including signal‐to‐background ratio and an implementation of the 3‐FLAB algorithm. Dice coefficients for the DTKNN averaged 0.65 and reached as high as 0.84. Conclusions: Incorporation of texture features from both modalities offers an improvement in segmentation accuracy over approaches that utilize each modality independently. The largest source of error was found to be the misregistration of PET to CT volumes and blurring of PET due to internal motion.

Machine vision 3D skin texture analysis for detection of melanoma

PurposeThe purpose of this paper is to describe innovative machine vision methods that have been employed for the capture and analysis of 3D skin textures; and the resulting potential for assisting with identification of suspicious lesions in the detection of skin cancer.Design/methodology/approachA machine vision approach has been employed for analysis of 3D skin textures. This involves an innovative application of photometric stereo for the capture of the textures, and a range of methods for analysing and quantifying them, including statistical methods and neural networks.Findings3D skin texture has been identified as a useful indicator of skin cancer. It can be used to improve realism of virtual skin reconstructions in tele‐dermatology. 3D texture features can also be combined with 2D features to obtain a more robust classifier for improving diagnostic accuracy, thereby assisting with the long‐term goal of implementing computer‐aided diagnostics for skin cancer.Originality/valueThe device developed for capturing 3D skin textures is known as the “Skin Analyser”, and as far as the authors know it is unique in the world in being able to recover 3D textures from pigmented lesions in vivo. There currently exist numerous methods for analysing lesions, including manual inspection (using established heuristics commonly known as ABCD rules), dermoscopy and SIAoscopy. The ability to capture and analyse 3D lesion textures complements these existing techniques and forms a valuable additional indicator for assisting with the early detection of dangerous skin cancers such as melanoma.

A computer assisted diagnosis system for malignant melanoma using 3D skin surface texture features and artificial neural network

It has been observed that disruptions in skin patterns are larger for malignant melanoma than for benign lesions. In contrast to existing work on 2D skin line patterns, this work proposes a computer assisted diagnosis system for malignant melanoma based on acquiring, analysing and classifying 3D skin surface texture features. Specifically, the 3D skin surface texture, in the form of surface normal vectors are acquired from a six-light photometric stereo device, the 3D features from the surface normals are extracted as the residuals between the acquired data and those from a 2D Gaussian model, while a three-layer feedforward neural classifier is used to classify the residuals. Preliminary studies on a sample set including 12 malignant melanomas and 34 benign lesions have given 91.7% sensitivity and 76.4% specificity using the proposed 3D skin surface normal features, which are better than 91.7% sensitivity and 25.7% specificity using the existing 2D skin line pattern features over the same lesion samples. This demonstrates that the proposed computer assisted diagnosis system of malignant melanoma based on 3D features offers an improvement over that based on 2D skin line patterns.

Computer-aided diagnosis of radiographic patterns of lung disease via MDCT images

With the increasing ability of imaging technologies to provide true volumetric image data, there is an opportunity to fill the need to advance the field of computer-aided texture classification using 3D feature information. We present a method allowing for characterising lung regions of interstitial lung diseases. We compared the inter-observer variation between experts and computer classification results; and analysed the expert's labelling error and computer classification error. The 3D adaptive multiple feature method was in agreement with an expert in 92%, whereas agreement between two experts was 67%. There was no significant classification difference for different selections of VOI sizes for 15 × 15, 21 × 21, and 31 × 31. We demonstrated that 3D texture features can successfully differentiate parenchymal pathologies associated with both emphysema and interstitial lung diseases as well as mixed patterns. The system may assist primary readers to quantify extent of lung disease, and as such could assist in monitoring of treatment effects.

Applied 3D texture features in ALS-based forest inventory

Airborne laser scanning (ALS) data are not usually considered to be very informative with respect to tree species, and this information is often obtained by combining such data with spectral image material. The aim was to test the ability of height, density, intensity and applied 2D and 3D texture variables derived solely from a very high-density ALS point cloud to describe the crown shape and structure characteristics required for tree species discrimination. Linear discriminant analysis was used to find optimal combinations of variables within the predictor groups, and classifications based on variables from different groups were compared. The third power of the tree diameter was used as a stem volume approximate, and rather than examining species alone, the classification was evaluated with respect to the volume approximates assigned to the predicted species. The sensitivity of pulse density to the methodology presented here was determined by simulating thinned data sets by reducing the initial pulse density. The reliability of the estimates was analysed both with functions generated using the original data and with new functions for each thinning level. Alpha shape metrics developed for describing tree crowns constructed from the 3D point clouds proved capable of discriminating between all three species groups evaluated, and several height distribution and textural variables were found to discriminate between the coniferous tree species. The results demonstrate the importance of species interpretation in forest inventories based on allometric modelling, but then indicate that species-specific estimation could be carried out using ALS-derived variables alone.

