Gray-level invariant Haralick texture features

Low-dose computed tomography perceptual image quality assessment

In computed tomography (CT) imaging, optimizing the balance between radiation dose and image quality is crucial due to the potentially harmful effects of radiation on patients. Although subjective assessments by radiologists are considered the gold standard in medical imaging, these evaluations can be time-consuming and costly. Thus, objective methods, such as the peak signal-to-noise ratio and structural similarity index measure, are often employed as alternatives. However, these metrics, initially developed for natural images, may not fully encapsulate the radiologists' assessment process. Consequently, interest in developing deep learning-based image quality assessment (IQA) methods that more closely align with radiologists' perceptions is growing. A significant barrier to this development has been the absence of open-source datasets and benchmark models specific to CT IQA. Addressing these challenges, we organized the Low-dose Computed Tomography Perceptual Image Quality Assessment Challenge in conjunction with the Medical Image Computing and Computer Assisted Intervention 2023. This event introduced the first open-source CT IQA dataset, consisting of 1,000 CT images of various quality, annotated with radiologists' assessment scores. As a benchmark, this challenge offers a comprehensive analysis of six submitted methods, providing valuable insight into their performance. This paper presents a summary of these methods and insights. This challenge underscores the potential for developing no-reference IQA methods that could exceed the capabilities of full-reference IQA methods, making a significant contribution to the research community with this novel dataset. The dataset is accessible at https://zenodo.org/records/7833096.

Capacity and capability of remote sensing to inform invasive plant species management in the Pacific Islands region

The Pacific Islands region is home to several of the world's biodiversity hotspots, yet its unique flora and fauna are under threat because of biological invasions. These invasions are likely to proliferate as human activity increases and large-scale natural disturbances unfold, exacerbated by climate change. Remote sensing data and techniques provide a feasible method to map and monitor invasive plant species and inform invasive plant species management across the Pacific Islands region. We used case studies taken from literature retrieved from Google Scholar, 3 regional agencies' digital libraries, and 2 online catalogs on invasive plant species management to examine the uptake and challenges faced in the implementation of remote sensing technology in the Pacific region. We synthesized remote sensing techniques and outlined their potential to detect and map invasive plant species based on species phenology, structural characteristics, and image texture algorithms. The application of remote sensing methods to detect invasive plant species was heavily reliant on species ecology, extent of invasion, and available geospatial and remotely sensed image data. However, current mechanisms that support invasive plant species management, including policy frameworks and geospatial data infrastructure, operated in isolation, leading to duplication of efforts and creating unsustainable solutions for the region. For remote sensing to support invasive plant species management in the region, key stakeholders including conservation managers, researchers, and practitioners; funding agencies; and regional organizations must invest, where possible, in the broader geospatial and environmental sector, integrate, and streamline policies and improve capacity and technology access.

Hybridization of Acoustic and Visual Features of Polish Sibilants Produced by Children for Computer Speech Diagnosis

Speech disorders are significant barriers to the balanced development of a child. Many children in Poland are affected by lisps (sigmatism)-the incorrect articulation of sibilants. Since speech therapy diagnostics is complex and multifaceted, developing computer-assisted methods is crucial. This paper presents the results of assessing the usefulness of hybrid feature vectors extracted based on multimodal (video and audio) data for the place of articulation assessment in sibilants /s/ and /ʂ/. We used acoustic features and, new in this field, visual parameters describing selected articulators' texture and shape. Analysis using statistical tests indicated the differences between various sibilant realizations in the context of the articulation pattern assessment using hybrid feature vectors. In sound /s/, 35 variables differentiated dental and interdental pronunciation, and 24 were visual (textural and shape). For sibilant /ʂ/, we found 49 statistically significant variables whose distributions differed between speaker groups (alveolar, dental, and postalveolar articulation), and the dominant feature type was noise-band acoustic. Our study suggests hybridizing the acoustic description with video processing provides richer diagnostic information.

Dispersion indices for universal quantification of fluorescently-labelled subcellular structure spatial distributions

In biology, accurate and robust quantification of biological images is critical for understanding distribution patterns and heterogeneity of subcellular structures within a cell. While various methods tailored to specific biological contexts have been employed for image analysis, there is a need for versatile approaches that transcend the constraints imposed by the intricacies of different biological systems. Here we report the application of dispersion indices - a statistical concept widely used to measure the income distribution within a population by economists - as a powerful and agnostic tool for quantifying biological images, which offers distinct advantages over traditional methods. In our approach, we substitute pixel intensity for income and number of pixels for population. We demonstrate the utility of dispersion indices in quantifying autophagic puncta, mitochondrial clustering, and microtubule dynamics, all of which are key measures relevant for maladies ranging from metabolic and neuronal diseases to cancer. Further, we show utility in 2D cell cultures and a 3D multicellular midbrain culture as well as measurement of a performance metric such as a half maximal effective concentration value (EC50).

AI-powered innovations in pancreatitis imaging: a comprehensive literature synthesis

Early identification of pancreatitis remains a significant clinical diagnostic challenge that impacts patient outcomes. The evolution of quantitative imaging followed by deep learning models has shown great promise in the non-invasive diagnosis of pancreatitis and its complications. We provide an overview of advancements in diagnostic imaging and quantitative imaging methods along with the evolution of artificial intelligence (AI). In this article, we review the current and future states of methodology and limitations of AI in improving clinical support in the context of early detection and management of pancreatitis.

Enhanced Diagnostic Precision: Assessing Tumor Differentiation in Head and Neck Squamous Cell Carcinoma Using Multi-Slice Spiral CT Texture Analysis

This study explores the efficacy of texture analysis by using preoperative multi-slice spiral computed tomography (MSCT) to non-invasively determine the grade of cellular differentiation in head and neck squamous cell carcinoma (HNSCC). In a retrospective study, MSCT scans of patients with HNSCC were analyzed and classified based on its histological grade as moderately differentiated, well-differentiated, or poorly differentiated. The location of the tumor was categorized as either in the bone or in soft tissues. Segmentation of the lesion areas was conducted, followed by texture analysis. Eleven GLCM parameters across five different distances were calculated. Median values and correlations of texture parameters were examined in relation to tumor differentiation grade by using Spearman's correlation coefficient and Kruskal-Wallis and Dunn tests. Forty-six patients were included, predominantly female (87%), with a mean age of 66.7 years. Texture analysis revealed significant parameter correlations with histopathological grades of tumor differentiation. The study identified no significant age correlation with tumor differentiation, which underscores the potential of texture analysis as an age-independent biomarker. The strong correlations between texture parameters and histopathological grades support the integration of this technique into the clinical decision-making process.

Estimating the Workability of Concrete with a Stereovision Camera during Mixing

The correct workability of concrete is an essential parameter for its placement and compaction. However, an absence of automatic and transparent measurement methods to estimate the workability of concrete hinders the adaptation from laborious traditional methods such as the slump test. In this paper, we developed a machine-learning framework for estimating the slump class of concrete in the mixer using a stereovision camera. Depth data from five different slump classes was transformed into Haralick texture features to train several machine-learning classifiers. The best-performing classifier achieved a multiclass classification accuracy of 0.8179 with the XGBoost algorithm. Furthermore, we found through statistical analysis that while the denoising of depth data has little effect on the accuracy, the feature extraction of mixer blades and the choice of region of interest significantly increase the accuracy and the efficiency of the classifiers. The proposed framework shows robust results, indicating that stereovision is a competitive solution to estimate the workability of concrete during concrete production.

Persistent fatigue in post-acute COVID syndrome is associated with altered T1 MRI texture in subcortical structures: a preliminary investigation

Post-acute COVID syndrome (PACS) is a global health concern and is often associated with debilitating symptoms. Post-COVID fatigue is a particularly frequent and troubling issue, and its underlying mechanisms remain incompletely understood. One potential contributor is micropathological injury of subcortical and brainstem structures, as has been identified in other patient populations. Texture-based analysis (TA) may be used to measure such changes in anatomical MRI data. The present study develops a methodology of voxel-wise TA mapping in subcortical and brainstem regions, which is then applied to T1-weighted MRI data from a cohort of 48 individuals who had PACS (32 with and 16 without ongoing fatigue symptoms) and 15 controls who had cold and flu-like symptoms but tested negative for COVID-19. Both groups were assessed an average of 4-5 months post-infection. There were no significant differences between PACS and control groups, but significant differences were observed within the PACS groups, between those with and without fatigue symptoms. This included reduced texture energy and increased entropy, along with reduced texture correlation, cluster shade and profile in the putamen, pallidum, thalamus and brainstem. These findings provide new insights into the neurophysiological mechanisms that underlie PACS, with altered tissue texture as a potential biomarker of this debilitating condition.

