Input density tunes Kenyon cell sensory responses in the Drosophila mushroom body.

The ability to discriminate sensory stimuli with overlapping features is thought to arise in brain structures called expansion layers, where neurons carrying information about sensory features make combinatorial connections onto a much larger set of cells. For 50 years, expansion coding has been a prime topic of theoretical neuroscience, which seeks to explain how quantitative parameters of the expansion circuit influence sensory sensitivity, discrimination, and generalization. Here, we investigate the developmental events that produce the quantitative parameters of the arthropod expansion layer, called the mushroom body. Using Drosophila melanogaster as a model, we employ genetic and chemical tools to engineer changes to circuit development. These allow us to produce living animals with hypothesis-driven variations on natural expansion layer wiring parameters. We then test the functional and behavioral consequences. By altering the number of expansion layer neurons (Kenyon cells) and their dendritic complexity, we find that input density, but not cell number, tunes neuronal odor selectivity. Simple odor discrimination behavior is maintained when the Kenyon cell number is reduced and augmented by Kenyon cell number expansion. Animals with increased input density to each Kenyon cell show increased overlap in Kenyon cell odor responses and become worse at odor discrimination tasks.

Deep multi-modal intermediate fusion of clinical record and time series data in mortality prediction.

In intensive care units (ICUs), mortality prediction is performed by combining information from these two sources of ICU patients by monitoring patient health. Respectively, time series data generated from each patient admission to the ICU and clinical records consisting of physician diagnostic summaries. However, existing mortality prediction studies mainly cascade the multimodal features of time series data and clinical records for prediction, ignoring thecross-modal correlation between the underlying features in different modal data. To address theseissues, we propose a multimodal fusion model for mortality prediction that jointly models patients' time-series data as well as clinical records. We apply a fine-tuned Bert model (Bio-Bert) to the patient's clinical record to generate a holistic embedding of the text part, which is then combined with the output of an LSTM model encoding the patient's time-series data to extract valid features. The global contextual information of each modal data is extracted using the improved fusion module to capture the correlation between different modal data. Furthermore, the improved fusion module can be easily added to the fusion features of any unimodal network and utilize existing pre-trained unimodal model weights. We use a real dataset containing 18904 ICU patients to train and evaluate our model, and the research results show that the representations obtained by themodel can achieve better prediction accuracy compared to the baseline.

Fabrication of an ultra-high quality MgF2 micro-resonator for a single soliton comb generation.

Crystalline micro-resonators are attractive for a wide range of applications due to their extremely high quality (Q) factor. In this paper, we develop a semi-automatic method for fabricating ultra-high Q-factor MgF2 crystalline micro-resonators. By utilizing a force feedback sensor and corresponding control, we made a semi-automatic precision grind-and-polishing machine, and successfully fabricated trapezoid MgF2 resonators with diameter of 9.5 mm and a root mean square surface roughness of 0.26 nm. The maximum difference of peaks and valleys is about 1.5 nm. The Q-factor was characterized to be 9.24 × 109at 1550 nm by the cavity ring-down spectroscopy. A single soliton optical frequency comb was generated by pumping the microcavity with 150 mW optical power.

Hacking brain development to test models of sensory coding.

Animals can discriminate myriad sensory stimuli but can also generalize from learned experience. You can probably distinguish the favorite teas of your colleagues while still recognizing that all tea pales in comparison to coffee. Tradeoffs between detection, discrimination, and generalization are inherent at every layer of sensory processing. During development, specific quantitative parameters are wired into perceptual circuits and set the playing field on which plasticity mechanisms play out. A primary goal of systems neuroscience is to understand how material properties of a circuit define the logical operations-computations--that it makes, and what good these computations are for survival. A cardinal method in biology-and the mechanism of evolution--is to change a unit or variable within a system and ask how this affects organismal function. Here, we make use of our knowledge of developmental wiring mechanisms to modify hard-wired circuit parameters in the Drosophila melanogaster mushroom body and assess the functional and behavioral consequences. By altering the number of expansion layer neurons (Kenyon cells) and their dendritic complexity, we find that input number, but not cell number, tunes odor selectivity. Simple odor discrimination performance is maintained when Kenyon cell number is reduced and augmented by Kenyon cell expansion.

Inverse design of soliton microcomb based on genetic algorithm and deep learning.

