Generative adversarial network: a statistical-based deep learning paradigm to improve detecting breast cancer in thermograms.

Thermography, as a harmless modality, thanks to its low equipment complexity in parallel with quick and cheap access, has been able to come up as a method with significant potential in the diagnosis of some cancers in recent years. However, the complexity of the images resulting from this method has caused the use of deep learning to interpret thermograms. A limiting factor in this process is the strong dependence of deep learning methods on the number of training data, which is a serious challenge in thermography due to the young age of this technology and the lack of available images. In this paper, an attempt is made to reduce the above challenge by utilizing the concept of statistical learning in such a way that the statistical distribution of the original data is estimated by using generative adversarial networks (i.e., GAN). Then, several fake images are reconstructed based on the estimated distribution in order to increase the training thermograms. Since the fake images are reconstructed based on similar statistics of real thermograms in each class, the effective features of each class are preserved to a significant extent in the reconstruction process. The use of this method indicates a significant improvement in the separation of healthy and cancerous thermograms compared to the benchmark method which does not use the concept of GAN in such a way that characteristics of sensitivity and accuracy are improved in ranges of 3-9% and 3-7%, respectively. In terms of specificity, although we have seen an improvement of up to 9%, in some cases, small drops of up to 2% have also been observed, which can still be justified due to the significant improvement in sensitivity and accuracy.

A New method for promote the performance of deep learning paradigm in diagnosing breast cancer: improving role of fusing multiple views of thermography images.

Purpose: Breast cancer is one of the deadliest cancers among women worldwide which its early detection may significantly reduce its mortality rate. Thermgraphy is a new, non-invasive, non-painful, and low-cost modality that detects abnormalities by detecting heat from the breast surface. Method: Recent research has applied deep learning to early breast cancer diagnosis via thermography, using only the frontal view of thermograms. We combine several views of thermal images to improve the performance of pre-trained deep learning architectures in this article. This goal is achieved by combining frontal-45 data with lateral-45 and lateral45 thermograms to construct a detection model that utilizes transfer learning. Result: Research in this area uses the Database for Mastology Research (DMR) with infrared images. In this study, transfer based deep learning methods are demonstrated to be effective in fusing several views of thermograms to diagnose breast cancer in a manner that can result in a sensitivity increase of 2-15 percent and a specificity increase of 2-30 percent compared to other deep learning-based or handcrafted schemes. Conclusion: Using multiple views of thermograms and transfer learning, this paper proposes a method for improving breast cancer diagnosis. Using methods based on deep learning and methods based on hand-crafted features, we evaluated the performance of the proposed model. Using the obtained results as a basis for future research, the proposed design can be improved and developed as a valid approach in interpreting breast thermography images.

A Novel Pulse-Taking Device for Persian Medicine Based on Convolutional Neural Networks.

Background: In Persian medicine (PM), measuring the wrist pulse is one of the main methods for determining a person's health status and temperament. One problem that can arise is the dependence of the diagnosis on the physician's interpretation of pulse wave features. Perhaps, this is one reason why this method has yet to be combined with modern medical methods. This paper addresses this concern and outlines a system for measuring pulse signals based on PM. Methods: A system that uses data from a customized device that logs the pulse wave on the wrist was designed and clinically implemented based on PM. Seven convolutional neural networks (CNNs) have been used for classification. Results: The pulse wave features of 34 participants were assessed by a specialist based on PM principles. Pulse taking was done on the wrist in the supine position (named Malmas in PM) under the supervision of the physician. Seven CNNs were implemented for each participant's pulse characteristic (pace, rate, vessel elasticity, strength, width, length, and height) assessment, and then, each participant was classified into three classes. Conclusion: It appears that the design and construction of a customized device combined with the deep learning algorithm can measure the pulse wave features according to PM and it can increase the reliability and repeatability of the diagnostic results based on PM.

Residual Learning: A New Paradigm to Improve Deep Learning-Based Segmentation of the Left Ventricle in Magnetic Resonance Imaging Cardiac Images.

BACKGROUND: Recently, magnetic resonance imaging (MRI) has become a useful tool for the early detection of heart failure. A vital step of this process is a valid measurement of the left ventricle's properties, which seriously depends on the accurate segmentation of the heart in captured images. Although various schemes have been tested for this segmentation so far, the latest proposed methods have used the concept of deep learning to estimate the range of the left ventricle in cardiac MRI images. While deep learning methods can lead to better results than their classical alternatives, but unfortunately, the gradient vanishing and exploding problems may hamper their efficiency for the accurate segmentation of the left ventricle in MRI heart images. METHODS: In this article, a new concept called residual learning is utilized to improve the performance of deep learning schemes against gradient vanishing problems. For this purpose, the Residual Network of Residual Network (i.e., Residual of Residual) substructure is utilized inside the main deep learning architecture (e.g., Unet), which provides more significant detection indexes. RESULTS AND CONCLUSION: The proposed method's performances and its alternatives were evaluated on Sunnybrook Cardiac Data as a reliable dataset in the left ventricle segmentation. The results show that the detection parameters are improved at least by 5%, 3.5%, 8.1%, and 11.4% compared to its deep alternatives in terms of Jaccard, Dice, precision, and false-positive rate indexes, respectively. These improvements were made when the recall parameter was reduced to a negligible value (i.e., approximately 1%). Overall, the proposed method can be used as a suitable tool for more accurate detection of the left ventricle in MRI images.

A New Method to Improve the Performance of Deep Neural Networks in Detecting P300 Signals: Optimizing Curvature of Error Surface Using Genetic Algorithm.

