No description available.

No description available.

No description available.

This contribution will report on further challenges if STEMSI is used for the acquisition of water transport in a heterogeneous root phantom and around the roots of a life plant system. While a 3D MRI image of the root system with sufficient spatial resolution is necessary it is also important to obtain the full 3D information of the velocity vector for the water movement in the vicinity of the plant roots. These requirements need to be balanced against the necessary acquisition time for this 6D data set since the plant is growing and therefore changing its root system over time. To meet this requirement the concepts for Stimulated Echo Acquisition Mode (STEAM) [4] have been fully incorporated into the STEMSI method, thus enabling rapid multi-slice acquisition while retaining sufficient signal to noise ratios.

No description available.

Biological systems constantly evolve and adapt in response to changed environment and external stimuli at the molecular and genomic levels. Building statistical models that characterize such dynamic changes in biological systems is one of the key objectives in bioinformatics and computational biology. Recent advances in high-throughput genomic and molecular profiling technologies such as gene expression and and copy number microarrays provide ample opportunities to study cellular activities at the individual gene and network levels. The aim of this dissertation is to formulate mathematically dynamic changes in biological networks and DNA copy numbers, to develop machine learning algorithms to learn these statistical models from high-throughput biological data, and to demonstrate their applications in systems biological studies. The first part (Chapters 2-4) of the dissertation focuses on the dynamic changes taking placing at the biological network level. Biological networks are context-specific and dynamic in nature. Under different conditions, different regulatory components and mechanisms are activated and the topology of the underlying gene regulatory network changes. We report a differential dependency network (DDN) analysis to detect statistically significant topological changes in the transcriptional networks between two biological conditions. Further, we formalize and extend the DDN approach to an effective learning strategy to extract structural changes in graphical models using l1-regularization based convex optimization. We discuss the key properties of this formulation and introduce an efficient implementation by the block coordinate descent algorithm. Another type of dynamic changes in biological networks is the observation that a group of genes involved in certain biological functions or processes coordinate to response to outside stimuli, producing distinct time course patterns. We apply the echo stat network, a new architecture of recurrent neural networks, to model temporal gene expression patterns and analyze the theoretical properties of echo state networks with random matrix theory. The second part (Chapter 5) of the dissertation focuses on the changes at the DNA copy number level, especially in cancer cells. Somatic DNA copy number alterations (CNAs) are key genetic events in the development and progression of human cancers, and frequently contribute to tumorigenesis. We propose a statistically-principled in silico approach, Bayesian Analysis of COpy number Mixtures (BACOM), to accurately detect genomic deletion type, estimate normal tissue contamination, and accordingly recover the true copy number profile in cancer cells. / Ph. D.

Dynamic spectrum access (DSA) is a promising technology to mitigate spectrum shortage and improve spectrum utilization. However, DSA users have to face two fundamental issues, interference coordination between DSA users and protections to primary users (PUs). These two issues are very challenging, since generally there is no powerful infrastructure in DSA networks to support centralized control. As a result, DSA users have to perform spectrum managements, including spectrum access and power allocations, independently without accurate channel state information. In this thesis, a novel spectrum management approach is proposed, in which Q-learning, a type of reinforcement learning, is utilized to enable DSA users to carry out effective spectrum managements individually and intelligently. For more efficient processes, powerful neural networks (NNs) are employed to implement Q-learning processes, so-called deep Q-network (DQN). Furthermore, I also investigate the optimal way to construct DQN considering both the performance of wireless communications and the difficulty of NN training. Finally, extensive simulation studies are conducted to demonstrate the effectiveness of the proposed spectrum management approach. / Generally, in dynamic spectrum access (DSA) networks, co-operations and centralized control are unavailable and DSA users have to carry out wireless transmissions individually. DSA users have to know other users’ behaviors by sensing and analyzing wireless environments, so that DSA users can adjust their parameters properly and carry out effective wireless transmissions. In this thesis, machine learning and deep learning technologies are leveraged in DSA network to enable appropriate and intelligent spectrum managements, including both spectrum access and power allocations. Accordingly, a novel spectrum management framework utilizing deep reinforcement learning is proposed, in which deep reinforcement learning is employed to accurately learn wireless environments and generate optimal spectrum management strategies to adapt to the variations of wireless environments. Due to the model-free nature of reinforcement learning, DSA users only need to directly interact with environments to obtain optimal strategies rather than relying on accurate channel estimations. In this thesis, Q-learning, a type of reinforcement learning, is adopted to design the spectrum management framework. For more efficient and accurate learning, powerful neural networks (NN) is employed to combine Q-learning and deep learning, also referred to as deep Q-network (DQN). The selection of NNs is crucial for the performance of DQN, since different types of NNs possess various properties and are applicable for different application scenarios. Therefore, in this thesis, the optimal way to construct DQN is also analyzed and studied. Finally, the extensive simulation studies demonstrate that the proposed spectrum management framework could enable users to perform proper spectrum managements and achieve better performance.