Soliton microcombs generated by the third-order nonlinearity of microresonators exhibit high coherence, low noise, and stable spectra envelopes, which can be designed for many applications. However, conventional dispersion engineering based design methods require iteratively solving Maxwell's equations through time-consuming electromagnetic field simulations until a local optimum is obtained. Moreover, the overall inverse design from soliton microcomb to the microcavity geometry has not been systematically investigated. In this paper, we propose a high accuracy microcomb-to-geometry inverse design method based on the genetic algorithm (GA) and deep neural network (DNN), which effectively optimizes dispersive wave position and power. The method uses the Lugiato-Lefever equation and GA (LLE-GA) to obtain second- and higher-order dispersions from a target microcomb, and it utilizes a pre-trained forward DNN combined with GA (FDNN-GA) to obtain microcavity geometry. The results show that the dispersive wave position deviations of the inverse designed MgF2 and Si3N4 microresonators are less than 0.5%, and the power deviations are less than 5 dB, which demonstrates good versatility and effectiveness of our method for various materials and structures.

A multi-head self-attention deep learning approach for detection and recommendation of neuromagnetic high frequency oscillations in epilepsy.

Magnetoencephalography is a noninvasive neuromagnetic technology to record epileptic activities for the pre-operative localization of epileptogenic zones, which has received increasing attention in the diagnosis and surgery of epilepsy. As reported by recent studies, pathological high frequency oscillations (HFOs), when utilized as a biomarker to localize the epileptogenic zones, result in a significant reduction in seizure frequency, even seizure elimination in around 80% of cases. Thus, objective, rapid, and automatic detection and recommendation of HFOs are highly desirable for clinicians to alleviate the burden of reviewing a large amount of MEG data from a given patient. Despite the advantage, the performance of existing HFOs rarely satisfies the clinical requirement. Consequently, no HFOs have been successfully applied to real clinical applications so far. In this work, we propose a multi-head self-attention-based detector for recommendation, termed MSADR, to detect and recommend HFO signals. Taking advantage of the state-of-the-art multi-head self-attention mechanism in deep learning, the proposed MSADR achieves a more superior accuracy of 88.6% than peer machine learning models in both detection and recommendation tasks. In addition, the robustness of MSADR is also extensively assessed with various ablation tests, results of which further demonstrate the effectiveness and generalizability of the proposed approach.

Automatic BASED scoring on scalp EEG in children with infantile spasms using convolutional neural network.

In recent years, the Burden of Amplitudes and Epileptiform Discharges (BASED) score has been used as a reliable, accurate, and feasible electroencephalogram (EEG) grading scale for infantile spasms. However, manual EEG annotation is, in general, very time-consuming, and BASED scoring is no exception. Convolutional neural networks (CNNs) have proven their great potential in many EEG classification problems. However, very few research studies have focused on the use of CNNs for BASED scoring, a challenging but vital task in the diagnosis and treatment of infantile spasms. This study proposes an automatic BASED scoring framework using EEG and a deep CNN. The feasibility of using CNN for automatic BASED scoring was investigated in 36 patients with infantile spasms by annotating their long-term EEG data with four levels of the BASED score (scores 5, 4, 3, and ≤2). In the validation set, the accuracy was 96.9% by applying a multi-layer CNN to classify the EEG data as a 4-label problem. The extensive experiments have demonstrated that our proposed approach offers high accuracy and, hence, is an important step toward an automatic BASED scoring algorithm. To the best of our knowledge, this is the first attempt to use a CNN to construct a BASED-based scoring model.

Gating mechanism and a modulatory niche of human GluN1-GluN2A NMDA receptors.

N-methyl-D-aspartate (NMDA) receptors are glutamate-gated calcium-permeable ion channels that are widely implicated in synaptic transmission and plasticity. Here, we report a gallery of cryo-electron microscopy (cryo-EM) structures of the human GluN1-GluN2A NMDA receptor at an overall resolution of 4 Å in complex with distinct ligands or modulators. In the full-length context of GluN1-GluN2A receptors, we visualize the competitive antagonists bound to the ligand-binding domains (LBDs) of GluN1 and GluN2A subunits, respectively. We reveal that the binding of positive allosteric modulator shortens the distance between LBDs and the transmembrane domain (TMD), which further stretches the opening of the gate. In addition, we unexpectedly visualize the binding cavity of the "foot-in-the-door" blocker 9-aminoacridine within the LBD-TMD linker region, differing from the conventional "trapping" blocker binding site at the vestibule within the TMD. Our study provides molecular insights into the crosstalk between LBDs and TMD during channel activation, inhibition, and allosteric transition.

