CT-Based Radiomics to Predict KRAS Mutation in CRC Patients Using a Machine Learning Algorithm: A Retrospective Study.

Colorectal cancer (CRC) is one of the most common types of cancer worldwide. The KRAS mutation is present in 30-50% of CRC patients. This mutation confers resistance to treatment with anti-EGFR therapy. This article aims at proving that computer tomography (CT)-based radiomics can predict the KRAS mutation in CRC patients. The piece is a retrospective study with 56 CRC patients from the Hospital of Santiago de Compostela, Spain. All patients had a confirmatory pathological analysis of the KRAS status. Radiomics features were obtained using an abdominal contrast enhancement CT (CECT) before applying any treatments. We used several classifiers, including AdaBoost, neural network, decision tree, support vector machine, and random forest, to predict the presence or absence of KRAS mutation. The most reliable prediction was achieved using the AdaBoost ensemble on clinical patient data, with a kappa and accuracy of 53.7% and 76.8%, respectively. The sensitivity and specificity were 73.3% and 80.8%. Using texture descriptors, the best accuracy and kappa were 73.2% and 46%, respectively, with sensitivity and specificity of 76.7% and 69.2%, also showing a correlation between texture patterns on CT images and KRAS mutation. Radiomics could help manage CRC patients, and in the future, it could have a crucial role in diagnosing CRC patients ahead of invasive methods.

Real-time botnet detection on large network bandwidths using machine learning.

Botnets are one of the most harmful cyberthreats, that can perform many types of cyberattacks and cause billionaire losses to the global economy. Nowadays, vast amounts of network traffic are generated every second, hence manual analysis is impossible. To be effective, automatic botnet detection should be done as fast as possible, but carrying this out is difficult in large bandwidths. To handle this problem, we propose an approach that is capable of carrying out an ultra-fast network analysis (i.e. on windows of one second), without a significant loss in the F1-score. We compared our model with other three literature proposals, and achieved the best performance: an F1 score of 0.926 with a processing time of 0.007 ms per sample. We also assessed the robustness of our model on saturated networks and on large bandwidths. In particular, our model is capable of working on networks with a saturation of 10% of packet loss, and we estimated the number of CPU cores needed to analyze traffic on three bandwidth sizes. Our results suggest that using commercial-grade cores of 2.4 GHz, our approach would only need four cores for bandwidths of 100 Mbps and 1 Gbps, and 19 cores on 10 Gbps networks.

Identification of the presence of ischaemic stroke lesions by means of texture analysis on brain magnetic resonance images.

BACKGROUND: The differential quantification of brain atrophy, white matter hyperintensities (WMH) and stroke lesions is important in studies of stroke and dementia. However, the presence of stroke lesions is usually overlooked by automatic neuroimage processing methods and the-state-of-the-art deep learning schemes, which lack sufficient annotated data. We explore the use of radiomics in identifying whether a brain magnetic resonance imaging (MRI) scan belongs to an individual that had a stroke or not. MATERIALS AND METHODS: We used 1800 3D sets of MRI data from three prospective studies: one of stroke mechanisms and two of cognitive ageing, evaluated 114 textural features in WMH, cerebrospinal fluid, deep grey and normal-appearing white matter, and attempted to classify the scans using a random forest and support vector machine classifiers with and without feature selection. We evaluated the discriminatory power of each feature independently in each population and corrected the result against Type 1 errors. We also evaluated the influence of clinical parameters in the classification results. RESULTS: Subtypes of ischaemic strokes (i.e. lacunar vs. cortical) cannot be discerned using radiomics, but the presence of a stroke-type lesion can be ascertained with accuracies ranging from 0.7 < AUC < 0.83. Feature selection, tissue type, stroke subtype and MRI sequence did not seem to determine the classification results. From all clinical variables evaluated, age correlated with the proportion of images classified correctly using either different or the same descriptors (Pearson r = 0.31 and 0.39 respectively, p < 0.001). CONCLUSIONS: Texture features in conventionally automatically segmented tissues may help in the identification of the presence of previous stroke lesions on an MRI scan, and should be taken into account in transfer learning strategies of the-state-of-the-art deep learning schemes.

A protocol for precise comparisons of small vessel disease lesions between ex vivo magnetic resonance imaging and histopathology.