BACKGROUND: Deep neural networks have been widely used in detection of P300 signal in Brain Machine Interface (BMI) systems which are rely on Event-Related Potentials (ERPs) (i.e. P300 signals). Such networks have high curvature variation in their error surface hampering their favorable performance. Therefore, the variations in curvature of the error surface must be minimized to improve the performance of these networks in detecting P300 signals. OBJECTIVE: The aim of this paper is to introduce a method for minimizing the curvature of the error surface during training Convolutional Neural Network (CNN). The curvature variation of the error surface is highly dependent on model parameters of deep neural network; therefore, we try to minimize this curvature by optimizing the model parameters. MATERIAL AND METHODS: In this experimental study an attempt is made to tune the CNN parameters affecting the curvature of its error surface in order to obtain the best possible learning. For achieving this goal, Genetic Algorithm is utilized to optimize the above parameters in order to minimize the curvature variations. RESULTS: The performance of the proposed algorithm was evaluated on EPFL dataset. The obtained results demonstrated that the proposed method detected the P300 signals with maximally 98.91% classification accuracy and 98.54% True Positive Ratio (TPR). CONCLUSION: The obtained results showed that using genetic algorithm for minimizing curvature of the error surface in CNN increased its accuracy in parallel with decreasing the variance of the results. Consequently, it may be concluded that the proposed method has considerable potential to be used as P300 detection module in BMI applications.

Mobile sensor based human activity recognition: distinguishing of challenging activities by applying long short-term memory deep learning modified by residual network concept.

Automated recognition of daily human tasks is a novel method for continuous monitoring of the health of elderly people. Nowadays mobile devices (i.e. smartphone and smartwatch) are equipped with a variety of sensors, therefore activity classification algorithms have become as useful, low-cost, and non-invasive diagnostic modality to implement as mobile software. The aim of this article is to introduce a new deep learning structure for recognizing challenging (i.e. similar) human activities based on signals which have been recorded by sensors mounted on mobile devices. In the proposed structure, the residual network concept is engaged as a new substructure inside the main proposed structure. This part is responsible to address the problem of accuracy saturation in convolutional neural networks, thanks to its ability in jump over some layers which leads to reducing vanishing gradients effect. Therefore the accuracy of the classification of several activities is increased by using the proposed structure. Performance of the proposed method is evaluated on real life recorded signals and is compared with existing techniques in two different scenarios. The proposed structure is applied on two well-known human activity datasets that have been prepared in university of Fordham. The first dataset contains the recorded signals which arise from six different activities including walking, jogging, upstairs, downstairs, sitting, and standing. The second dataset also contains walking, jogging, stairs, sitting, standing, eating soup, eating sandwich, and eating chips. In the first scenario, the performance of the proposed structures is compared with deep learning schemes. The obtained results show that the proposed method may improve the recognition rate at least 5% for the first dataset against its own family alternatives in distinguishing challenging activities (i.e. downstairs and upstairs). For the second data set similar improvements is obtained for some challenging activities (i.e. eating sandwich and eating chips). These superiorities even reach to at least 28% when the capability of the proposed method in recognizing downstairs and upstairs is compared to its non-family methods for the first dataset. Increasing the recognition rate of the proposed method for challenging activities (i.e. downstairs and upstairs, eating sandwich and eating chips) in parallel with its acceptable performance for other non-challenging activities shows its effectiveness in mobile sensor-based health monitoring systems.

A New Method for Detecting P300 Signals by Using Deep Learning: Hyperparameter Tuning in High-Dimensional Space by Minimizing Nonconvex Error Function.

BACKGROUND: P300 signal detection is an essential problem in many fields of Brain-Computer Interface (BCI) systems. Although deep neural networks have almost ubiquitously used in P300 detection, in such networks, increasing the number of dimensions leads to growth ratio of saddle points to local minimums. This phenomenon results in slow convergence in deep neural network. Hyperparameter tuning is one of the approaches in deep learning, which leads to fast convergence because of its ability to find better local minimums. In this paper, a new adaptive hyperparameter tuning method is proposed to improve training of Convolutional Neural Networks (CNNs). METHODS: The aim of this paper is to introduce a novel method to improve the performance of deep neural networks in P300 signal detection. To reach this purpose, the proposed method transferred the non-convex error function of CNN) into Lagranging paradigm, then, Newton and dual active set techniques are utilized for hyperparameter tuning in order to minimize error of objective function in high dimensional space of CNN. RESULTS: The proposed method was implemented on MATLAB 2017 package and its performance was evaluated on dataset of Ecole Polytechnique Fédérale de Lausanne (EPFL) BCI group. The obtained results depicted that the proposed method detected the P300 signals with 95.34% classification accuracy in parallel with high True Positive Rate (i.e., 92.9%) and low False Positive Rate (i.e., 0.77%). CONCLUSIONS: To estimate the performance of the proposed method, the achieved results were compared with the results of Naive Hyperparameter (NHP) tuning method. The comparisons depicted the superiority of the proposed method against its alternative, in such way that the best accuracy by using the proposed method was 6.44%, better than the accuracy of the alternative method.

A new method for sperm characterization for infertility treatment: hypothesis testing by using combination of watershed segmentation and graph theory.

Shape and movement features of sperms are important parameters for infertility study and treatment. In this article, a new method is introduced for characterization sperms in microscopic videos. In this method, first a hypothesis framework is defined to distinguish sperms from other particles in captured video. Then decision about each hypothesis is done in following steps: Selecting some primary regions as candidates for sperms by watershed-based segmentation, pruning of some false candidates during successive frames using graph theory concept and finally confirming correct sperms by using their movement trajectories. Performance of the proposed method is evaluated on real captured images belongs to semen with high density of sperms. The obtained results show the proposed method may detect 97% of sperms in presence of 5% false detections and track 91% of moving sperms. Furthermore, it can be shown that better characterization of sperms in proposed algorithm doesn't lead to extracting more false sperms compared to some present approaches.