Three-dimensional Texture Analysis of Renal Cell Carcinoma Cell Nuclei for Computerized Automatic Grading

The extraction of important features in cancer cell image analysis is a key process in grading renal cell carcinoma. In this study, we analyzed the three-dimensional chromatin texture of cell nuclei based on digital image cytometry. Individual images of 2,423 cell nuclei were extracted from 80 renal cell carcinomas (RCCs) using confocal laser scanning microscopy (CLSM). First, we applied the 3D texture mapping method to render the volume of entire tissue sections. Then, we determined the chromatin texture quantitatively by calculating 3D gray level co-occurrence matrices and 3D run length matrices. Finally, to demonstrate the suitability of 3D texture features for classification, we performed a discriminant analysis. In addition, we conducted a principal component analysis to obtain optimized texture features. Automatic grading of cell nuclei using 3D texture features had an accuracy of 78.30%. Combining 3D textural and 3D morphological features improved the accuracy to 82.19%.

Enhancing the discrimination accuracy between metastases, gliomas and meningiomas on brain MRI by volumetric textural features and ensemble pattern recognition methods

Three-dimensional (3D) texture analysis of volumetric brain magnetic resonance (MR) images has been identified as an important indicator for discriminating among different brain pathologies. The purpose of this study was to evaluate the efficiency of 3D textural features using a pattern recognition system in the task of discriminating benign, malignant and metastatic brain tissues on T1 postcontrast MR imaging (MRI) series. The dataset consisted of 67 brain MRI series obtained from patients with verified and untreated intracranial tumors. The pattern recognition system was designed as an ensemble classification scheme employing a support vector machine classifier, specially modified in order to integrate the least squares features transformation logic in its kernel function. The latter, in conjunction with using 3D textural features, enabled boosting up the performance of the system in discriminating metastatic, malignant and benign brain tumors with 77.14%, 89.19% and 93.33% accuracy, respectively. The method was evaluated using an external cross-validation process; thus, results might be considered indicative of the generalization performance of the system to "unseen" cases. The proposed system might be used as an assisting tool for brain tumor characterization on volumetric MRI series.

TERRAIN CLASSIFICATION USING 3D CO-OCCURRENCE FEATURES AND NEURAL NETWORKS

Texture analysis has been efficiently utilized in the area of terrain classification. In this application, features have been obtained in the 2D image domain. In this paper we suggest 3D co-occurrence texture features by extending the concept of co-occurrence feature to the 3D world. The suggested 3D features are described as a 3D co-occurrence matrix by using a co-occurrence histogram of digital elevations at two contiguous positions. The practical construction of the co-occurrence matrix limits the number of levels of digital elevation. If the digital elevation is quantized into a few levels over the whole DEM (Digital Elevation Map), distinctive features cannot be obtained. To resolve this quantization problem, we employ the local quantization technique which can preserve the variation of elevations with a small number of quantization levels. SOM (Self-Organizing Maps), FCM (Fuzzy C-mean) and GBFCM (Gradient Based Fuzzy C-mean) clustering algorithms are employed to implement the terrain classifier, since these ANN clustering algorithms are known as robust against the high dimensionality problem in the classification process. Experimental results show that the classification accuracy with the addition of 3D co-occurrence features is significantly improved over the conventional classification method only with 2D features.

Relevance of 2D radiographic texture analysis for the assessment of 3D bone micro-architecture

Although the diagnosis of osteoporosis is mainly based on dual x-ray absorptiometry, it has been shown that trabecular bone micro-architecture is also an important factor in regard to fracture risk. In vivo, techniques based on high-resolution x-ray radiography associated to texture analysis have been proposed to investigate bone micro-architecture, but their relevance for giving pertinent 3D information is unclear. Thirty-three calcaneus and femoral neck bone samples including the cortical shells (diameter: 14 mm, height: 30-40 mm) were imaged using 3D-synchrotron x-ray micro-CT at the ESRF. The 3D reconstructed images with a cubic voxel size of 15 microm were further used for two purposes: (1) quantification of three-dimensional trabecular bone micro-architecture, (2) simulation of realistic x-ray radiographs under different acquisition conditions. The simulated x-ray radiographs were then analyzed using a large variety of texture analysis methods (co-occurrence, spectral density, fractal, morphology, etc.). The range of micro-architecture parameters was in agreement with previous studies and rather large, suggesting that the population was representative. More than 350 texture parameters were tested. A small number of them were selected based on their correlation to micro-architectural morphometric parameters. Using this subset of texture parameters, multiple regression allowed one to predict up to 93% of the variance of micro-architecture parameters using three texture features. 2D texture features predicting 3D micro-architecture parameters other than BV/TV were identified. The methodology proposed for evaluating the relationships between 3D micro-architecture and 2D texture parameters may also be used for optimizing the conditions for radiographic imaging. Further work will include the application of the method to physical radiographs. In the future, this approach could be used in combination with DXA to refine osteoporosis diagnosis.