RATIONALE: Neuroimaging and clinical studies have defined human sporadic cerebral small vessel disease but the pathophysiology remains relatively poorly understood. To develop effective therapies and preventative strategies, we must better understand the heterogeneity and development of small vessel disease at a cellular level. HYPOTHESIS: Small vessel disease lesions as seen on neuroimaging have specific neuropathological correlates. METHODS AND DESIGN: Standard histological samples are taken from strategic areas of the brain typically affected by small vessel disease, in cases with a range of disease from mild to severe and controls. Tissue is formalin fixed, scanned using 7-tesla magnetic resonance imaging and processed for histology. Histological slides are digitalized then registered with the corresponding magnetic resonance image. Small vessel disease burden is assessed and lesions are precisely identified on the ex vivo imaging and microscopy independently then compared. The tissue can be interrogated using multiple magnetic resonance sequences and histological methods targeting the gliovascular unit. STUDY OUTCOMES: The primary outcome is identifying and defining the cellular characteristics of small vessel disease lesions compared to imaging. Secondary outcomes are related to obtaining information about abnormalities of protein expression in the gliovascular unit, defining groups of small vessel disease severity in our cohorts for future analysis and developing a reliable, reproducible protocol for accurate radiological-histological lesion comparison, which can be applied to other neurological diseases in the future. DISCUSSION: Comprehensive, precise pathological-radiological-clinical correlations in small vessel disease will provide greater insight into associations and pathophysiology underlying magnetic resonance imaging findings in normal- and abnormal-appearing tissue, ex vivo and in vivo.

Machine learning of neuroimaging for assisted diagnosis of cognitive impairment and dementia: A systematic review.

INTRODUCTION: Advanced machine learning methods might help to identify dementia risk from neuroimaging, but their accuracy to date is unclear. METHODS: We systematically reviewed the literature, 2006 to late 2016, for machine learning studies differentiating healthy aging from dementia of various types, assessing study quality, and comparing accuracy at different disease boundaries. RESULTS: Of 111 relevant studies, most assessed Alzheimer's disease versus healthy controls, using AD Neuroimaging Initiative data, support vector machines, and only T1-weighted sequences. Accuracy was highest for differentiating Alzheimer's disease from healthy controls and poor for differentiating healthy controls versus mild cognitive impairment versus Alzheimer's disease or mild cognitive impairment converters versus nonconverters. Accuracy increased using combined data types, but not by data source, sample size, or machine learning method. DISCUSSION: Machine learning does not differentiate clinically relevant disease categories yet. More diverse data sets, combinations of different types of data, and close clinical integration of machine learning would help to advance the field.

Application of Texture Analysis to Study Small Vessel Disease and Blood-Brain Barrier Integrity.

OBJECTIVES: We evaluate the alternative use of texture analysis for evaluating the role of blood-brain barrier (BBB) in small vessel disease (SVD). METHODS: We used brain magnetic resonance imaging from 204 stroke patients, acquired before and 20 min after intravenous gadolinium administration. We segmented tissues, white matter hyperintensities (WMH) and applied validated visual scores. We measured textural features in all tissues pre- and post-contrast and used ANCOVA to evaluate the effect of SVD indicators on the pre-/post-contrast change, Kruskal-Wallis for significance between patient groups and linear mixed models for pre-/post-contrast variations in cerebrospinal fluid (CSF) with Fazekas scores. RESULTS: Textural "homogeneity" increase in normal tissues with higher presence of SVD indicators was consistently more overt than in abnormal tissues. Textural "homogeneity" increased with age, basal ganglia perivascular spaces scores (p < 0.01) and SVD scores (p < 0.05) and was significantly higher in hypertensive patients (p < 0.002) and lacunar stroke (p = 0.04). Hypertension (74% patients), WMH load (median = 1.5 ± 1.6% of intracranial volume), and age (mean = 65.6 years, SD = 11.3) predicted the pre/post-contrast change in normal white matter, WMH, and index stroke lesion. CSF signal increased with increasing SVD post-contrast. CONCLUSION: A consistent general pattern of increasing textural "homogeneity" with increasing SVD and post-contrast change in CSF with increasing WMH suggest that texture analysis may be useful for the study of BBB integrity.

Reliability of an automatic classifier for brain enlarged perivascular spaces burden and comparison with human performance.

