Differentiation of Cartilage Repair Techniques Using Texture Analysis from T2 Maps.

A multichannel statistical classifier for detecting prostate cancer was developed and validated by combining information from three different magnetic resonance (MR) methodologies: T2-weighted, T2-mapping, and line scan diffusion imaging (LSDI). From these MR sequences, four different sets of image intensities were obtained: T2-weighted (T2W) from T2-weighted imaging, Apparent Diffusion Coefficient (ADC) from LSDI, and proton density (PD) and T2 (T2 Map) from T2-mapping imaging. Manually segmented tumor labels from a radiologist, which were validated by biopsy results, served as tumor "ground truth." Textural features were extracted from the images using co-occurrence matrix (CM) and discrete cosine transform (DCT). Anatomical location of voxels was described by a cylindrical coordinate system. A statistical jack-knife approach was used to evaluate our classifiers. Single-channel maximum likelihood (ML) classifiers were based on 1 of the 4 basic image intensities. Our multichannel classifiers: support vector machine (SVM) and Fisher linear discriminant (FLD), utilized five different sets of derived features. Each classifier generated a summary statistical map that indicated tumor likelihood in the peripheral zone (PZ) of the prostate gland. To assess classifier accuracy, the average areas under the receiver operator characteristic (ROC) curves over all subjects were compared. Our best FLD classifier achieved an average ROC area of 0.839(+/-0.064), and our best SVM classifier achieved an average ROC area of 0.761(+/-0.043). The T2W ML classifier, our best single-channel classifier, only achieved an average ROC area of 0.599(+/-0.146). Compared to the best single-channel ML classifier, our best multichannel FLD and SVM classifiers have statistically superior ROC performance (P=0.0003 and 0.0017, respectively) from pairwise two-sided t-test. By integrating the information from multiple images and capturing the textural and anatomical features in tumor areas, summary statistical maps can potentially aid in image-guided prostate biopsy and assist in guiding and controlling delivery of localized therapy under image guidance.

New noninvasive biochemical quantitative magnetic resonance imaging (MRI) techniques are increasingly applied in cartilage repair and osteoarthritis (OA), where quantitative T2 mapping and initially also T2* mapping are seen as valid biomarkers. These quantitative MR biomarkers detect early degeneration of articular cartilage, mainly represented by increasing water content and a disruption of the collagen network. Recently, feasibility of biochemical MR imaging of repair tissue and the surrounding cartilage after cartilage repair procedures has been demonstrated. Different ultrastructural properties of the repair tissue after different repair procedures, such as microfracturing and autologous chondrocyte implantation, resulted in differences in imaging characteristics. Further, these biochemical imaging characteristics varied during follow-up and are able to depict repair tissue maturation. Within these approaches, zonal stratification analysis was considered crucial. Comparably in OA, T2 mapping seems to provide additional information during different stages of cartilage degeneration and can be used in patient subgroups as a predictor for disease progression. Although the results in various studies employing T2 and T2* mapping are very promising and valuable additional information on cartilage ultrastructure can be obtained, (comparable to all the other quantitative cartilage imaging techniques) further multimodal investigation is needed especially since a standard of reference is hard to define.

Prediction of Malignant Cell Infiltration Patterns with Texture Features of Biopsy-Correlated Positron Emission Tomography of Osteolytic Lesions in Multiple Myeloma

Background: Several patient- and defect-specific factors influencing clinical outcomes after matrix-associated chondrocyte transplantation (MACT) have been identified, including the patient’s age, location of the defect, or duration of symptoms before surgery. Little is known, however, about the influence of cell-specific characteristics on clinical results after transplantation. Purpose: The aim of the present study was to investigate the influence of cell differentiation and interleukin-1 β (IL-1β) expression on clinical outcomes up to 5 years after MACT. Study Design: Case series; Level of evidence, 4. Methods: Twenty-seven patients who underwent MACT of the tibiofemoral joint area of the knee were included in this study. Clinical assessments were performed preoperatively as well as 6, 12, 24, and 60 months after transplantation by using the following scores: the Knee injury and Osteoarthritis Outcome Score (KOOS), the International Knee Documentation Committee (IKDC) Subjective Knee Form, the Noyes sports activity rating scale, the Brittberg clinical score, and a visual analog scale (VAS) for pain. The quality of repair tissue was assessed by magnetic resonance imaging using the magnetic resonance observation of cartilage repair tissue (MOCART) score at 1 and 5 years. Cell differentiation (defined as collagen type II:type I expression ratio), aggrecan, and IL-1β expression were determined by real-time polymerase chain reaction in transplant residuals and were correlated with clinical outcomes. Results: The largest improvements in clinical scores were found during the first year. Two years postoperatively, a stable improvement was reached until 5 years after transplantation, with a mean IKDC score of 34.4 ± 8.6 preoperatively to 77.9 ± 16 after 24 months (P < .001). Cell differentiation showed a significant positive correlation with nearly all clinical scores at different time points, especially after 12 months (P < .05). IL-1β expression negatively influenced clinical outcomes at 24 months (Brittberg score) and 60 months (Brittberg and VAS scores) after surgery (P < .05). No correlation was found between the MOCART score and clinical outcomes or gene expression. Conclusion: Our data demonstrate that cell differentiation and IL-1β expression influence clinical outcomes up to 5 years after MACT.