Evaluating Bacterial Spore Preparation Methods for Scanning Electron Microscopy

Scanning electron microscopy (SEM) can reveal the ultrastructure of bacterial spores, including morphology, surface features, texture, spore damage, germination, and appendages. Understanding these features can provide a basis for adherence, how physical and environmental stressors affect spore viability, integrity, and functionality, as well as the distribution and function of surface appendages. However, the spore sample preparation method can significantly impact the SEM images' appearance, resolution, and overall quality. In this study, we compare different spore preparation methods to identify optimal approaches for preparation time, spore appearance and resolved features, including the exosporium and spore pili, for SEM imaging. We use Bacillus paranthracis as model species and evaluate the efficacy of preparation protocols using different fixation and drying methods, as well as imaging under room- and cryogenic temperatures. We compare and assess method complexity to the visibility of the spore exosporium and spore appendages across different methods. Additionally, we use Haralick texture features to quantify the differences in spore surface appearance and determine the most suitable method for preserving spore structures and surface features during SEM evaluation. The findings from this study will help establish protocols for preparing bacterial spores for SEM and facilitating accurate and reliable analysis of spores' characteristics.

Examining feature extraction and classification modules in machine learning for diagnosis of low-dose computed tomographic screening-detected in vivo lesions

Purpose

Medical imaging-based machine learning (ML) for computer-aided diagnosis of in vivo lesions consists of two basic components or modules of (i) feature extraction from non-invasively acquired medical images and (ii) feature classification for prediction of malignancy of lesions detected or localized in the medical images. This study investigates their individual performances for diagnosis of low-dose computed tomography (CT) screening-detected lesions of pulmonary nodules and colorectal polyps.

Approach

Three feature extraction methods were investigated. One uses the mathematical descriptor of gray-level co-occurrence image texture measure to extract the Haralick image texture features (HFs). One uses the convolutional neural network (CNN) architecture to extract deep learning (DL) image abstractive features (DFs). The third one uses the interactions between lesion tissues and X-ray energy of CT to extract tissue-energy specific characteristic features (TFs). All the above three categories of extracted features were classified by the random forest (RF) classifier with comparison to the DL-CNN method, which reads the images, extracts the DFs, and classifies the DFs in an end-to-end manner. The ML diagnosis of lesions or prediction of lesion malignancy was measured by the area under the receiver operating characteristic curve (AUC). Three lesion image datasets were used. The lesions' tissue pathological reports were used as the learning labels.

Results

Experiments on the three datasets produced AUC values of 0.724 to 0.878 for the HFs, 0.652 to 0.965 for the DFs, and 0.985 to 0.996 for the TFs, compared to the DL-CNN of 0.694 to 0.964. These experimental outcomes indicate that the RF classifier performed comparably to the DL-CNN classification module and the extraction of tissue-energy specific characteristic features dramatically improved AUC value.

Conclusions

The feature extraction module is more important than the feature classification module. Extraction of tissue-energy specific characteristic features is more important than extraction of image abstractive and characteristic features.

MR histology reveals tissue features beneath heterogeneous MRI signal in genetically engineered mouse models of sarcoma

Purpose

To identify significant relationships between quantitative cytometric tissue features and quantitative MR (qMRI) intratumorally in preclinical undifferentiated pleomorphic sarcomas (UPS).

Materials and methods

In a prospective study of genetically engineered mouse models of UPS, we registered imaging libraries consisting of matched multi-contrast in vivo MRI, three-dimensional (3D) multi-contrast high-resolution ex vivo MR histology (MRH), and two-dimensional (2D) tissue slides. From digitized histology we generated quantitative cytometric feature maps from whole-slide automated nuclear segmentation. We automatically segmented intratumoral regions of distinct qMRI values and measured corresponding cytometric features. Linear regression analysis was performed to compare intratumoral qMRI and tissue cytometric features, and results were corrected for multiple comparisons. Linear correlations between qMRI and cytometric features with p values of <0.05 after correction for multiple comparisons were considered significant.

Results

Three features correlated with ex vivo apparent diffusion coefficient (ADC), and no features correlated with in vivo ADC. Six features demonstrated significant linear relationships with ex vivo T2*, and fifteen features correlated significantly with in vivo T2*. In both cases, nuclear Haralick texture features were the most prevalent type of feature correlated with T2*. A small group of nuclear topology features also correlated with one or both T2* contrasts, and positive trends were seen between T2* and nuclear size metrics.

Conclusion

Registered multi-parametric imaging datasets can identify quantitative tissue features which contribute to UPS MR signal. T2* may provide quantitative information about nuclear morphology and pleomorphism, adding histological insights to radiological interpretation of UPS.

Texture-preserving low dose CT image denoising using Pearson divergence

Objective.The mean squared error (MSE), also known asL2loss, has been widely used as a loss function to optimize image denoising models due to its strong performance as a mean estimator of the Gaussian noise model. Recently, various low-dose computed tomography (LDCT) image denoising methods using deep learning combined with the MSE loss have been developed; however, this approach has been observed to suffer from the regression-to-the-mean problem, leading to over-smoothed edges and degradation of texture in the image.Approach.To overcome this issue, we propose a stochastic function in the loss function to improve the texture of the denoised CT images, rather than relying on complicated networks or feature space losses. The proposed loss function includes the MSE loss to learn the mean distribution and the Pearson divergence loss to learn feature textures. Specifically, the Pearson divergence loss is computed in an image space to measure the distance between two intensity measures of denoised low-dose and normal-dose CT images. The evaluation of the proposed model employs a novel approach of multi-metric quantitative analysis utilizing relative texture feature distance.Results.Our experimental results show that the proposed Pearson divergence loss leads to a significant improvement in texture compared to the conventional MSE loss and generative adversarial network (GAN), both qualitatively and quantitatively.Significance.Achieving consistent texture preservation in LDCT is a challenge in conventional GAN-type methods due to adversarial aspects aimed at minimizing noise while preserving texture. By incorporating the Pearson regularizer in the loss function, we can easily achieve a balance between two conflicting properties. Consistent high-quality CT images can significantly help clinicians in diagnoses and supporting researchers in the development of AI-diagnostic models.

Physics-Informed Discretization for Reproducible and Robust Radiomic Feature Extraction Using Quantitative MRI

OBJECTIVE

Given the limited repeatability and reproducibility of radiomic features derived from weighted magnetic resonance imaging (MRI), there may be significant advantages to using radiomics in conjunction with quantitative MRI. This study introduces a novel physics-informed discretization (PID) method for reproducible radiomic feature extraction and evaluates its performance using quantitative MRI sequences including magnetic resonance fingerprinting (MRF) and apparent diffusion coefficient (ADC) mapping.

MATERIALS AND METHODS

A multiscanner, scan-rescan dataset comprising whole-brain 3D quantitative (MRF T1, MRF T2, and ADC) and weighted MRI (T1w MPRAGE, T2w SPACE, and T2w FLAIR) from 5 healthy subjects was prospectively acquired. Subjects underwent 2 repeated acquisitions on 3 distinct 3 T scanners each, for a total of 6 scans per subject (30 total scans). First-order statistical (n = 23) and second-order texture (n = 74) radiomic features were extracted from 56 brain tissue regions of interest using the proposed PID method (for quantitative MRI) and conventional fixed bin number (FBN) discretization (for quantitative MRI and weighted MRI). Interscanner radiomic feature reproducibility was measured using the intraclass correlation coefficient (ICC), and the effect of image sequence (eg, MRF T1 vs T1w MPRAGE), as well as image discretization method (ie, PID vs FBN), on radiomic feature reproducibility was assessed using repeated measures analysis of variance. The robustness of PID and FBN discretization to segmentation error was evaluated by simulating segmentation differences in brainstem regions of interest. Radiomic features with ICCs greater than 0.75 following simulated segmentation were determined to be robust to segmentation.

