Alzheimer's diagnosis using deep learning in segmenting and classifying 3D brain MR images.

BACKGROUND AND OBJECTIVES: Dementia is one of the brain diseases with serious symptoms such as memory loss, and thinking problems. According to the World Alzheimer Report 2016, in the world, there are 47 million people having dementia and it can be 131 million by 2050. There is no standard method to diagnose dementia, and consequently unable to access the treatment effectively. Hence, the computational diagnosis of the disease from brain Magnetic Resonance Image (MRI) scans plays an important role in supporting the early diagnosis. Alzheimer's Disease (AD), a common type of Dementia, includes problems related to disorientation, mood swings, not managing self-care, and behavioral issues. In this article, we present a new computational method to diagnosis Alzheimer's disease from 3D brain MR images. METHODS: An efficient approach to diagnosis Alzheimer's disease from brain MRI scans is proposed comprising two phases: I) segmentation and II) classification, both based on deep learning. After the brain tissues are segmented by a model that combines Gaussian Mixture Model (GMM) and Convolutional Neural Network (CNN), a new model combining Extreme Gradient Boosting (XGBoost) and Support Vector Machine (SVM) is used to classify Alzheimer's disease based on the segmented tissues. RESULTS: We present two evaluations for segmentation and classification. For comparison, the new method was evaluated using the AD-86 and AD-126 datasets leading to Dice 0.96 for segmentation in both datasets and accuracies 0.88, and 0.80 for classification, respectively. CONCLUSION: Deep learning gives prominent results for segmentation and feature extraction in medical image processing. The combination of XGboost and SVM improves the results obtained.

Convolutional Neural Networks and Heuristic Methods for Crowd Counting: A Systematic Review.

The crowd counting task has become a pillar for crowd control as it provides information concerning the number of people in a scene. It is helpful in many scenarios such as video surveillance, public safety, and future event planning. To solve such tasks, researchers have proposed different solutions. In the beginning, researchers went with more traditional solutions, while recently the focus is on deep learning methods and, more specifically, on Convolutional Neural Networks (CNNs), because of their efficiency. This review explores these methods by focusing on their key differences, advantages, and disadvantages. We have systematically analyzed algorithms and works based on the different models suggested and the problems they are trying to solve. The main focus is on the shift made in the history of crowd counting methods, moving from the heuristic models to CNN models by identifying each category and discussing its different methods and architectures. After a deep study of the literature on crowd counting, the survey partitions current datasets into sparse and crowded ones. It discusses the reviewed methods by comparing their results on the different datasets. The findings suggest that the heuristic models could be even more effective than the CNN models in sparse scenarios.

Using Deep Neural Networks for Human Fall Detection Based on Pose Estimation.

Requests for caring for and monitoring the health and safety of older adults are increasing nowadays and form a topic of great social interest. One of the issues that lead to serious concerns is human falls, especially among aged people. Computer vision techniques can be used to identify fall events, and Deep Learning methods can detect them with optimum accuracy. Such imaging-based solutions are a good alternative to body-worn solutions. This article proposes a novel human fall detection solution based on the Fast Pose Estimation method. The solution uses Time-Distributed Convolutional Long Short-Term Memory (TD-CNN-LSTM) and 1Dimentional Convolutional Neural Network (1D-CNN) models, to classify the data extracted from image frames, and achieved high accuracies: 98 and 97% for the 1D-CNN and TD-CNN-LSTM models, respectively. Therefore, by applying the Fast Pose Estimation method, which has not been used before for this purpose, the proposed solution is an effective contribution to accurate human fall detection, which can be deployed in edge devices due to its low computational and memory demands.

9-DOF IMU-Based Attitude and Heading Estimation Using an Extended Kalman Filter with Bias Consideration.

The attitude and heading reference system (AHRS) is an important concept in the area of navigation, image stabilization, and object detection and tracking. Many studies and works have been conducted in this regard to estimate the accurate orientation of rigid bodies. In most research in this area, low-cost MEMS sensors are employed, but since the system's response will diverge over time due to integration drift, it is necessary to apply proper estimation algorithms. A two-step extended Kalman Filter (EKF) algorithm is used in this study to estimate the orientation of an IMU. A 9-DOF device is used for this purpose, including a 6-DOF IMU with a three-axis gyroscope and a three-axis accelerometer, and a three-axis magnetometer. In addition, to have an accurate algorithm, both IMU and magnetometer biases and disturbances are modeled and considered in the real-time filter. After applying the algorithm to the sensor's output, an accurate orientation as well as unbiased angular velocity, linear acceleration, and magnetic field were achieved. In order to demonstrate the reduction of noise power, fast Fourier transform (FFT) diagrams are used. The effect of the initial condition on the response of the system is also investigated.