Echo in telephones is generally undesirable but inevitable. There are two possible sources of echo in a telephone system. The impedance mismatch in hybrids generates network (electric) echo. The acoustic coupling between loudspeaker and microphone, in hands-free telephones, produces acoustic echo. Echo cancelers are used to control these echoes. In this thesis, we analyze the Least Mean Squares (LMS), Normalized LMS (NLMS), Recursive Least Squares (RLS), and Subband NLMS (SNLMS) algorithms, and evaluate their performance as acoustic and network echo cancelers. The algorithms are compared based on their convergence rate, steady state echo return loss (ERL), and complexity of implementation. While LMS is simple, its convergence rate is dependent on the eigenvalue spread of the signal. In particular, it converges slowly with speech as input. This problem is mitigated in NLMS. The complexity of NLMS is comparable to that of LMS. The convergence rate of RLS is independent of the eigenvalue spread, and it has the fastest convergence. On the other hand, RLS is highly computation intensive. Among the four algorithms considered here, SNLMS has the least complexity of implementation, as well as the slowest rate of convergence. Switching between the NLMS and SNLMS algorithms is used to achieve fast convergence with low computational requirements. For a given computational power, it is shown that switching between algorithms can give better performance than using either of the two algorithms exclusively, especially in rooms with long reverberation times. We also discuss various implementation issues associated with an integrated echo cancellation system, such as double-talk detection, finite precision effects, nonlinear processing, and howling detection and control. The use of a second adaptive filter is proposed, to reduce near-end ambient noise. Simulation results indicate that this approach can reduce the ambient noise by about 20 dB. A configuration is presented for the real time single-chip DSP implementation of acoustic and network echo cancelers, and an interface between the echo canceler and the telephone is proposed. Finally, some results obtained from simulations and implementations of individual modules, on the TMS320C31 and ADSP 2181 processors, are reported. The real time NLMS DSP implementations provide 15 dB of echo return loss. / Master of Science

Yes / Objectives: To assess the test-retest variability of both diffusion parameters and fat fraction (FF) estimates in normal muscle, and to assess differences in normal values between muscles in the thigh. Methods: 29 healthy volunteers (mean age 37 years, range 20-60 years, 17/29 males) completed the study. Magnetic resonance images of the mid-thigh were acquired using a stimulated echo acquisition mode-echoplanar imaging (STEAM-EPI) imaging sequence, to assess diffusion, and 2-point Dixon imaging, to assess FF. Imaging was repeated in 19 participants after a 30 min interval in order to assess test-retest variability of the measurements. Results: Intraclass correlation coefficients (ICCs) for test-retest variability were 0.99 [95% confidence interval, (CI): 0.98, 1] for FF, 0.94 (95% CI: 0.84, 0.97) for mean diffusivity and 0.89 (95% CI: 0.74, 0.96) for fractional anisotropy (FA). FF was higher in the hamstrings than the quadriceps by a mean difference of 1.81% (95% CI:1.63, 2.00)%, p < 0.001. Mean diffusivity was significantly lower in the hamstrings than the quadriceps (0.26 (0.13, 0.39) x10-3 mm2s-1, p < 0.001) whereas fractional anisotropy was significantly higher in the hamstrings relative to the quadriceps with a mean difference of 0.063 (0.05, 0.07), p < 0.001. Conclusions: This study has shown excellent test-retest, variability in MR-based FF and diffusion measurements and demonstrated significant differences in these measures between hamstrings and quadriceps in the healthy thigh. Advances in knowledge: Test-retest variability is excellent for STEAM-EPI diffusion and 2-point Dixon-based FF measurements in the healthy muscle. Inter- and intraobserver variability were excellent for region of interest placement for STEAM-EPI diffusion and 2-point Dixon-based FF measurements in the healthy muscle. There are significant differences in FF and diffusion measurements between the hamstrings and quadriceps in the normal muscle. / ICA-CL-2016-02-017/DH_/Department of Health/United Kingdom; NIHR

No description available.