Detecting High Frequency Oscillations for Stereoelectroencephalography in Epilepsy via Hypergraph Learning.

Successful epilepsy surgeries depend highly on pre-operative localization of epileptogenic zones. Stereoelectroencephalography (SEEG) records interictal and ictal activities of the epilepsy in order to precisely find and localize epileptogenic zones in clinical practice. While it is difficult to find distinct ictal onset patterns generated the seizure onset zone from SEEG recordings in a confined region, high frequency oscillations are commonly considered as putative biomarkers for the identification of epileptogenic zones. Therefore, automatic and accurate detection of high frequency oscillations in SEEG signals is crucial for timely clinical evaluation. This work formulates the detection of high frequency oscillations as a signal segment classification problem and develops a hypergraph-based detector to automatically detect high frequency oscillations such that human experts can visually review SEEG signals. We evaluated our method on 4,000 signal segments from clinical SEEG recordings that contain both ictal and interictal data obtained from 19 patients who suffer from refractory focal epilepsy. The experimental results demonstrate the effectiveness of the proposed detector that can successfully localize interictal high frequency oscillations and outperforms multiple peer machine learning methods. In particular, the proposed detector achieved 90.7% in accuracy, 80.9% in sensitivity, and 96.9% in specificity.

Multi-Head Self-Attention Model for Classification of Temporal Lobe Epilepsy Subtypes.

As a long-standing chronic disease, Temporal Lobe Epilepsy (TLE), resulting from abnormal discharges of neurons and characterized by recurrent episodic central nervous system dysfunctions, has affected more than 70% of drug-resistant epilepsy patients across the world. As the etiology and clinical symptoms are complicated, differential diagnosis of TLE mainly relies on experienced clinicians, and specific diagnostic biomarkers remain unclear. Though great effort has been made regarding the genetics, pathology, and neuroimaging of TLE, an accurate and effective diagnosis of TLE, especially the TLE subtypes, remains an open problem. It is of a great importance to explore the brain network of TLE, since it can provide the basis for diagnoses and treatments of TLE. To this end, in this paper, we proposed a multi-head self-attention model (MSAM). By integrating the self-attention mechanism and multilayer perceptron method, the MSAM offers a promising tool to enhance the classification of TLE subtypes. In comparison with other approaches, including convolutional neural network (CNN), support vector machine (SVM), and random forest (RF), experimental results on our collected MEG dataset show that the MSAM achieves a supreme performance of 83.6% on accuracy, 90.9% on recall, 90.7% on precision, and 83.4% on F1-score, which outperforms its counterparts. Furthermore, effectiveness of varying head numbers of multi-head self-attention is assessed, which helps select the optimal number of multi-head. The self-attention aspect learns the weights of different signal locations which can effectively improve classification accuracy. In addition, the robustness of MSAM is extensively assessed with various ablation tests, which demonstrates the effectiveness and generalizability of the proposed approach.

Automatic and Accurate Epilepsy Ripple and Fast Ripple Detection via Virtual Sample Generation and Attention Neural Networks.

About 1% of the population around the world suffers from epilepsy. The success of epilepsy surgery depends critically on pre-operative localization of epileptogenic zones. High frequency oscillations including ripples (80-250 Hz) and fast ripples (250-500 Hz) are commonly used as biomarkers to localize epileptogenic zones. Recent literature demonstrated that fast ripples indicate epileptogenic zones better than ripples. Thus, it is crucial to accurately detect fast ripples from ripples signals of magnetoencephalography for improving outcome of epilepsy surgery. This paper proposes an automatic and accurate ripple and fast ripple detection method that employs virtual sample generation and neural networks with an attention mechanism. We evaluate our proposed detector on patient data with 50 ripples and 50 fast ripples labeled by two experts. The experimental results show that our new detector outperforms multiple traditional machine learning models. In particular, our method can achieve a mean accuracy of 89.3% and an average area under the receiver operating characteristic curve of 0.88 in 50 repeats of random subsampling validation. In addition, we experimentally demonstrate the effectiveness of virtual sample generation, attention mechanism, and architecture of neural network models.