Sound Classification and Processing of Urban Environments: A Systematic Literature Review.

Audio recognition can be used in smart cities for security, surveillance, manufacturing, autonomous vehicles, and noise mitigation, just to name a few. However, urban sounds are everyday audio events that occur daily, presenting unstructured characteristics containing different genres of noise and sounds unrelated to the sound event under study, making it a challenging problem. Therefore, the main objective of this literature review is to summarize the most recent works on this subject to understand the current approaches and identify their limitations. Based on the reviewed articles, it can be realized that Deep Learning (DL) architectures, attention mechanisms, data augmentation techniques, and pretraining are the most crucial factors to consider while creating an efficient sound classification model. The best-found results were obtained by Mushtaq and Su, in 2020, using a DenseNet-161 with pretrained weights from ImageNet, and NA-1 and NA-2 as augmentation techniques, which were of 97.98%, 98.52%, and 99.22% for UrbanSound8K, ESC-50, and ESC-10 datasets, respectively. Nonetheless, the use of these models in real-world scenarios has not been properly addressed, so their effectiveness is still questionable in such situations.

Transformers for Urban Sound Classification-A Comprehensive Performance Evaluation.

Many relevant sound events occur in urban scenarios, and robust classification models are required to identify abnormal and relevant events correctly. These models need to identify such events within valuable time, being effective and prompt. It is also essential to determine for how much time these events prevail. This article presents an extensive analysis developed to identify the best-performing model to successfully classify a broad set of sound events occurring in urban scenarios. Analysis and modelling of Transformer models were performed using available public datasets with different sets of sound classes. The Transformer models' performance was compared to the one achieved by the baseline model and end-to-end convolutional models. Furthermore, the benefits of using pre-training from image and sound domains and data augmentation techniques were identified. Additionally, complementary methods that have been used to improve the models' performance and good practices to obtain robust sound classification models were investigated. After an extensive evaluation, it was found that the most promising results were obtained by employing a Transformer model using a novel Adam optimizer with weight decay and transfer learning from the audio domain by reusing the weights from AudioSet, which led to an accuracy score of 89.8% for the UrbanSound8K dataset, 95.8% for the ESC-50 dataset, and 99% for the ESC-10 dataset, respectively.

Binaural Acoustic Scene Classification Using Wavelet Scattering, Parallel Ensemble Classifiers and Nonlinear Fusion.

The analysis of ambient sounds can be very useful when developing sound base intelligent systems. Acoustic scene classification (ASC) is defined as identifying the area of a recorded sound or clip among some predefined scenes. ASC has huge potential to be used in urban sound event classification systems. This research presents a hybrid method that includes a novel mathematical fusion step which aims to tackle the challenges of ASC accuracy and adaptability of current state-of-the-art models. The proposed method uses a stereo signal, two ensemble classifiers (random subspace), and a novel mathematical fusion step. In the proposed method, a stable, invariant signal representation of the stereo signal is built using Wavelet Scattering Transform (WST). For each mono, i.e., left and right, channel, a different random subspace classifier is trained using WST. A novel mathematical formula for fusion step was developed, its parameters being found using a Genetic algorithm. The results on the DCASE 2017 dataset showed that the proposed method has higher classification accuracy (about 95%), pushing the boundaries of existing methods.

Classification of Heart Sounds Using Chaogram Transform and Deep Convolutional Neural Network Transfer Learning.

Heart sounds convey important information regarding potential heart diseases. Currently, heart sound classification attracts many researchers from the fields of telemedicine, digital signal processing, and machine learning-among others-mainly to identify cardiac pathology as quickly as possible. This article proposes chaogram as a new transform to convert heart sound signals to colour images. In the proposed approach, the output image is, therefore, the projection of the reconstructed phase space representation of the phonocardiogram (PCG) signal on three coordinate planes. This has two major benefits: (1) it makes possible to apply deep convolutional neural networks to heart sounds and (2) it is also possible to employ a transfer learning scheme by converting a heart sound signal to an image. The performance of the proposed approach was verified on the PhysioNet dataset. Due to the imbalanced data on this dataset, it is common to assess the results quality using the average of sensitivity and specificity, which is known as score, instead of accuracy. In this study, the best results were achieved using the InceptionV3 model, which achieved a score of 88.06%.