To evaluate the relationship between T1 after intravenous contrast administration (T1Gd) and Delta relaxation rate (DeltaR1) = (1/T1(Gd) - 1/T1o) in the delayed Gadolinium-Enhanced MRI of cartilage (dGEMRIC) evaluation of cartilage repair tissue. Thirty single MR examinations from 30 patients after matrix-associated autologous chondrocyte transplantations of the knee joint with different postoperative intervals were examined using an 8-channel knee-coil at 3T. T1 mapping using a 3D GRE sequence with a 35/10 degrees flip angle excitation pulse combination was performed before and after contrast administration (dGEMRIC technique). T1 postcontrast (T1(Gd)) and the DeltaR1 (relative index of pre- and postcontrast R1 value) were calculated for repair tissue and the weight-bearing normal appearing control cartilage. For evaluation of the different postoperative intervals, MR exams were subdivided into 3 groups (up to 12 months, 12-24 months, more than 24 months). For statistical analysis Spearman correlation coefficients were calculated. The mean value for T1 postcontrast was 427 +/- 159 ms, for DeltaR1 1.85 +/- 1.0; in reference cartilage 636 +/- 181 ms for T1 postcontrast and 0.83 +/- 0.5 for DeltaR1.The correlation coefficients were highly significant between T1 (Gd) and DeltaR1 for repair tissue (0.969) as well as normal reference cartilage (0.928) in total, and for the reparative cartilage in the early, middle postoperative, and late postoperative interval after surgery (R values: -0.986, -0.970, and -0.978, respectively). Using either T1(Gd) or DeltaR1, the 2 metrics resulted in similar conclusions regarding the time course of change of repair tissue and control tissue, namely that highly significant (P > 0.01) differences between cartilage repair tissue and reference cartilage were found for all follow-up groups. Additionally, for both metrics highly significant differences (P < 0.01) between early follow up and the 2 later postoperative groups for cartilage repair tissue were found. No statistical differences were found between the 2 later follow-up groups of reparative cartilage either for T1 (Gd) or DeltaR1. The high correlation between T1 (Gd) and DeltaR1 and the comparable conclusions reached utilizing metric implies that T1 mapping before intravenous administration of MR contrast agent is not necessary for the evaluation of repair tissue. This will help to reduce costs, inconvenience for the patients, simplifies the examination procedure, and makes dGEMRIC more attractive for follow-up of patients after cartilage repair surgeries.

Abstract. Tomato yellow leaf curl disease spreads very fast and often causes huge yield losses. It does not represent obvious symptoms on leaves until the disease becomes serious. This paper investigated the possibility of discriminating tomato yellow leaf curl disease by texture and leaf vein features in fluorescence image. A fluorescence imaging system was used to collect fluorescence images of healthy and diseased tomato leaves. It includes a high-speed camera (30 frames per second), filters (center wavelength: 690 nm), and excitation LEDs (center wavelength: 430 nm). By automatic iterative threshold selection, the leaf pixels were separated from back ground. After region filling, the binary mask images were acquired. Then AND operator was used to extract leaf vein images with gray information. Sub-images were segmented into 40X40 pixels. Leaf vein images were extracted by hit-or-miss transform (HMT) from sub-images. 8 texture features and 8 leaf vein features were calculated by gray level co-occurrence matrix (GLCM). The discrimination model was evaluated by support vector machine (SVM). The results showed that the infected leaf represented a clear leaf vein in sub-images and leaf vein images, especially in the main vein. All texture features and leaf vein features had significant dispersion. The differences between infected and healthy leaves were obvious for 5 features (T_ENT_MEAN, T_ENT_DEV, T_INE_DEV, L_ASM_DEV, and L_ENT_DEV). The method obtained good classification accuracy of 82.61%.