RESULTS

First-order features demonstrated higher reproducibility in quantitative MRI than weighted MRI sequences, with 30% (n = 7/23) features being more reproducible in MRF T1 and MRF T2 than weighted MRI. Gray level co-occurrence matrix (GLCM) texture features extracted from MRF T1 and MRF T2 were significantly more reproducible using PID compared with FBN discretization; for all quantitative MRI sequences, PID yielded the highest number of texture features with excellent reproducibility (ICC > 0.9). Comparing texture reproducibility of quantitative and weighted MRI, a greater proportion of MRF T1 (n = 225/370, 61%) and MRF T2 (n = 150/370, 41%) texture features had excellent reproducibility (ICC > 0.9) compared with T1w MPRAGE (n = 148/370, 40%), ADC (n = 115/370, 32%), T2w SPACE (n = 98/370, 27%), and FLAIR (n = 102/370, 28%). Physics-informed discretization was also more robust than FBN discretization to segmentation error, as 46% (n = 103/222, 46%) of texture features extracted from quantitative MRI using PID were robust to simulated 6 mm segmentation shift compared with 19% (n = 42/222, 19%) of weighted MRI texture features extracted using FBN discretization.

CONCLUSIONS

The proposed PID method yields radiomic features extracted from quantitative MRI sequences that are more reproducible and robust than radiomic features extracted from weighted MRI using conventional (FBN) discretization approaches. Quantitative MRI sequences also demonstrated greater scan-rescan robustness and first-order feature reproducibility than weighted MRI.

A novel method for vegetable and fruit classification based on using diffusion maps and machine learning

Vegetable and fruit classification can help all links of agricultural product circulation to better carry out inventory management, logistics planning and supply chain coordination, and improve the efficiency and response speed of the supply chain. However, the current classification of vegetables and fruits mainly relies on manual classification, which inevitably introduces the influence of human subjective factors, resulting in errors and misjudgments in the classification of vegetables and fruits. In response to this serious problem, this research proposes an efficient and reproducible novel model to classify multiple vegetables and fruits using handcrafted features. In the proposed model, preprocessing operations such as Gaussian filtering, grayscale and binarization are performed on the pictures of vegetables and fruits to improve the quality of the pictures; statistical texture features representing vegetable and fruit categories, wavelet transform features and shape features are extracted from the preprocessed images; the feature dimension reduction method of diffusion maps is used to reduce the redundant information of the combined features composed of statistical texture features, wavelet transform features and shape features; five effective machine learning methods were used to classify the types of vegetables and fruits. In this research, the proposed method was rigorously verified experimentally and the results show that the SVM classifier achieves 96.25% classification accuracy of vegetables and fruits, which proves that the proposed method is helpful to improve the quality and management level of vegetables and fruits, and provide strong support for agricultural production and supply chain.

AI-based diagnosis and phenotype - Genotype correlations in syndromic craniosynostoses

Apert (AS), Crouzon (CS), Muenke (MS), Pfeiffer (PS), and Saethre Chotzen (SCS) are among the most frequently diagnosed syndromic craniosynostoses. The aims of this study were (1) to train an innovative model using artificial intelligence (AI)-based methods on two-dimensional facial frontal, lateral, and external ear photographs to assist diagnosis for syndromic craniosynostoses vs controls, and (2) to screen for genotype/phenotype correlations in AS, CS, and PS. We included retrospectively and prospectively, from 1979 to 2023, all frontal and lateral pictures of patients genetically diagnosed with AS, CS, MS, PS and SCS syndromes. After a deep learning-based preprocessing, we extracted geometric and textural features and used XGboost (eXtreme Gradient Boosting) to classify patients. The model was tested on an independent international validation set of genetically confirmed patients and non-syndromic controls. Between 1979 and 2023, we included 2228 frontal and lateral facial photographs corresponding to 541 patients. In all, 70.2% [0.593-0.797] (p < 0.001) of patients in the validation set were correctly diagnosed. Genotypes linked to a splice donor site of FGFR2 in Crouzon-Pfeiffer syndrome (CPS) caused a milder phenotype in CPS. Here we report a new method for the automatic detection of syndromic craniosynostoses using AI.

A machine learning-based approach to identify peripheral artery disease using texture features from contrast-enhanced magnetic resonance imaging

Diagnosing and assessing the risk of peripheral artery disease (PAD) has long been a focal point for medical practitioners. The impaired blood circulation in PAD patients results in altered microvascular perfusion patterns in the calf muscles which is the primary location of intermittent claudication pain. Consequently, we hypothesized that changes in perfusion and increase in connective tissue could lead to alterations in the appearance or texture patterns of the skeletal calf muscles, as visualized with non-invasive imaging techniques. We designed an automatic pipeline for textural feature extraction from contrast-enhanced magnetic resonance imaging (CE-MRI) scans and used the texture features to train machine learning models to detect the heterogeneity in the muscle pattern among PAD patients and matched controls. CE-MRIs from 36 PAD patients and 20 matched controls were used for preparing training and testing data at a 7:3 ratio with cross-validation (CV) techniques. We employed feature arrangement and selection methods to optimize the number of features. The proposed method achieved a peak accuracy of 94.11% and a mean testing accuracy of 84.85% in a 2-class classification approach (controls vs. PAD). A three-class classification approach was performed to identify a high-risk PAD sub-group which yielded an average test accuracy of 83.23% (matched controls vs. PAD without diabetes vs. PAD with diabetes). Similarly, we obtained 78.60% average accuracy among matched controls, PAD treadmill exercise completers, and PAD exercise treadmill non-completers. Machine learning and imaging-based texture features may be of interest in the study of lower extremity ischemia.

Next generation phenotyping for diagnosis and phenotype-genotype correlations in Kabuki syndrome

The field of dysmorphology has been changed by the use Artificial Intelligence (AI) and the development of Next Generation Phenotyping (NGP). The aim of this study was to propose a new NGP model for predicting KS (Kabuki Syndrome) on 2D facial photographs and distinguish KS1 (KS type 1, KMT2D-related) from KS2 (KS type 2, KDM6A-related). We included retrospectively and prospectively, from 1998 to 2023, all frontal and lateral pictures of patients with a molecular confirmation of KS. After automatic preprocessing, we extracted geometric and textural features. After incorporation of age, gender, and ethnicity, we used XGboost (eXtreme Gradient Boosting), a supervised machine learning classifier. The model was tested on an independent validation set. Finally, we compared the performances of our model with DeepGestalt (Face2Gene). The study included 1448 frontal and lateral facial photographs from 6 centers, corresponding to 634 patients (527 controls, 107 KS); 82 (78%) of KS patients had a variation in the KMT2D gene (KS1) and 23 (22%) in the KDM6A gene (KS2). We were able to distinguish KS from controls in the independent validation group with an accuracy of 95.8% (78.9-99.9%, p < 0.001) and distinguish KS1 from KS2 with an empirical Area Under the Curve (AUC) of 0.805 (0.729-0.880, p < 0.001). We report an automatic detection model for KS with high performances (AUC 0.993 and accuracy 95.8%). We were able to distinguish patients with KS1 from KS2, with an AUC of 0.805. These results outperform the current commercial AI-based solutions and expert clinicians.

Machine and Deep Learning Approaches Applied to Classify Gougerot-Sjögren Syndrome and Jointly Segment Salivary Glands

To diagnose Gougerot-Sjögren syndrome (GSS), ultrasound imaging (US) is a promising tool for helping physicians and experts. Our project focuses on the automatic detection of the presence of GSS using US. Ultrasound imaging suffers from a weak signal-to-noise ratio. Therefore, any classification or segmentation task based on these images becomes a difficult challenge. To address these two tasks, we evaluate different approaches: a classification using a machine learning method along with feature extraction based on a set of measurements following the radiomics guidance and a deep-learning-based classification. We propose, therefore, an innovative method to enhance the training of a deep neural network with a two phases: multiple supervision using joint classification and a segmentation implemented as pretraining. We highlight the fact that our learning methods provide segmentation results similar to those performed by human experts. We obtain proficient segmentation results for salivary glands and promising detection results for Gougerot-Sjögren syndrome; we observe maximal accuracy with the model trained in two phases. Our experimental results corroborate the fact that deep learning and radiomics combined with ultrasound imaging can be a promising tool for the above-mentioned problems.