Weakly Supervised Video Anomaly Detection Based on 3D Convolution and LSTM.

Weakly supervised video anomaly detection is a recent focus of computer vision research thanks to the availability of large-scale weakly supervised video datasets. However, most existing research works are limited to the frame-level classification with emphasis on finding the presence of specific objects or activities. In this article, a new neural network architecture is proposed to efficiently extract the prominent features for detecting whether a video contains anomalies. A video is treated as an integral input and the detection follows the procedure of video-label assignment. The extraction of spatial and temporal features is carried out by three-dimensional convolutions, and then their relationship is further modeled using an LSTM network. The concise structure of the proposed method enables high computational efficiency, and extensive experiments demonstrate its effectiveness.

Urban Safety: An Image-Processing and Deep-Learning-Based Intelligent Traffic Management and Control System.

With the rapid growth and development of cities, Intelligent Traffic Management and Control (ITMC) is becoming a fundamental component to address the challenges of modern urban traffic management, where a wide range of daily problems need to be addressed in a prompt and expedited manner. Issues such as unpredictable traffic dynamics, resource constraints, and abnormal events pose difficulties to city managers. ITMC aims to increase the efficiency of traffic management by minimizing the odds of traffic problems, by providing real-time traffic state forecasts to better schedule the intersection signal controls. Reliable implementations of ITMC improve the safety of inhabitants and the quality of life, leading to economic growth. In recent years, researchers have proposed different solutions to address specific problems concerning traffic management, ranging from image-processing and deep-learning techniques to forecasting the traffic state and deriving policies to control intersection signals. This review article studies the primary public datasets helpful in developing models to address the identified problems, complemented with a deep analysis of the works related to traffic state forecast and intersection-signal-control models. Our analysis found that deep-learning-based approaches for short-term traffic state forecast and multi-intersection signal control showed reasonable results, but lacked robustness for unusual scenarios, particularly during oversaturated situations, which can be resolved by explicitly addressing these cases, potentially leading to significant improvements of the systems overall. However, there is arguably a long path until these models can be used safely and effectively in real-world scenarios.

Diagnosis of Leukaemia in Blood Slides Based on a Fine-Tuned and Highly Generalisable Deep Learning Model.

Leukaemia is a dysfunction that affects the production of white blood cells in the bone marrow. Young cells are abnormally produced, replacing normal blood cells. Consequently, the person suffers problems in transporting oxygen and in fighting infections. This article proposes a convolutional neural network (CNN) named LeukNet that was inspired on convolutional blocks of VGG-16, but with smaller dense layers. To define the LeukNet parameters, we evaluated different CNNs models and fine-tuning methods using 18 image datasets, with different resolution, contrast, colour and texture characteristics. We applied data augmentation operations to expand the training dataset, and the 5-fold cross-validation led to an accuracy of 98.61%. To evaluate the CNNs generalisation ability, we applied a cross-dataset validation technique. The obtained accuracies using cross-dataset experiments on three datasets were 97.04, 82.46 and 70.24%, which overcome the accuracies obtained by current state-of-the-art methods. We conclude that using the most common and deepest CNNs may not be the best choice for applications where the images to be classified differ from those used in pre-training. Additionally, the adopted cross-dataset validation approach proved to be an excellent choice to evaluate the generalisation capability of a model, as it considers the model performance on unseen data, which is paramount for CAD systems.

Versatile Convolutional Networks Applied to Computed Tomography and Magnetic Resonance Image Segmentation.

Medical image segmentation has seen positive developments in recent years but remains challenging with many practical obstacles to overcome. The applications of this task are wide-ranging in many fields of medicine, and used in several imaging modalities which usually require tailored solutions. Deep learning models have gained much attention and have been lately recognized as the most successful for automated segmentation. In this work we show the versatility of this technique by means of a single deep learning architecture capable of successfully performing segmentation on two very different types of imaging: computed tomography and magnetic resonance. The developed model is fully convolutional with an encoder-decoder structure and high-resolution pathways which can process whole three-dimensional volumes at once, and learn directly from the data to find which voxels belong to the regions of interest and localize those against the background. The model was applied to two publicly available datasets achieving equivalent results for both imaging modalities, as well as performing segmentation of different organs in different anatomic regions with comparable success.

