Simulated four-dimensional CT for markerless tumor tracking using a deep learning network with multi-task learning

IntroductionOur markerless tumor tracking algorithm requires 4DCT data to train models. 4DCT cannot be used for markerless tracking for respiratory-gated treatment due to inaccuracies and a high radiation dose. We developed a deep neural network (DNN) to generate 4DCT from 3DCT data.MethodsWe used 2420 thoracic 4DCT datasets from 436 patients to train a DNN, designed to export 9 deformation vector fields (each field representing one-ninth of the respiratory cycle) from each CT dataset based on a 3D convolutional autoencoder with shortcut connections using deformable image registration. Then 3DCT data at exhale were transformed using the predicted deformation vector fields to obtain simulated 4DCT data. We compared markerless tracking accuracy between original and simulated 4DCT datasets for 20 patients. Our tracking algorithm used a machine learning approach with patient-specific model parameters. For the training stage, a pair of digitally reconstructed radiography images was generated using 4DCT for each patient. For the prediction stage, the tracking algorithm calculated tumor position using incoming fluoroscopic image data.ResultsDiaphragmatic displacement averaged over 40 cases for the original 4DCT were slightly higher (<1.3 mm) than those for the simulated 4DCT. Tracking positional errors (95th percentile of the absolute value of displacement, “simulated 4DCT” minus “original 4DCT”) averaged over the 20 cases were 0.56 mm, 0.65 mm, and 0.96 mm in the X, Y and Z directions, respectively.ConclusionsWe developed a DNN to generate simulated 4DCT data that are useful for markerless tumor tracking when original 4DCT is not available. Using this DNN would accelerate markerless tumor tracking and increase treatment accuracy in thoracoabdominal treatment.     

1D-CADCapsNet: One dimensional deep capsule networks for coronary artery disease detection using ECG signals

PurposeCardiovascular disease (CVD) is a leading cause of death globally. Electrocardiogram (ECG), which records the electrical activity of the heart, has been used for the diagnosis of CVD. The automated and robust detection of CVD from ECG signals plays a significant role for early and accurate clinical diagnosis. The purpose of this study is to provide automated detection of coronary artery disease (CAD) from ECG signals using capsule networks (CapsNet).MethodsDeep learning-based approaches have become increasingly popular in computer aided diagnosis systems. Capsule networks are one of the new promising approaches in the field of deep learning. In this study, we used 1D version of CapsNet for the automated detection of coronary artery disease (CAD) on two second (95,300) and five second-long (38,120) ECG segments. These segments are obtained from 40 normal and 7 CAD subjects. In the experimental studies, 5-fold cross validation technique is employed to evaluate performance of the model.ResultsThe proposed model, which is named as 1D-CADCapsNet, yielded a promising 5-fold diagnosis accuracy of 99.44% and 98.62% for two- and five-second ECG signal groups, respectively. We have obtained the highest performance results using 2 s ECG segment than the state-of-art studies reported in the literature.Conclusions1D-CADCapsNet model automatically learns the pertinent representations from raw ECG data without using any hand-crafted technique and can be used as a fast and accurate diagnostic tool to help cardiologists.     

Simple low-cost approaches to semantic segmentation in radiation therapy planning for prostate cancer using deep learning with non-contrast planning CT images

PurposeDeep learning has shown great efficacy for semantic segmentation. However, there are difficulties in the collection, labeling and management of medical imaging data, because of ethical complications and the limited number of imaging studies available at a single facility.This study aimed to find a simple and low-cost method to increase the accuracy of deep learning semantic segmentation for radiation therapy of prostate cancer.MethodsIn total, 556 cases with non-contrast CT images for prostate cancer radiation therapy were examined using a two-dimensional U-Net. Initially, all slices were used for the input data. Then, we removed slices of the cranial portions, which were beyond the margins of the bladder and rectum. Finally, the ground truth labels for the bladder and rectum were added as channels to the input for the prostate training dataset.ResultsThe highest mean dice similarity coefficients (DSCs) for each organ in the test dataset of 56 cases were 0.85 ± 0.05, 0.94 ± 0.04 and 0.85 ± 0.07 for the prostate, bladder and rectum, respectively. Removal of the cranial slices from the original images significantly increased the DSC of the rectum from 0.83 ± 0.09 to 0.85 ± 0.07 (p < 0.05). Adding bladder and rectum information to prostate training without removing the slices significantly increased the DSC of the prostate from 0.79 ± 0.05 to 0.85 ± 0.05 (p < 0.05).ConclusionsThese cost-free approaches may be useful for new applications, which may include updated models and datasets. They may be applicable to other organs at risk (OARs) and clinical targets such as elective nodal irradiation.     

Modeling of dynamical systems through deep learning

This review presents a modern perspective on dynamical systems in the context of current goals and open challenges. In particular, our review focuses on the key challenges of discovering dynamics from data and finding data-driven representations that make nonlinear systems amenable to linear analysis. We explore various challenges in modern dynamical systems, along with emerging techniques in data science and machine learning to tackle them. The two chief challenges are (1) nonlinear dynamics and (2) unknown or partially known dynamics. Machine learning is providing new and powerful techniques for both challenges. Dimensionality reduction methods are used for projecting dynamical methods in reduced form, and these methods perform computational efficiency on real-world data. Data-driven models drive to discover the governing equations and give laws of physics. The identification of dynamical systems through deep learning techniques succeeds in inferring physical systems. Machine learning provides advanced new and powerful algorithms for nonlinear dynamics. Advanced deep learning methods like autoencoders, recurrent neural networks, convolutional neural networks, and reinforcement learning are used in modeling of dynamical systems.     