Haralick texture analysis for microdosimetry: characterization of Monte Carlo generated 3D specific energy distributions

Objective.Explore the application of Haralick textural analysis to 3D distributions of specific energy (energy imparted per unit mass) scored in cell-scale targets considering varying mean specific energy (absorbed dose), target volume, and incident spectrum.Approach.Monte Carlo simulations are used to generate specific energy distributions in cell-scale water voxels ((1μm)3-(15μm)3) irradiated by photon sources (mean energies: 0.02-2 MeV) to varying mean specific energies (10-400 mGy). Five Haralick features (homogeneity, contrast, entropy, correlation, local homogeneity) are calculated using an implementation of Haralick analysis designed to reduce sensitivity to grey level quantization and are interpreted using fundamental radiation physics.Main results.Haralick measures quantify differences in 3D specific energy distributions observed with varying voxel volume, absorbed dose magnitude, and source spectrum. For example, specific energy distributions in small (1-3μm) voxels with low magnitudes of absorbed dose (10 mGy) have relatively high measures of homogeneity and local homogeneity and relatively low measures of contrast and entropy (all relative to measures for larger voxels), reflecting the many voxels with zero specific energy in an otherwise sporadic distribution. With increasing target size, energy is shared across more target voxels, and trends in Haralick measures, such as decreasing homogeneity and increasing contrast and entropy, reflect characteristics of each 3D specific energy distribution. Specific energy distributions for sources of differing mean energy are characterized by Haralick measures, e.g. contrast generally decreases with increasing source energy, correlation and homogeneity are often (not always) higher for higher energy sources.Significance.Haralick texture analysis successfully quantifies spatial trends in 3D specific energy distributions characteristic of radiation source, target size, and absorbed dose magnitude, thus offering new avenues to quantify microdosimetric data beyond first order histogram features. Promising future directions include investigations of multiscale tissue models, targeted radiation therapy techniques, and biological response to radiation.

Graph 'texture' features as novel metrics that can summarize complex biological graphs

Objective.Image texture features, such as those derived by Haralicket al, are a powerful metric for image classification and are used across fields including cancer research. Our aim is to demonstrate how analogous texture features can be derived for graphs and networks. We also aim to illustrate how these new metrics summarize graphs, may aid comparative graph studies, may help classify biological graphs, and might assist in detecting dysregulation in cancer.Approach.We generate the first analogies of image texture for graphs and networks. Co-occurrence matrices for graphs are generated by summing over all pairs of neighboring nodes in the graph. We generate metrics for fitness landscapes, gene co-expression and regulatory networks, and protein interaction networks. To assess metric sensitivity we varied discretization parameters and noise. To examine these metrics in the cancer context we compare metrics for both simulated and publicly available experimental gene expression and build random forest classifiers for cancer cell lineage.Main results.Our novel graph 'texture' features are shown to be informative of graph structure and node label distributions. The metrics are sensitive to discretization parameters and noise in node labels. We demonstrate that graph texture features vary across different biological graph topologies and node labelings. We show how our texture metrics can be used to classify cell line expression by lineage, demonstrating classifiers with 82% and 89% accuracy.Significance.New metrics provide opportunities for better comparative analyzes and new models for classification. Our texture features are novel second-order graph features for networks or graphs with ordered node labels. In the complex cancer informatics setting, evolutionary analyses and drug response prediction are two examples where new network science approaches like this may prove fruitful.

Using Intratumor Heterogeneity of Immunohistochemistry Biomarkers to Classify Laryngeal and Hypopharyngeal Tumors Based on Histologic Features

Haralick texture features are used to quantify the spatial distribution of signal intensities within an image. In this study, the heterogeneity of proliferation (Ki-67 expression) and immune cells (CD45 expression) within tumors was quantified and used to classify histologic characteristics of larynx and hypopharynx carcinomas. Of 21 laryngectomy specimens, 74 whole-mount tumor slides were scored on histologic characteristics. Ki-67 and CD45 immunohistochemistry was performed, and all sections were digitized. The tumor area was annotated in QuPath. Haralick features independent of the diaminobenzidine intensity were extracted from the isolated diaminobenzidine signal to quantify intratumor heterogeneity. Haralick features from both Ki-67 and CD45 were used as input for a principal component analysis. A linear support vector machine was fitted to the first 4 principal components for classification and validated with a leave-one-patient-out cross-validation method. Significant differences in individual Haralick features were found between cohesive and noncohesive tumors for CD45 (angular second motion: P =.03, inverse difference moment: P =.009, and entropy: P =.02) and between the larynx and hypopharynx tumors for both CD45 (angular second motion: P =.03, inverse difference moment: P =.007, and entropy: P =.005) and Ki-67 (correlation: P =.003). Therefore, these features were used for classification. The linear classifier resulted in a classification accuracy of 85% for site of origin and 81% for growth pattern. A leave-one-patient-out cross-validation resulted in an error rate of 0.27 and 0.35 for both classifiers, respectively. In conclusion, we show a method to quantify intratumor heterogeneity of immunohistochemistry biomarkers using Haralick features. This study also shows the feasibility of using these features to classify tumors by histologic characteristics. The classifiers created in this study are a proof of concept because more data are needed to create robust classifiers, but the method shows potential for automated tumor classification.

Drought stress detection technique for wheat crop using machine learning

The workflow of this research is based on numerous hypotheses involving the usage of pre-processing methods, wheat canopy segmentation methods, and whether the existing models from the past research can be adapted to classify wheat crop water stress. Hence, to construct an automation model for water stress detection, it was found that pre-processing operations known as total variation with L1 data fidelity term (TV-L1) denoising with a Primal-Dual algorithm and min-max contrast stretching are most useful. For wheat canopy segmentation curve fit based K-means algorithm (Cfit-kmeans) was also validated for the most accurate segmentation using intersection over union metric. For automated water stress detection, rapid prototyping of machine learning models revealed that there is a need only to explore nine models. After extensive grid search-based hyper-parameter tuning of machine learning algorithms and 10 K fold cross validation it was found that out of nine different machine algorithms tested, the random forest algorithm has the highest global diagnostic accuracy of 91.164% and is the most suitable for constructing water stress detection models.

Distinguishing Parkinson's Disease with GLCM Features from the Hankelization of EEG Signals

This study proposes a novel method that uses electroencephalography (EEG) signals to classify Parkinson's Disease (PD) and demographically matched healthy control groups. The method utilizes the reduced beta activity and amplitude decrease in EEG signals that are associated with PD. The study involved 61 PD patients and 61 demographically matched controls groups, and EEG signals were recorded in various conditions (eyes closed, eyes open, eyes both open and closed, on-drug, off-drug) from three publicly available EEG data sources (New Mexico, Iowa, and Turku). The preprocessed EEG signals were classified using features obtained from gray-level co-occurrence matrix (GLCM) features through the Hankelization of EEG signals. The performance of classifiers with these novel features was evaluated using extensive cross-validations (CV) and leave-one-out cross-validation (LOOCV) schemes. This method under 10 × 10 fold CV, the method was able to differentiate PD groups from healthy control groups using a support vector machine (SVM) with an accuracy of 92.4 ± 0.01, 85.7 ± 0.02, and 77.1 ± 0.06 for New Mexico, Iowa, and Turku datasets, respectively. After a head-to-head comparison with state-of-the-art methods, this study showed an increase in the classification of PD and controls.