In the brain, enlarged perivascular spaces (PVS) relate to cerebral small vessel disease (SVD), poor cognition, inflammation and hypertension. We propose a fully automatic scheme that uses a support vector machine (SVM) to classify the burden of PVS in the basal ganglia (BG) region as low or high. We assess the performance of three different types of descriptors extracted from the BG region in T2-weighted MRI images: (i) statistics obtained from Wavelet transform's coefficients, (ii) local binary patterns and (iii) bag of visual words (BoW) based descriptors characterizing local keypoints obtained from a dense grid with the scale-invariant feature transform (SIFT) characteristics. When the latter were used, the SVM classifier achieved the best accuracy (81.16%). The output from the classifier using the BoW descriptors was compared with visual ratings done by an experienced neuroradiologist (Observer 1) and by a trained image analyst (Observer 2). The agreement and cross-correlation between the classifier and Observer 2 (κ = 0.67 (0.58-0.76)) were slightly higher than between the classifier and Observer 1 (κ = 0.62 (0.53-0.72)) and comparable between both the observers (κ = 0.68 (0.61-0.75)). Finally, three logistic regression models using clinical variables as independent variable and each of the PVS ratings as dependent variable were built to assess how clinically meaningful were the predictions of the classifier. The goodness-of-fit of the model for the classifier was good (area under the curve (AUC) values: 0.93 (model 1), 0.90 (model 2) and 0.92 (model 3)) and slightly better (i.e. AUC values: 0.02 units higher) than that of the model for Observer 2. These results suggest that, although it can be improved, an automatic classifier to assess PVS burden from brain MRI can provide clinically meaningful results close to those from a trained observer.

On the computational assessment of white matter hyperintensity progression: difficulties in method selection and bias field correction performance on images with significant white matter pathology.

INTRODUCTION: Subtle inhomogeneities in the scanner's magnetic fields (B0 and B1) alter the intensity levels of the structural magnetic resonance imaging (MRI) affecting the volumetric assessment of WMH changes. Here, we investigate the influence that (1) correcting the images for the B1 inhomogeneities (i.e. bias field correction (BFC)) and (2) selection of the WMH change assessment method can have on longitudinal analyses of WMH progression and discuss possible solutions. METHODS: We used brain structural MRI from 46 mild stroke patients scanned at stroke onset and 3 years later. We tested three BFC approaches: FSL-FAST, N4 and exponentially entropy-driven homomorphic unsharp masking (E(2)D-HUM) and analysed their effect on the measured WMH change. Separately, we tested two methods to assess WMH changes: measuring WMH volumes independently at both time points semi-automatically (MCMxxxVI) and subtracting intensity-normalised FLAIR images at both time points following image gamma correction. We then combined the BFC with the computational method that performed best across the whole sample to assess WMH changes. RESULTS: Analysis of the difference in the variance-to-mean intensity ratio in normal tissue between BFC and uncorrected images and visual inspection showed that all BFC methods altered the WMH appearance and distribution, but FSL-FAST in general performed more consistently across the sample and MRI modalities. The WMH volume change over 3 years obtained with MCMxxxVI with vs. without FSL-FAST BFC did not significantly differ (medians(IQR)(with BFC) = 3.2(6.3) vs. 2.9(7.4)ml (without BFC), p = 0.5), but both differed significantly from the WMH volume change obtained from subtracting post-processed FLAIR images (without BFC)(7.6(8.2)ml, p < 0.001). This latter method considerably inflated the WMH volume change as subtle WMH at baseline that became more intense at follow-up were counted as increase in the volumetric change. CONCLUSIONS: Measurement of WMH volume change remains challenging. Although the overall volumetric change was not significantly affected by the application of BFC, these methods distorted the image intensity distribution affecting subtle WMH. Subtracting the FLAIR images at both time points following gamma correction seems a promising technique but is adversely affected by subtle WMH. It is important to take into account not only the changes in volume but also in the signal intensity.

Texture and moments-based classification of the acrosome integrity of boar spermatozoa images.

The automated assessment of the sperm quality is an important challenge in the veterinary field. In this paper, we explore how to describe the acrosomes of boar spermatozoa using image analysis so that they can be automatically categorized as intact or damaged. Our proposal aims at characterizing the acrosomes by means of texture features. The texture is described using first order statistics and features derived from the co-occurrence matrix of the image, both computed from the original image and from the coefficients yielded by the Discrete Wavelet Transform. Texture descriptors are evaluated and compared with moments-based descriptors in terms of the classification accuracy they provide. Experimental results with a Multilayer Perceptron and the k-Nearest Neighbours classifiers show that texture descriptors outperform moment-based descriptors, reaching an accuracy of 94.93%, which makes this approach very attractive for the veterinarian community.