Deep Learning in Radiation Oncology Treatment Planning for Prostate Cancer: A Systematic Review.

Radiation oncology for prostate cancer is important as it can decrease the morbidity and mortality associated with this disease. Planning for this modality of treatment is both fundamental, time-consuming and prone to human-errors, leading to potentially avoidable delays in start of treatment. A fundamental step in radiotherapy planning is contouring of radiation targets, where medical specialists contouring, i.e., segment, the boundaries of the structures to be irradiated. Automating this step can potentially lead to faster treatment planning without a decrease in quality, while increasing time available to physicians and also more consistent treatment results. This can be framed as an image segmentation task, which has been studied for many decades in the fields of Computer Vision and Machine Learning. With the advent of Deep Learning, there have been many proposals for different network architectures achieving high performance levels. In this review, we searched the literature for those methods and describe them briefly, grouping those based on Computed Tomography (CT) or Magnetic Resonance Imaging (MRI). This is a booming field, evidenced by the date of the publications found. However, most publications use data from a very limited number of patients, which presents an obstacle to deep learning models training. Although the performance of the models has achieved very satisfactory results, there is still room for improvement, and there is arguably a long way before these models can be used safely and effectively in clinical practice.

Extraction, selection and comparison of features for an effective automated computer-aided diagnosis of Parkinson's disease based on [123I]FP-CIT SPECT images.

PURPOSE: This work aimed to assess the potential of a set of features extracted from [123I]FP-CIT SPECT brain images to be used in the computer-aided "in vivo" confirmation of dopaminergic degeneration and therefore to assist clinical decision to diagnose Parkinson's disease. METHODS: Seven features were computed from each brain hemisphere: five standard features related to uptake ratios on the striatum and two features related to the estimated volume and length of the striatal region with normal uptake. The features were tested on a dataset of 652 [123I]FP-CIT SPECT brain images from the Parkinson's Progression Markers Initiative. The discrimination capacities of each feature individually and groups of features were assessed using three different machine learning techniques: support vector machines (SVM), k-nearest neighbors and logistic regression. RESULTS: Cross-validation results based on SVM have shown that, individually, the features that generated the highest accuracies were the length of the striatal region (96.5%), the putaminal binding potential (95.4%) and the striatal binding potential (93.9%) with no statistically significant differences among them. The highest classification accuracy was obtained using all features simultaneously (accuracy 97.9%, sensitivity 98% and specificity 97.6%). Generally, slightly better results were obtained using the SVM with no statistically significant difference to the other classifiers for most of the features. CONCLUSIONS: The length of the striatal region uptake is clinically useful and highly valuable to confirm dopaminergic degeneration "in vivo" as an aid to the diagnosis of Parkinson's disease. It compares fairly well to the standard uptake ratio-based features, reaching, at least, similar accuracies and is easier to obtain automatically. Thus, we propose its day to day clinical use, jointly with the uptake ratio-based features, in the computer-aided diagnosis of dopaminergic degeneration in Parkinson's disease.

Medical cyber-physical systems: A survey.

Medical cyber-physical systems (MCPS) are healthcare critical integration of a network of medical devices. These systems are progressively used in hospitals to achieve a continuous high-quality healthcare. The MCPS design faces numerous challenges, including inoperability, security/privacy, and high assurance in the system software. In the current work, the infrastructure of the cyber-physical systems (CPS) are reviewed and discussed. This article enriched the researches of the networked Medical Device (MD) systems to increase the efficiency and safety of the healthcare. It also can assist the specialists of medical device to overcome crucial issues related to medical devices, and the challenges facing the design of the medical device's network. The concept of the social networking and its security along with the concept of the wireless sensor networks (WSNs) are addressed. Afterward, the CPS systems and platforms have been established, where more focus was directed toward CPS-based healthcare. The big data framework of CPSs is also included.

The role of the ipsilesional side in the rehabilitation of post-stroke subjects.