Texture feature of the froth image is widely used in the working condition recognition of froth flotation. However, due to the complexity of the froth image, the current texture features vary greatly and are difficult to identify the work condition accurately. Therefore, we propose a dynamic texture feature named LBP on the TOP and GLCM Histograms (LTGH) which integrates the local binary patterns (LBPs) and gray-level co-occurrence matrix (GLCM) histograms on the three orthogonal planes (TOP). First, we use the rotation invariant LBPs to enhance rotation invariance and illumination robustness. Then, we implement the TOP on the enhanced texture feature map to generate the multiple dimensional enhanced feature maps. After that, we calculate the GLCM and supplementary features (SFs) on the multiple dimensional enhanced feature map. Finally, we integrate the histogram of the GLCM and SFs to discriminate the texture feature. The LTGH feature considers the froth structures both in the macrolevel and microlevel and captures the temporal information between the froth images. Experiments have demonstrated the effectiveness and stability of the proposed texture feature for work condition recognition in froth flotation. Compared with other traditional texture features, the accuracy of the LTGH feature has been increased by at least 7.76%.

Gray Level Co-occurrence Matrix (GLCM), one of the best known texture analysis methods, estimates image properties related to second-order statistics. These image properties commonly known as texture features can be used for image classification, image segmentation, and remote sensing applications. In this paper, we present an FPGA based co-processor to accelerate the extraction of texture features from GLCM. Handel-C, a recently developed C-like programming language for hardware design, has been used for the FPGA implementation of GLCM texture features measurement. Results show that the FPGA has better speed performances when compared to a general purpose processor for the extraction of GLCM features.

We investigate the reconstruction of low-count positron emission tomography (PET) projection, which is an important, but challenging, task. Using the texture feature extraction method of radiomics, i.e. the gray-level co-occurrence matrix (GLCM), texture features can be extracted from magnetic resonance imaging images with high-spatial resolution. In this work, we propose a kernel reconstruction method combining autocorrelation texture features derived from the GLCM. The new kernel function includes the correlations of both the intensity and texture features from the prior image. By regarding the GLCM as a discrete approximation of a probability density function, the asymptotically gray-level-invariant autocorrelation texture feature is generated, which can maintain the accuracy of texture features extracted from small image regions by reducing the number of quantized image gray levels. A computer simulation shows that the proposed method can effectively reduce the noise in the reconstructed image compared to the maximum likelihood expectation maximum method and improve the image quality and tumor region accuracy compared to the original kernel method for low-count PET reconstruction. A simulation study on clinical patient images also shows that the proposed method can improve the whole image quality and that the reconstruction of a high-uptake lesion is more accurate than that achieved by the original kernel method.

Background: Cartilage repair still presents a challenge to clinicians and researchers alike. A more effective, simpler procedure that can produce hyaline-like cartilage is needed for articular cartilage repair. Hypothesis: A technique combining microfracture with a biomaterial scaffold of perforated decalcified cortical-cancellous bone matrix (DCCBM; composed of cortical and cancellous parts) would create a 1-step procedure for hyaline-like cartilage repair. Study Design: Controlled laboratory study. Methods: For the in vitro portion of this study, mesenchymal stem cells (MSCs) were isolated from bone marrow aspirates of New Zealand White rabbits. Scanning electron microscopy (SEM), confocal microscopy, and 1,9-dimethylmethylene blue assay were used to assess the attachment, proliferation, and cartilage matrix production of MSCs grown on a DCCBM scaffold. For the in vivo experiment, full-thickness defects were produced in the articular cartilage of the trochlear groove of 45 New Zealand White rabbits, and the rabbits were then assigned to 1 of 3 treatment groups: perforated DCCBM combined with microfracture (DCCBM+M group), perforated DCCBM alone (DCCBM group), and microfracture alone (M group). Five rabbits in each group were sacrificed at 6, 12, or 24 weeks after the operation, and the repair tissues were analyzed by histological examination, assessment of matrix staining, SEM, and nanoindentation of biomechanical properties. Results: The DCCBM+M group showed hyaline-like articular cartilage repair, and the repair tissues appeared to have better matrix staining and revealed biomechanical properties close to those of the normal cartilage. Compared with the DCCBM+M group, there was unsatisfactory repair tissues with less matrix staining in the DCCBM group and no matrix staining in the M group, as well as poor integration with normal cartilage and poor biomechanical properties. Conclusion: The DCCBM scaffold is suitable for MSC growth and hyaline-like cartilage repair induction when combined with microfracture. Clinical Relevance: Microfracture combined with a DCCBM scaffold is a promising method that can be performed and adopted into clinical treatment for articular cartilage injuries.