Texture analysis of ultrasound images obtained with different beamforming techniques and dynamic ranges - A robustness study

Texture analysis of medical images gives quantitative information about the tissue characterization for possible pathology discrimination. Ultrasound B-mode images are generated through a process called beamforming. Then, to obtain the final 8-bit image, the dynamic range value must be set. It is currently unknown how different beamforming techniques or dynamic range values may alter the final image texture. We provide here a robustness analysis of first and higher order texture features using six beamforming methods and seven dynamic range values, on experimental phantom and in vivo musculoskeletal images acquired using two different ultrasound research scanners. To investigate the repeatability of the texture parameters, we applied the multivariate analysis of variance (MANOVA) and estimated the intraclass correlation coefficient (ICC) on the texture features calculated on the B-mode images created with different beamforming methods and dynamic range values. We demonstrated the high repeatability of texture features when varying the dynamic range and showed texture features can differentiate between beamforming methods through a MANOVA analysis, hinting at the potential future clinical application of specific beamformers.

The CellPhe toolkit for cell phenotyping using time-lapse imaging and pattern recognition

With phenotypic heterogeneity in whole cell populations widely recognised, the demand for quantitative and temporal analysis approaches to characterise single cell morphology and dynamics has increased. We present CellPhe, a pattern recognition toolkit for the unbiased characterisation of cellular phenotypes within time-lapse videos. CellPhe imports tracking information from multiple segmentation and tracking algorithms to provide automated cell phenotyping from different imaging modalities, including fluorescence. To maximise data quality for downstream analysis, our toolkit includes automated recognition and removal of erroneous cell boundaries induced by inaccurate tracking and segmentation. We provide an extensive list of features extracted from individual cell time series, with custom feature selection to identify variables that provide greatest discrimination for the analysis in question. Using ensemble classification for accurate prediction of cellular phenotype and clustering algorithms for the characterisation of heterogeneous subsets, we validate and prove adaptability using different cell types and experimental conditions.

Radiogenomics Reveals Correlation between Quantitative Texture Radiomic Features of Biparametric MRI and Hypoxia-Related Gene Expression in Men with Localised Prostate Cancer

OBJECTIVES

To perform multiscale correlation analysis between quantitative texture feature phenotypes of pre-biopsy biparametric MRI (bpMRI) and targeted sequence-based RNA expression for hypoxia-related genes.

MATERIALS AND METHODS

Images from pre-biopsy 3T bpMRI scans in clinically localised PCa patients of various risk categories (n = 15) were used to extract textural features. The genomic landscape of hypoxia-related gene expression was obtained using post-radical prostatectomy tissue for targeted RNA expression profiling using the TempO-sequence method. The nonparametric Games Howell test was used to correlate the differential expression of the important hypoxia-related genes with 28 radiomic texture features. Then, cBioportal was accessed, and a gene-specific query was executed to extract the Oncoprint genomic output graph of the selected hypoxia-related genes from The Cancer Genome Atlas (TCGA). Based on each selected gene profile, correlation analysis using Pearson's coefficients and survival analysis using Kaplan-Meier estimators were performed.

RESULTS

The quantitative bpMR imaging textural features, including the histogram and grey level co-occurrence matrix (GLCM), correlated with three hypoxia-related genes (ANGPTL4, VEGFA, and P4HA1) based on RNA sequencing using the TempO-Seq method. Further radiogenomic analysis, including data accessed from the cBioportal genomic database, confirmed that overexpressed hypoxia-related genes significantly correlated with a poor survival outcomes, with a median survival ratio of 81.11:133.00 months in those with and without alterations in genes, respectively.

CONCLUSION

This study found that there is a correlation between the radiomic texture features extracted from bpMRI in localised prostate cancer and the hypoxia-related genes that are differentially expressed. The analysis of expression data based on cBioportal revealed that these hypoxia-related genes, which were the focus of the study, are linked to an unfavourable survival outcomes in prostate cancer patients.

Haralick texture feature analysis for characterization of specific energy and absorbed dose distributions across cellular to patient length scales

Objective.To investigate an approach for quantitative characterization of the spatial distribution of dosimetric data by introducing Haralick texture feature analysis in this context.Approach.Monte Carlo simulations are used to generate 3D arrays of dosimetric data for 2 scenarios: (1) cell-scale microdosimetry: specific energy (energy imparted per unit mass) in cell-scale targets irradiated by photon spectra (¹²⁵I,¹⁹²Ir, 6 MV); (2) tumour-scale dosimetry: absorbed dose in voxels for idealized models of¹²⁵I permanent implant prostate brachytherapy, considering 'TG186' (realistic tissues including 0% to 5% intraprostatic calcifications; interseed attenuation) and 'TG43' (water model, no interseed attenuation) conditions. Five prominent Haralick features (homogeneity, contrast, correlation, local homogeneity, entropy) are computed and trends are interpreted using fundamental radiation physics.Main results.In the cell-scale scenario, the Haralick measures quantify differences in 3D specific energy distributions due to source spectra. For example, contrast and entropy are highest for¹²⁵I reflecting the large variations in specific energy in adjacent voxels (photoelectric interactions; relatively short range of electrons), while 6 MV has the highest homogeneity with smaller variations in specific energy between voxels (Compton scattering dominates; longer range of electrons). For the tumour-scale scenario, the Haralick measures quantify differences due to TG186/TG43 simulation conditions and the presence of calcifications. For example, as calcifications increase from 0% to 5%, contrast increases while correlation decreases, reflecting the large differences in absorbed dose in adjacent voxels (higher absorbed dose in voxels with calcification due to photoelectric interactions).Significance.Haralick texture analysis provides a quantitative method for the characterization of 3D dosimetric distributions across cellular to tumour length scales, with promising future applications including analyses of multiscale tissue models, patient-specific data, and comparison of treatment approaches.

Fault-Level Grading of Photovoltaic Cells Employing Lightweight Deep Learning Models

The deployment of photovoltaic (PV) cells as a renewable energy resource has been boosted recently, which enhanced the need to develop an automatic and swift fault detection system for PV cells. Prior to isolation for repair or replacement, it is critical to judge the level of the fault that occurred in the PV cell. The aim of this research study is the fault-level grading of PV cells employing deep neural network models. The experiment is carried out using a publically available dataset of 2,624 electroluminescence images of PV cells, which are labeled with four distinct defect probabilities defined as the defect levels. The deep architectures of the classical artificial neural networks are developed while employing hand-crafted texture features extracted from the EL image data. Moreover, optimized architectures of the convolutional neural network are developed with a specific emphasis on lightweight models for real-time processing. The experiments are performed for two-way binary classification and multiclass classification. For the first binary categorization, the proposed CNN model outperformed the state-of-the-art solution with a margin of 1.3% in accuracy with a significant 50% less computational complexity. In the second binary classification task, the CPU-based proposed model outperformed the GPU-based solution with a margin of 0.9% accuracy with an 8× lighter architecture. Finally, the multiclass categorization of PV cells is performed and the state-of-the-art results with 83.5% accuracy are achieved. The proposed models offer a lightweight, efficient, and computationally cheaper CPU-based solution for the real-time fault-level categorization of PV cells.

Development of an image classification pipeline for atherosclerotic plaques assessment using supervised machine learning

Background

During atherosclerosis, the narrowing of the arterial lumen is observed through the accumulation of bio compounds and the formation of plaque within artery walls. A non-linear optical imaging modality (NLOM), coherent anti-stokes Raman scattering (CARS) microscopy, can be used to image lipid-rich structures commonly found in atherosclerotic plaques. By matching the lipid’s molecular vibrational frequencies (CH bonds), it is possible to map the accumulation of lipid-rich structures without the need for exogenous labelling and/or processing of the samples. CARS allows for the visualization of the morphological features of plaque. In combination with supervised machine learning, CARS imaged morphological features can be used to characterize the progression of atherosclerotic plaques. 