Most stroke lesions occur in the middle cerebral artery territory, presenting a high probability of damage of pathways with predominant ipsilesional disposition, mainly related to postural control. Despite the high probability of bilateral postural control dysfunction based on neuroanatomical fundaments, both research and clinical rehabilitation involving stroke subjects have been focused on contralesional side (also named affected side) impairments, while ipsilesional side (also named non-affected side) impairments have been attributed to an adaptive strategy. This paper aims to present a critical understanding about the state-of-the-art that sustains the hypothesis that stroke subjects with middle cerebral artery territory lesion at the subcortical level show an atypical behaviour in the ipsilateral side associated with the lesion itself and the possible implications.

Cognitive predictors of balance in Parkinson's disease.

Postural instability is one of the most incapacitating symptoms of Parkinson's disease (PD) and appears to be related to cognitive deficits. This study aims to determine the cognitive factors that can predict deficits in static and dynamic balance in individuals with PD. A sociodemographic questionnaire characterized 52 individuals with PD for this work. The Trail Making Test, Rule Shift Cards Test, and Digit Span Test assessed the executive functions. The static balance was assessed using a plantar pressure platform, and dynamic balance was based on the Timed Up and Go Test. The results were statistically analysed using SPSS Statistics software through linear regression analysis. The results show that a statistically significant model based on cognitive outcomes was able to explain the variance of motor variables. Also, the explanatory value of the model tended to increase with the addition of individual and clinical variables, although the resulting model was not statistically significant The model explained 25-29% of the variability of the Timed Up and Go Test, while for the anteroposterior displacement it was 23-34%, and for the mediolateral displacement it was 24-39%. From the findings, we conclude that the cognitive performance, especially the executive functions, is a predictor of balance deficit in individuals with PD.

Effects of dual-task training on balance and executive functions in Parkinson's disease: A pilot study.

The aim of this study was to analyze the efficacy of cognitive-motor dual-task training compared with single-task training on balance and executive functions in individuals with Parkinson's disease. Fifteen subjects, aged between 39 and 75 years old, were randomly assigned to the dual-task training group (n = 8) and single-task training group (n = 7). The training was run twice a week for 6 weeks. The single-task group received balance training and the dual-task group performed cognitive tasks simultaneously with the balance training. There were no significant differences between the two groups at baseline. After the intervention, the results for mediolateral sway with eyes closed were significantly better for the dual-task group and anteroposterior sway with eyes closed was significantly better for the single-task group. The results suggest superior outcomes for the dual-task training compared to the single-task training for static postural control, except in anteroposterior sway with eyes closed.

Ankle antagonist coactivation in the double-support phase of walking: Stroke vs. healthy subjects.

INTRODUCTION: Lesions in ipsilateral systems related to postural control in the ipsilesional side may justify the lower performance of stroke subjects during walking. PURPOSE: To analyze bilateral ankle antagonist coactivation during double support in stroke subjects. METHODS: Sixteen (8 females; 8 males) subjects with a first isquemic stroke and 22 controls (12 females; 10 males) participated in this study. The double-support phase was assessed through ground reaction forces and the electromyography of ankle muscles was assessed in both limbs. RESULTS: The ipsilesional limb presented statistically significant differences from the control when assuming specific roles during double support. The tibialis anterior and soleus pair was the one in which this atypical behavior was more pronounced. CONCLUSION: The ipsilesional limb presents a dysfunctional behavior when a higher postural control activity was demanded.

Ultrasonic sensor signals and optimum path forest classifier for the microstructural characterization of thermally-aged inconel 625 alloy.

Secondary phases, such as laves and carbides, are formed during the final solidification stages of nickel-based superalloy coatings deposited during the gas tungsten arc welding cold wire process. However, when aged at high temperatures, other phases can precipitate in the microstructure, like the γ'' and δ phases. This work presents an evaluation of the powerful optimum path forest (OPF) classifier configured with six distance functions to classify background echo and backscattered ultrasonic signals from samples of the inconel 625 superalloy thermally aged at 650 and 950 °C for 10, 100 and 200 h. The background echo and backscattered ultrasonic signals were acquired using transducers with frequencies of 4 and 5 MHz. The potentiality of ultrasonic sensor signals combined with the OPF to characterize the microstructures of an inconel 625 thermally aged and in the as-welded condition were confirmed by the results. The experimental results revealed that the OPF classifier is sufficiently fast (classification total time of 0.316 ms) and accurate (accuracy of 88.75%" and harmonic mean of 89.52) for the application proposed.