Elman Recurrent Neural Network (ERNN) is a model that accommodates network output to be network input to produce the next network output. Genetic algorithms are computational approaches to solving problems that are modeled after the process of biological evolution. The purpose of this research is to classify brain cancer from Magnetic Resonance Imaging (MRI) brain with ERNN model. To improve the accuracy of the classification results, the ERNN model parameters are optimized using genetic algorithms. The data used in the model is the result of extracting MRI images using Gray Level Co-occurrence Matrix (GLCM) method, which consists of 13 variables, those are contrast, correlation, energy, homogeneity, entropy, sum of square variance, Inverse Difference Moment (IDM), sum average, sum entropy, sum variance, difference entropy, maximum probability, and dissimilarity. Classification results are evaluated using accuracy, sensitivity, and specificity on training and testing data. The study has perfect results on training data with no misclassification, while on the testing data the results are also very satisfying with sensitivity, specificity, and accuracy values are 100%, 93.33%, 96.43%.

Biometrics is the study of unique characteristics present in the human body such as fingerprint, palm-print, retina, iris, footprint, etc. While other traits have been explored widely, only a few people have been considered the foot-palm region, despite having unique properties. Prior work has explored the foot shape features using length, width, major axis, minor axis, centroid, etc. but they are not reliable for personal verification due to similarity in the physical composition of two persons. It increases the demand for more unique features based on the footprint. Footprint texture features coming from creases of foot palm are unique and permanent like palmprint texture features. Hence the main objective of the paper is to investigate various kinds of texture feature techniques. These techniques will be further used in correct extraction of footprint features. After extraction of footprint features a detailed experimental analysis is performed to discover the uniqueness in foot texture. It is further utilized to test its viability as a human recognition trait. We describe a detailed feature extraction and classification technique applied to a collected footprint data-set. For feature extraction, we use three techniques: Gray Level Co-occurrence Matrix (GLCM), Histogram Oriented Gradient (HOG), and Local Binary Patterns (LBP). Feature classification is performed using four techniques: Linear Discriminant Analysis (LDA), Support Vector Machine (SVM), K-Nearest Neighbor (KNN), and Ensemble Subspace Discriminant (ESD). GLCM provides less accuracy, while HOG generates a big feature vector which takes more execution time. LBP provides a trade-off between the accuracy and the execution time. Detailed quantitative experiments show: GLCM with LDA provides an accuracy of $$88.5\%$$ , HOG with Fine-KNN achieves $$86.5\%$$ accuracy and LBP with LDA achieves the accuracy of $$97.9\%$$ .

Texture analysis using non-contrast MRI to predict estrogen receptor status in breast cancer lesions: Is it feasible?

This paper proposes to improve the classification accuracy of the leaf images by extracting texture and statistical features by utilizing the presence of striking features on the dorsal and ventral sides of the leaves, which on other types of objects may not be that prominent. The texture features have been extracted from dorsal, ventral and a combination of dorsal-ventral sides of leaf images using Gray level co-occurrence matrix. In addition to this, this work also uses certain general statistical features for discriminating them into various classes. The feature selection work has been performed separately for the dorsal, ventral and combined data sets (for both texture and statistical features) using the most common feature selection algorithms. After selecting the relevant features, the classification has been done using the classification algorithms: K-Nearest Neighbor, J48, Naive Bayes, Partial Least Square (PLS), Classification and Regression Tree (CART), Classification Tree(CT). The classification accuracy has been calculated and compared to find which side of the leaf image (dorsal or ventral) gives better results with which type of features(texture or statistical). This study reveals that the ventral leaf features can be another alternative in discriminating the leaf images into various classes.

Batik is a traditional craft of designing patterned fabrics which hold high artistic value in Sri Lankan culture, where hand-painted wax patterns are coloured using specialist dyeing methods to create the finished product. This paper presents a study of vision-based feature extraction of Batik images considering colour, texture and shape features to develop a comprehensive feature descriptor of Batik motifs. Wax drawn patterns are identified from the digital images of Batik motifs to retrieve an outline of patterns demarcating the different coloured layers generated by multiple stages of dyeing. Motifs with repetitive patterns are identified using the Local Binary Pattern (LBP) as a texture feature vector. Both RGB and L*a*b* colour schemes are studied in the representation of Batik motifs. The colour description is presented using Mini Batch K-Means which out-performed the widely used K-Means clustering method. Hu Moment Invariants are used for shape feature extraction, and Gray Level Co-occurrence Matrix (GLCM) for texture feature extraction. A comprehensive feature descriptor is developed to represent Batik designs, which could be used to recommend similar designs based on the shape and texture features of query images presented by the user.