Results

Based on a set of label-free CARS images of atherosclerotic plaques (i.e. foam cell clusters) from a Watanabe heritable hyperlipidemic rabbit model, we developed an automated pipeline to classify atherosclerotic lesions based on their major morphological features. Our method uses image preprocessing to first improve the quality of the CARS-imaged plaque, followed by the segmentation of the plaque using Otsu thresholding, marker-controlled watershed, K-means segmentation and a novel independent foam cell thresholding segmentation. To define relevant morphological features, 27 quantitative features were extracted and further refined by a novel coefficient of variation feature refinement method in accordance with filter-type feature selection. Refined morphological features were supplied into three supervised machine learning algorithms; K-nearest neighbour, support vector machine and decision tree classifier. The classification pipeline showcased the ability to exploit relevant plaque morphological features to accurately classify 3 pre-defined stages of atherosclerosis: early fatty streak development (EFS) and advancing atheroma (AA) with a greater than 85% class accuracy

Conclusions

Through the combination of CARS microscopy and computational methods, a powerful classification tool was developed to identify the progression of atherosclerotic plaque in an automated manner. Using a curated dataset, the classification pipeline demonstrated the ability to differentiate between EFS, EF and AA. Thus, presenting the opportunity to classify the onset of atherosclerosis at an earlier stage of development

Investigation of the Inter- and Intrascanner Reproducibility and Repeatability of Radiomics Features in T1-Weighted Brain MRI

BACKGROUND

Radiomics is the high throughput analysis of medical images using computer algorithms, which specifically assess textural features. It has increasingly been proposed as a tool for the development of imaging biomarkers. However, an important acknowledged limitation of radiomics is the lack of reproducibility of features produced.

PURPOSE

To assess reproducibility and repeatability of radiomics variables in brain MRI through a multivisit, multicenter study.

STUDY TYPE

Retrospective.

POPULATION

Fourteen individuals visiting three institutions twice, 10 males with the mean age of 36.3 years and age range 25-51.

FIELD STRENGTH

3D T1W inversion recovery on three 1.5-T General Electric scanners.

ASSESSMENT

Radiomics analysis by a consultant radiologist performed on the T1W images of the whole brain on all visits. All possible radiomics features were generated.

STATISTICAL TEST

Concordance correlation coefficient (CCC) and dynamic range (DR) for all variables were calculated to assess the test-retest repeatability. Intraclass correlation coefficients (ICCs) were calculated to investigate the reproducibility of features across centers.

RESULTS

Of 1596 features generated, 57 from center 1, 15 from center 2, and 22 from center 3 had a CCC > 0.9 and DR > 0.9. Eight variables had CCC > 0.9 and DR > 0.9 in all centers. Forty-one variables had an ICC of >0.9. No variables had CCC > 0.9, DR > 0.9, and ICC > 0.9.

DATA CONCLUSION

Repeatability and reproducibility of variables is a significant limitation of radiomics analysis in 3DT1W brain MRI. Careful selection of radiomic features is required.

LEVEL OF EVIDENCE

4 TECHNICAL EFFICACY STAGE: 2.

Simulation of land use/land cover changes and urban expansion in Estonia by a hybrid ANN-CA-MCA model and utilizing spectral-textural indices

Over the recent two decades, land use/land cover (LULC) drastically changed in Estonia. Even though the population decreased by 11%, noticeable agricultural and forest land areas were turned into urban land. In this work, we analyzed those LULC changes by mapping the spatial characteristics of LULC and urban expansion in the years 2000-2019 in Estonia. Moreover, using the revealed spatiotemporal transitions of LULC, we simulated LULC and urban expansion for 2030. Landsat 5 and 8 data were used to estimate 147 spectral-textural indices in the Google Earth Engine cloud computing platform. After that, 19 selected indices were used to model LULC changes by applying the hybrid artificial neural network, cellular automata, and Markov chain analysis (ANN-CA-MCA). While determining spectral-textural indices is quite common for LULC classifications, utilization of these continues indices in LULC change detection and examining these indices at the landscape scale is still in infancy. This country-wide modeling approach provided the first comprehensive projection of future LULC utilizing spectral-textural indices. In this work, we utilized the hybrid ANN-CA-MCA model for predicting LULC in Estonia for 2030; we revealed that the predicted changes in LULC from 2019 to 2030 were similar to the observed changes from 2011 to 2019. The predicted change in the area of artificial surfaces was an increased rate of 1.33% to reach 787.04 km² in total by 2030. Between 2019 and 2030, the other significant changes were the decrease of 34.57 km² of forest lands and the increase of agricultural lands by 14.90 km² and wetlands by 9.31 km². These findings can develop a proper course of action for long-term spatial planning in Estonia. Therefore, a key policy priority should be to plan for the stable care of forest lands to maintain biodiversity.

Multimodal NASH prognosis using 3D imaging flow cytometry and artificial intelligence to characterize liver cells

To improve the understanding of the complex biological process underlying the development of non-alcoholic steatohepatitis (NASH), 3D imaging flow cytometry (3D-IFC) with transmission and side-scattered images were used to characterize hepatic stellate cell (HSC) and liver endothelial cell (LEC) morphology at single-cell resolution. In this study, HSC and LEC were obtained from biopsy-proven NASH subjects with early-stage NASH (F2-F3) and healthy controls. Here, we applied single-cell imaging and 3D digital reconstructions of healthy and diseased cells to analyze a spatially resolved set of morphometric cellular and texture parameters that showed regression with disease progression. By developing a customized autoencoder convolutional neural network (CNN) based on label-free cell transmission and side scattering images obtained from a 3D imaging flow cytometer, we demonstrated key regulated cell types involved in the development of NASH and cell classification performance superior to conventional machine learning methods.

Early Osteogenic Marker Expression in hMSCs Cultured onto Acid Etching-Derived Micro- and Nanotopography 3D-Printed Titanium Surfaces

Polyetheretherketone (PEEK) titanium composite (PTC) is a novel interbody fusion device that combines a PEEK core with titanium alloy (Ti6Al4V) endplates. The present study aimed to investigate the in vitro biological reactivity of human bone-marrow-derived mesenchymal stem cells (hBM-MSCs) to micro- and nanotopographies produced by an acid-etching process on the surface of 3D-printed PTC endplates. Optical profilometer and scanning electron microscopy were used to assess the surface roughness and identify the nano-features of etched or unetched PTC endplates, respectively. The viability, morphology and the expression of specific osteogenic markers were examined after 7 days of culture in the seeded cells. Haralick texture analysis was carried out on the unseeded endplates to correlate surface texture features to the biological data. The acid-etching process modified the surface roughness of the 3D-printed PTC endplates, creating micro- and nano-scale structures that significantly contributed to sustaining the viability of hBM-MSCs and triggering the expression of early osteogenic markers, such as alkaline phosphatase activity and bone-ECM protein production. Finally, the topography of 3D-printed PTC endplates influenced Haralick's features, which in turn correlated with the expression of two osteogenic markers, osteopontin and osteocalcin. Overall, these data demonstrate that the acid-etching process of PTC endplates created a favourable environment for osteogenic differentiation of hBM-MSCs and may potentially have clinical benefit.

Radiomics Nomogram Based on Contrast-enhanced CT to Predict the Malignant Potential of Gastrointestinal Stromal Tumor: A Two-center Study

RATIONALE AND OBJECTIVES

Contrast-enhanced computed tomography (CE-CT) was used to establish radiomics nomogram to evaluate the malignant potential of gastrointestinal stromal tumors (GISTs).

MATERIALS AND METHODS

A total of 500 GIST patients were enrolled in this study and divided into training cohort (n = 346, our center) and validation cohort (n = 154, another center). Minimum redundancy maximum relevance (mRMR) and least absolute shrinkage and selection operator (LASSO) algorithms were used to select the feature subset with the best discriminant features from the three phases image, and five classifiers were used to establish four radiomics signatures. Preoperative radiomics nomogram was constructed by adding the clinical features determined by multivariate logistic regression analysis. The performance of radiomics signatures and nomogram were evaluated by area under the curve (AUC) of the receiver operating characteristic (ROC). The calibration of nomogram was appraised by calibration curve.

RESULTS

A total of 13 radiomic features were extracted from tri-phase combined CE-CT images. Tri-phase combined CE-CT features + Support Vector Machine (SVM) was the best combination at predicting the malignant potential of GIST, with an AUC of 0.895 (95% CI 0.858-0.931) in the training cohort and 0.847 (95% CI 0.778-0.917) in the validation cohort. The nomogram also had good calibration. In the training cohort and the validation cohort, preoperative radiomics nomogram reached AUCs of 0.927 and 0.905, respectively, which were higher than clinical.

CONCLUSION

The radiomics nomogram had a good predictive effect and generalization on the malignant potential of GIST, which could effectively help guide preoperative clinical decision.

