Abstract:Raman microspectroscopy is a powerful tool for the analysis of tissue sections providing a molecular map of the investigated samples. Nevertheless, data pre processing, and particularly the removal of broad background to the spectra remains problematic. Indeed, the physical origin of the background has not been satisfactorily determined. Using 785nm as source in a confocal geometry, it is demonstrated, that for the example of the protein kappa-elastin, the background and resulting quality of the recorded spectrum is dependent on the morphology of the sample. Whereas a fine powder yields a dominant broad background, compressed pellets and solution cast thin films produce respectively improved quality spectra with significantly reduced spectral background. As the chemical composition of the samples is identical, the background is ascribed to stray light due to diffuse scattering rather than an intrinsic photoluminescence. Recorded spectra from tissue sample exhibit a large and spatially variable background, resulting in poorly defined spectral features. A significant reduction of the background signal and improvement of the spectral quality is achieved by immersion in water, and measurement with an immersion objective. The significant improvement in signal to background is attributed to a reduction of the diffuse scattering due to a change in the effective morphology as a result of an improved index matching between the water/ tissue interface compared to the air/tissue interface. Compared to sections measured in air, the background is reduced to that of the water, and preprocessing is reduced to the subtraction of the substrate and water signal, and correction for the instrument response, all of which are highly reproducible. Data preprocessing is thus greatly simplified and the results significantly more reliable.

Alzheimer’s disease (AD) may cause damage to the memory cells permanently, which results in the form of dementia. The diagnosis of Alzheimer’s disease at an early stage is a problematic task for researchers. For this, machine learning and deep convolutional neural network (CNN) based approaches are readily available to solve various problems related to brain image data analysis. In clinical research, magnetic resonance imaging (MRI) is used to diagnose AD. For accurate classification of dementia stages, we need highly discriminative features obtained from MRI images. Recently advanced deep CNN-based models successfully proved their accuracy. However, due to a smaller number of image samples available in the datasets, there exist problems of over-fitting hindering the performance of deep learning approaches. In this research, we developed a Siamese convolutional neural network (SCNN) model inspired by VGG-16 (also called Oxford Net) to classify dementia stages. In our approach, we extend the insufficient and imbalanced data by using augmentation approaches. Experiments are performed on a publicly available dataset open access series of imaging studies (OASIS), by using the proposed approach, an excellent test accuracy of 99.05% is achieved for the classification of dementia stages. We compared our model with the state-of-the-art models and discovered that the proposed model outperformed the state-of-the-art models in terms of performance, efficiency, and accuracy.

Brain tumors have become a leading cause of death around the globe. The main reason for this epidemic is the difficulty conducting a timely diagnosis of the tumor. Fortunately, magnetic resonance images (MRI) are utilized to diagnose tumors in most cases. The performance of a Convolutional Neural Network (CNN) depends on many factors (i.e., weight initialization, optimization, batches and epochs, learning rate, activation function, loss function, and network topology), data quality, and specific combinations of these model attributes. When we deal with a segmentation or classification problem, utilizing a single optimizer is considered weak testing or validity unless the decision of the selection of an optimizer is backed up by a strong argument. Therefore, optimizer selection processes are considered important to validate the usage of a single optimizer in order to attain these decision problems. In this paper, we provides a comprehensive comparative analysis of popular optimizers of CNN to benchmark the segmentation for improvement. In detail, we perform a comparative analysis of 10 different state-of-the-art gradient descent-based optimizers, namely Adaptive Gradient (Adagrad), Adaptive Delta (AdaDelta), Stochastic Gradient Descent (SGD), Adaptive Momentum (Adam), Cyclic Learning Rate (CLR), Adaptive Max Pooling (Adamax), Root Mean Square Propagation (RMS Prop), Nesterov Adaptive Momentum (Nadam), and Nesterov accelerated gradient (NAG) for CNN. The experiments were performed on the BraTS2015 data set. The Adam optimizer had the best accuracy of 99.2% in enhancing the CNN ability in classification and segmentation.

The rapid growth of electronic documents are causing problems like unstructured data that need more time and effort to search a relevant document. Text Document Classification (TDC) has a great significance in information processing and retrieval where unstructured documents are organized into predefined classes. Urdu is the most favorite research language in South Asian languages because of its complex morphology, unique features, and lack of linguistic resources like standard datasets. As compared to short text, like sentiment analysis, long text classification needs more time and effort because of large vocabulary, more noise, and redundant information. Machine Learning (ML) and Deep Learning (DL) models have been widely used in text processing. Despite the major limitations of ML models, like learn directed features, these are the favorite methods for Urdu TDC. To the best of our knowledge, it is the first study of Urdu TDC using DL model. In this paper, we design a large multipurpose and multi-format dataset that contain more than ten thousand documents organize into six classes. We use Single-layer Multisize Filters Convolutional Neural Network (SMFCNN) for classification and compare its performance with sixteen ML baseline models on three imbalanced datasets of various sizes. Further, we analyze the effects of preprocessing methods on SMFCNN performance. SMFCNN outperformed the baseline classifiers and achieved 95.4%, 91.8%, and 93.3% scores of accuracy on medium, large and small size dataset respectively. The designed dataset would be publically and freely available in different formats for future research in Urdu text processing. INDEX TERMS Convolutional neural network, deep learning, machine learning, natural language processing, text document classification, Urdu text classification.

No abstract

Highway agencies are continually facing safety problems on highways, especially on horizontal alignments. Traditionally, the geometric design implicitly considers safety through satisfying minimum design requirements for different geometric elements. This article presents a new substantive-safety approach for the design of horizontal alignments based not only on minimum design guidelines, but also on actual collision experience. The curve radii, spiral lengths, lane width, shoulder width, and tangent lengths are determined to optimize the mean collision frequency along the highway. The model allows the parameters of the horizontal alignment to vary within specified ranges. The model also considers any specified physical obstructions in selecting the optimal alignment. Collision experience is addressed using existing collision prediction models for horizontal alignments and cross sections. The model is applicable to two-lane rural highways for which collision prediction models exist. Application of the model is presented using numerical examples. The proposed substantive-safety approach takes horizontal alignment design one step further beyond the minimum-guideline concept, and therefore should be of interest to highway designers.

No abstract

One of the main objectives to promote traffic safety at roundabouts is design consistency. Design consistency ensures that the speed differences along a vehicle path or between conflicting paths are smaller than a specified criterion. Traditionally, roundabout design involves an iterative process. Initially, the vehicle path radii are determined from drawing each path by freehand on the proposed roundabout geometry, and then the speeds along path elements are calculated. The process is repeated until design consistency is satisfied. This paper presents an optimization model that directly provides the design parameters that maximize design consistency. The radii of different vehicle paths (through, right, and left) are mathematically modeled for given design parameters. The optimum design parameters include vehicle path radii, approach entry widths, inscribed circle diameter, circulating width, and central island diameter. The optimization model is applicable to single-lane roundabouts with four legs intersecting at right angles. The proposed model not only makes the design process more efficient, but also guarantees the optimum design consistency.Key words: geometric design, roundabouts, horizontal curve radius, consistency, safety, optimization.