Historical demography and climate driven distributional changes in a widespread Neotropical freshwater species with high economic importance

The Neotropical region exhibits the greatest worldwide diversity and the diversification history of several clades is related to the puzzling geomorphologic and climatic history of this region. The freshwater Amazon ecoregion contains the main hydrographic basins of the Neotropical region that are highly dendritic and ecologically diverse. It contains a rich and endemic fish fauna, including one of its most iconic and economically important representatives, the bony-tongue Arapaima gigas (Teleostei, Osteoglossiformes). Here, we evaluated the projected distribution of the genus in different historical periods (Present, Last Glacial Maximum, Last Interglacial Maximum and Near Future) and interpreted these results in light of the genomic diversity and modeled historical demography. For that, we combined species distribution models, population genetic analysis using SNPs and deep learning model selection. We analyzed a representative sample of the genus from the two basins where it naturally occurs, four localities in the Amazon (Am) and three in the Tocantins-Araguaia (To-Ar) basin, as well as individuals from three fish farms. We inferred a potentially smaller distribution in the glacial period, with a possible refuge in central Am. Our genetic data agrees with this result, suggesting a higher level of genetic diversity in the Am basin, compared to that observed in To-Ar. Our deep learning model comparison indicated that the To-Ar basin was colonized by the population from the Am basin. Considering a global warming scenario in the near future, A. gigas could reach an even larger range, especially if anthropogenic related dispersal occurs, potentially invading new areas and impacting their communities.     

Three critical factors affecting automated image species recognition performance for camera traps

Ecological camera traps are increasingly used by wildlife biologists to unobtrusively monitor an ecosystems animal population. However, manual inspection of the images produced is expensive, laborious, and time‐consuming. The success of deep learning systems using camera trap images has been previously explored in preliminary stages. These studies, however, are lacking in their practicality. They are primarily focused on extremely large datasets, often millions of images, and there is little to no focus on performance when tasked with species identification in new locations not seen during training. Our goal was to test the capabilities of deep learning systems trained on camera trap images using modestly sized training data, compare performance when considering unseen background locations, and quantify the gradient of lower bound performance to provide a guideline of data requirements in correspondence to performance expectations. We use a dataset provided by Parks Canada containing 47,279 images collected from 36 unique geographic locations across multiple environments. Images represent 55 animal species and human activity with high‐class imbalance. We trained, tested, and compared the capabilities of six deep learning computer vision networks using transfer learning and image augmentation: DenseNet201, Inception‐ResNet‐V3, InceptionV3, NASNetMobile, MobileNetV2, and Xception. We compare overall performance on “trained” locations where DenseNet201 performed best with 95.6% top‐1 accuracy showing promise for deep learning methods for smaller scale research efforts. Using trained locations, classifications with <500 images had low and highly variable recall of 0.750 ± 0.329, while classifications with over 1,000 images had a high and stable recall of 0.971 ± 0.0137. Models tasked with classifying species from untrained locations were less accurate, with DenseNet201 performing best with 68.7% top‐1 accuracy. Finally, we provide an open repository where ecologists can insert their image data to train and test custom species detection models for their desired ecological domain.     

Flexible Analytic Wavelet Transform Based Features for Physical Action Identification Using sEMG Signals

ObjectivesElectromyography (EMG) is recording of the electrical activity produced by skeletal muscles. The classification of the EMG signals for different physical actions can be useful in restoring some or all of the lost motor functionalities in these individuals. Accuracy in classifying the EMG signal indicates efficient control of prosthesis.Material and methodsThe flexible analytic wavelet transform (FAWT) is used for classification of surface electromyography (sEMG) signals for identification of physical actions. FAWT is an efficient method for decomposition of sEMG signal into eight sub-bands, features namely neg-entropy, mean absolute value (MAV), variance (VAR), modified mean absolute value type 1 (MAV1), waveform length (WL), simple square integral (SSI), Tsallis entropy, integrated EMG (IEMG) are extracted from the sub-bands. Extracted features are fed into an extreme learning machine (ELM) classifier with sigmoid activation function.ResultsComprehensive experiments are conducted on the input sEMG signals and the accuracy, sensitivity and specificity scores are used for performance measurement. Experiments showed that among all sub-bands, the seventh sub-band provided the best performance where the recorded accuracy, sensitivity and specificity values were 99.36%, 99.36% and 99.93%, respectively. The comparison results showed best efficiency of proposed method as compared to other methods on the same dataset.ConclusionThis paper investigates the usage of the FAWT and ELM on sEMG signal classification. The results show that the proposed method is quite efficient in classification of the sEMG signals. It is also observed that the seventh sub-band of the FAWT provides the best discrimination property. In the future works, recent wavelet transform methods will be used for improving the classification performance.