Modified Gray-Level Haralick Texture Features for Early Detection of Diabetes Mellitus and High Cholesterol with Iris Image

Iris has specific advantages, which can record all organ conditions, body construction, and psychological disorders. Traces related to the intensity or deviation of organs caused by the disease are recorded systematically and patterned on the iris and its surroundings. The pattern that appears on the iris can be recognized by using image processing techniques. Based on the pattern in the iris image, this paper aims to provide an alternative noninvasive method for the early detection of DM and HC. In this paper, we perform detection based on iris images for two diseases, DM and HC simultaneously, by developing the invariant Haralick feature on quantized images with 256, 128, 64, 32, and 16 gray levels. The feature extraction process does early detection based on iris images. Researchers and scientists have introduced many methods, one of which is the feature extraction of the gray-level co-occurrence matrix (GLCM). Early detection based on the iris is done using the volumetric GLCM development, namely, 3D-GLCM. Based on 3D-GLCM, which is formed at a distance of d = 1 and in the direction of 0°, 45°, 90°, 135°, 180°, 225°, 270°, and 315°, it is used to calculate Haralick features and develop Haralick features which are invariant to the number of quantization gray levels. The test results show that the invariant feature with a gray level of 256 has the best identification performance. In dataset I, the accuracy value is 97.92, precision is 96.88, and recall is 95.83, while in dataset II, the accuracy value is 95.83, precision is 89.69, and recall is 91.67. The identification of DM and HC trained on invariant features showed higher accuracy than the original features.

Haralick's texture analysis to predict cellular proliferation on randomly oriented electrospun nanomaterials

Using a computer vision approach we have extracted the Haralick's texture features of randomly oriented electrospun nanomaterials in order to predict the proliferative behavior of cells which were subsequently seeded onto the nanosurfaces.

Ultrasound Based Radiomics Features of Chronic Kidney Disease

RATIONALE AND OBJECTIVES

Interstitial fibrosis, common to most chronic kidney diseases, can potentially affect the speckle patterns of kidney ultrasound (US). Here we use Radiomics features derived from US images to identify kidneys with chronic kidney disease.

MATERIALS AND METHODS

B-mode US without speckle reduction was performed on a cohort of CKD patients (n = 75) and healthy subjects (n = 27). Images of the patients with renal cysts, agenesis and calculi were excluded. After background subtraction, regions of interest were selected from each kidney. Four hundred and sixty-five Radiomics features including first and second-order gray level statistics were calculated on the selected regions. Second-order features were also calculated on wavelet transformed images. A random forest model was used to identify the most important features that can differentiate healthy and diseased kidneys. The ten most important features, based on the Gini index, were used to train a support vector machine. Synthetic minority oversampling technique was used to remove over fitting.

RESULTS

Wavelet transformed, Gray Level Run Length Matrix based Normalized Run Length Non-uniformity, WT (LH) (GRLN) was identified as the most significant feature in differentiating CKD and healthy kidneys (accuracy - 0.85, sensitivity - 1.0). The mean WT (LH) GRLN of healthy kidneys (0.40 ± 0.01) was significantly higher (p < 0.01) than that of the CKD kidneys (0.24 ± 0.01). According to the Gini Index, the differentiability of WT (LH) GRLN was highest when the long axis of the kidney was oriented perpendicular to the columns of the image matrix.

CONCLUSION

Radiomics features based on wavelet transformation are sensitive to directionality of US speckle patters and can be successfully used to differentiate CKD and healthy US kidney images.

Establishing ADC-Based Histogram and Texture Features for Early Treatment-Induced Changes in Head and Neck Squamous Cell Carcinoma

The purpose of this study was to assess baseline variability in histogram and texture features derived from apparent diffusion coefficient (ADC) maps from diffusion-weighted MRI (DW-MRI) examinations and to identify early treatment-induced changes to these features in patients with head and neck squamous cell carcinoma (HNSCC) undergoing definitive chemoradiation. Patients with American Joint Committee on Cancer Stage III–IV (7^th edition) HNSCC were prospectively enrolled on an IRB-approved study to undergo two pre-treatment baseline DW-MRI examinations, performed 1 week apart, and a third early intra-treatment DW-MRI examination during the second week of chemoradiation. Forty texture and six histogram features were derived from ADC maps. Repeatability of the features from the baseline ADC maps was assessed with the intra-class correlation coefficient (ICC). A Wilcoxon signed-rank test compared average baseline and early treatment feature changes. Data from nine patients were used for this study. Comparison of the two baseline ADC maps yielded 11 features with an ICC ≥ 0.80, indicating that these features had excellent repeatability: Run Gray-Level Non-Uniformity, Coarseness, Long Zone High Gray-Level, Variance (Histogram Feature), Cluster Shade, Long Zone, Variance (Texture Feature), Run Length Non-Uniformity, Correlation, Cluster Tendency, and ADC Median. The Wilcoxon signed-rank test resulted in four features with significantly different early treatment-induced changes compared to the baseline values: Run Gray-Level Non-Uniformity (p = 0.005), Run Length Non-Uniformity (p = 0.005), Coarseness (p = 0.006), and Variance (Histogram) (p = 0.006). The feasibility of histogram and texture analysis as a potential biomarker is dependent on the baseline variability of each metric, which disqualifies many features.

Technical Note: On the development of an outcome-driven frequency filter for improving Radiomics-based modeling of Human Papilloma Virus (HPV) in patients with oropharyngeal squamous cell carcinomas

PURPOSE

To implement an outcome-driven frequency filter for improving radiomics-based modeling of human papilloma virus (HPV) for patients with oropharyngeal squamous cell carcinoma (OPSCC).

METHODS AND MATERIALS

One hundred twenty-eight OPSCC patients with known HPV status (60-HPV+ and 68-HPV-, confirmed by immunohistochemistry-P16 protein testing) were retrospectively studied. A 3D Discrete Fourier Transform was applied on contrast-enhanced CT images of patient gross tumor volumes (GTV's) to transform intensity distributions to the frequency domain and estimate frequency power spectrums of HPV- and HPV+ patient cohorts. Statistical analyses were performed to rank frequency bands contributing towards prediction of HPV status. An outcome-driven frequency filter was designed accordingly and applied to GTV frequency information. 3D Inverse-Discrete-Fourier-Transform was applied to reconstruct HPV-related frequency-filtered images. Radiomics features (11 feature-categories) were extracted from pre- and post- frequency filtered images using our previously published 'ROdiomiX' software. Least-Absolute-Shrinkage-and-Selection-Operation (Lasso) combined with a Generalized-Linear-Model (Lasso-GLM) was developed to identify and rank feature subsets with optimal information for prediction of HPV+/-. Radiomics-based Lasso-GLM classifiers (pre- and post-frequency filtered) were constructed and validated using a random permutation sampling and nested cross-validation techniques. Average Area Under Receiver Operating Characteristic (AUC), and Positive and Negative Predictive values (PPV, NPV) were computed to estimate generalization error and prediction performance.

RESULTS

Among 192 radiomic features, 15 features were found to be statistically significant discriminators between HPV+/- cohorts on post-frequency filtered CE-CT images; 14 such radiomic features were observed on pre-frequency filtered datasets. Discriminant features included tumor morphology and intensity contrast. Performances for prediction of HPV for the pre- and post-frequency filtered Lasso-GLM classifiers were: AUC/PPV/NPV = 0.789/0.755/0.805 and 0.850/0.808/0.877 respectively. Nested-CV performances for prediction of HPV for the pre- and post-frequency filtered Lasso-GLM classifiers were: AUC/PPV/NPV = 0.814/0.725/0.877 and 0.890/0.820/0.911 respectively.

CONCLUSION

Albeit subject to confirmation in a larger cohort, this pilot study presents encouraging results on the importance of frequency analysis prior to radiomic feature extraction toward enhancement of model performance for characterizing HPV in patients with OPSCC. This article is protected by copyright. All rights reserved.

MRI-aided kernel PET image reconstruction method based on texture 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.

Accurate Diagnosis of Diabetic Retinopathy and Glaucoma Using Retinal Fundus Images Based on Hybrid Features and Genetic Algorithm

Diabetic retinopathy (DR) and glaucoma can both be incurable if they are not detected early enough. Therefore, ophthalmologists worldwide are striving to detect them by personally screening retinal fundus images. However, this procedure is not only tedious, subjective, and labor-intensive, but also error-prone. Worse yet, it may not even be attainable in some countries where ophthalmologists are in short supply. A practical solution to this complicated problem is a computer-aided diagnosis (CAD) system—the objective of this work. We propose an accurate system to detect at once any of the two diseases from retinal fundus images. The accuracy stems from two factors. First, we calculate a large set of hybrid features belonging to three groups: first-order statistics (FOS), higher-order statistics (HOS), and histogram of oriented gradient (HOG). Then, these features are skillfully reduced using a genetic algorithm scheme that selects only the most relevant and significant of them. Finally, the selected features are fed to a classifier to detect one of three classes: DR, glaucoma, or normal. Four classifiers are tested for this job: decision tree (DT), naive Bayes (NB), k-nearest neighbor (kNN), and linear discriminant analysis (LDA). The experimental work, conducted on three publicly available datasets, two of them merged into one, shows impressive performance in terms of four standard classification metrics, each computed using k-fold crossvalidation for added credibility. The highest accuracy has been provided by DT—96.67% for DR, 100% for glaucoma, and 96.67% for normal.

Classification of Chronic Kidney Disease in Sonography Using the GLCM and Artificial Neural Network

Chronic kidney disease (CKD) can be treated if it is detected early, but as the disease progresses, recovery becomes impossible. Eventually, renal replacement therapy such as transplantation or dialysis is necessary. Ultrasound is a test method with which to diagnose kidney cancer, inflammatory disease, nodular disease, chronic kidney disease, etc. It is used to determine the degree of inflammation using information such as the kidney size and internal echo characteristics. The degree of the progression of chronic kidney disease in the current clinical trial is based on the value of the glomerular filtration rate. However, changes in the degree of inflammation and disease can even be observed with ultrasound. In this study, from a total of 741 images, 251 normal kidney images, 328 mild and moderate CKD images, and 162 severe CKD images were tested. In order to diagnose CKD in clinical practice, three ROIs were set: the cortex of the kidney, the boundary between the cortex and medulla, and the medulla, which are areas examined to obtain information from ultrasound images. Parameters were extracted from each ROI using the GLCM algorithm, which is widely used in ultrasound image analysis. When each parameter was extracted from the three areas, a total of 57 GLCM parameters were extracted. Finally, a total of 58 parameters were used by adding information on the size of the kidney, which is important for the diagnosis of chronic kidney disease. The artificial neural network (ANN) was composed of 58 input parameters, 10 hidden layers, and 3 output layers (normal, mild and moderate CKD, and severe CKD). Using the ANN model, the final classification rate was 95.4%, the epoch needed for training was 38 times, and the misclassification rate was 4.6%.

Face Mask Assistant: Detection of Face Mask Service Stage Based on Mobile Phone

Coronavirus Disease 2019 (COVID-19) has spread all over the world since it broke out massively in December 2019, which has caused a large loss to the whole world. Both the confirmed cases and death cases have reached a relatively frightening number. Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the cause of COVID-19, can be transmitted by small respiratory droplets. To curb its spread at the source, wearing masks is a convenient and effective measure. In most cases, people use face masks in a high-frequent but short-time way. Aimed at solving the problem that we do not know which service stage of the mask belongs to, we propose a detection system based on the mobile phone. We first extract four features from the gray level co-occurrence matrixes (GLCMs) of the face mask's micro-photos. Next, a three-result detection system is accomplished by using K Nearest Neighbor (KNN) algorithm. The results of validation experiments show that our system can reach an accuracy of 82.87% (measured by macro-measures) on the testing dataset. The precision of Type I 'normal use' and the recall of type III 'not recommended' reach 92.00% and 92.59%. In future work, we plan to expand the detection objects to more mask types. This work demonstrates that the proposed mobile microscope system can be used as an assistant for face mask being used, which may play a positive role in fighting against COVID-19.

Detection and classification of burnt skin via sparse representation of signals by over-redundant dictionaries

Skin burns in color images must be accurately detected and classified according to burn degree in order to assist clinicians during diagnosis and early treatment. Especially in emergency cases in which clinical experience might not be available to conduct a thorough examination with high accuracy, an automated assessment may benefit patient outcomes. In this work, detection and classification of burnt areas are performed by using the sparse representation of feature vectors by over-redundant dictionaries. Feature vectors are extracted from image patches so that each patch is assigned to a class representing a burn degree. Using color and texture information as features, detection and classification achieved 95.65% sensitivity and 94.02% precision. Experiments used two methods to build dictionaries for burn severity classes to apply to observed skin regions: (1) direct collection of feature vectors from patches in various images and locations and (2) collection of feature vectors followed by dictionary learning accompanied by K-singular value decomposition.

Impact of neoadjuvant androgen deprivation therapy on magnetic resonance imaging features in prostate cancer before radiotherapy

Background and purpose

In locally advanced prostate cancer (PC), androgen deprivation therapy (ADT) in combination with whole prostate radiotherapy (RT) is the standard treatment. ADT affects the prostate as well as the tumour on multiparametric magnetic resonance imaging (MRI) with decreased PC conspicuity and impaired localisation of the prostate lesion. Image texture analysis has been suggested to be of aid in separating tumour from normal tissue. The aim of the study was to investigate the impact of ADT on baseline defined MRI features in prostate cancer with the goal to investigate if it might be of use in radiotherapy planning.

Materials and methods

Fifty PC patients were included. Multiparametric MRI was performed before, and three months after ADT. At baseline, a tumour volume was delineated on apparent diffusion coefficient (ADC) maps with suspected tumour content and a reference volume in normal prostatic tissue. These volumes were transferred to MRIs after ADT and were analysed with first-order -and invariant Haralick -features.

Results

At baseline, the median value and several of the invariant Haralick features of ADC, showed a significant difference between tumour and reference volumes. After ADT, only ADC median value could significantly differentiate the two volumes.

Conclusions

Invariant Haralick -features could not distinguish between baseline MRI defined PC and normal tissue after ADT. First-order median value remained significantly different in tumour and reference volumes after ADT, but the difference was less pronounced than before ADT.

Classification of PPMI MRI scans with voxel-based morphometry and machine learning to assist in the diagnosis of Parkinson's disease

BACKGROUND AND OBJECTIVES

Qualitative and quantitative analyses of Magnetic Resonance Imaging (MRI) scans are carried out to study and understand Parkinson's Disease, the second most common neurodegenerative disorder in people at their 60's. Some quantitative analyses are based on the application of voxel-based morphometry (VBM) on magnetic resonance images to determine the regions of interest, within gray matter, where there is a loss of the nerve cells that generate dopamine. This loss of dopamine is indicative of Parkinson's disease. The purpose of this research is the introduction of a new method to classify the 3-D magnetic resonance scans of an individual, as an assisting tool for diagnosis of Parkinson's disease by using the largest MRI dataset (Parkinson's Progression Markers Initiative) from a population of patients with Parkinson's disease and control individuals. A contribution is that separate studies are conducted for men and women since gender plays a significant role within Neurobiology, which is demonstrated by the fact that men are more prone to Parkinson's disease than women are.

METHODS

Previous to classification, VBM is conducted on magnetic resonance images to detect the regions where features are extracted by using first- and second-order statistics methods. Furthermore, the number of features is considerably reduced by using feature selection techniques. Seven classifiers are used and we are conducting separate experiments for men and women.

RESULTS

The best detection performance achieved in men is 99.01% of accuracy, 99.35% of sensitivity, 100% of specificity, and 100% of precision. The best detection performance achieved in women is 96.97% of accuracy, 100% of sensitivity, 96.15% of specificity, and 97.22% of precision. During classification of magnetic resonance images, the corresponding computational complexity is reduced since few features are selected.

CONCLUSIONS

The proposed method provides high performance as an assisting tool in the diagnosis of Parkinson's disease, by conducting separate experiments in men and women. While previous works have focused their analysis to the striatum region of the brain (the largest nuclear complex of the basal ganglia), the proposed approach is based on analysis over the whole brain by looking for decreases of tissue thickness, with the consequence of finding other regions of interest such as the cortex.