Photonic thermometer with a sub-millikelvin resolution and broad temperature range by waveguide-microring Fano resonance.

Fano resonance theoretically is an effective approach for sensitivity enhancement in photonic sensing applications, but the reported methods suffer from complicated structure and fabrication, narrow dynamic range, etc. In this article, we propose a photonic thermometer with sub-millikelvin resolution and broad temperature measurement range implemented by a simple waveguide-microring Fano structure. An air hole is introduced at the center of the coupling region of the waveguide of an all-pass microring resonator. The effective refractive index theory is used to design its equivalent phase shift and therefore the lineshape of the Fano resonance. Experimental results showed that the quality factor and the Fano parameter of the structure were invariant in a broad temperature range. The wavelength-temperature sensitivity was 75.3 pm/℃, the intensity-temperature sensitivity at the Fano asymmetric edge was 7.49 dB/℃, and the temperature resolution was 0.25 mK within 10℃ to 90℃.

Deep learning in bioinformatics: Introduction, application, and perspective in the big data era.

Deep learning, which is especially formidable in handling big data, has achieved great success in various fields, including bioinformatics. With the advances of the big data era in biology, it is foreseeable that deep learning will become increasingly important in the field and will be incorporated in vast majorities of analysis pipelines. In this review, we provide both the exoteric introduction of deep learning, and concrete examples and implementations of its representative applications in bioinformatics. We start from the recent achievements of deep learning in the bioinformatics field, pointing out the problems which are suitable to use deep learning. After that, we introduce deep learning in an easy-to-understand fashion, from shallow neural networks to legendary convolutional neural networks, legendary recurrent neural networks, graph neural networks, generative adversarial networks, variational autoencoder, and the most recent state-of-the-art architectures. After that, we provide eight examples, covering five bioinformatics research directions and all the four kinds of data type, with the implementation written in Tensorflow and Keras. Finally, we discuss the common issues, such as overfitting and interpretability, that users will encounter when adopting deep learning methods and provide corresponding suggestions. The implementations are freely available at https://github.com/lykaust15/Deep_learning_examples.

[De novo construction of mammalian synthetic inhibitory transcription factor and promoter pairs].

Transcriptional regulation is crucial for regulated gene expression. Due to the complexity, it has been difficult to engineer eukaryotic transcription factor (TF) and promoter pairs. The few availabilities of eukaryotic TF and promotor pairs limit their practical use for clinical or industrial applications. Here, we report a de novo construction of synthetic inhibitory transcription factor and promoter pairs for mammalian transcriptional regulation. The design of synthetic TF was based on the fusion of DNA binding domain and Kruppel associated box transcription regulating domain (KRAB). The synthetic promoter was constructed by inserting the corresponding TF response element after SV40 promoter. We constructed and tested five synthetic inhibitory transcription factor and promoter pairs in cultured mammalian cells. The inhibition capability and orthogonality were verified by flow cytometry. In summary, we demonstrate the feasibility of constructing mammalian inhibitory TF and promoter pairs, which could be standardized for advanced gene-circuit design and various applications in the mammalian synthetic biology.

Edge extraction using a time-varying vortex beam in incoherent digital holography.

Edge extraction using a time-varying vortex beam (TV-VB) is demonstrated in optical scanning holography (OSH) operating in an incoherent mode. OSH is a two-pupil heterodyne scanning optical system. We propose that one of the pupil functions used is a delta function and the other pupil function is a spiral phase plate (SPP). The interference of these pupils creates a TV-VB to scan over an object to record the edge-only information of an object holographically. We also find that a reconstructed edge with better contrast is achieved by translating the SPP away from the pupil plane. Experimental results are compared with computer simulations and found to be in good agreement.

Improved full analytical polygon-based method using Fourier analysis of the three-dimensional affine transformation.

Previous research [Appl. Opt.52, A290 (2013)] has revealed that Fourier analysis of three-dimensional affine transformation theory can be used to improve the computation speed of the traditional polygon-based method. In this paper, we continue our research and propose an improved full analytical polygon-based method developed upon this theory. Vertex vectors of primitive and arbitrary triangles and the pseudo-inverse matrix were used to obtain an affine transformation matrix representing the spatial relationship between the two triangles. With this relationship and the primitive spectrum, we analytically obtained the spectrum of the arbitrary triangle. This algorithm discards low-level angular dependent computations. In order to add diffusive reflection to each arbitrary surface, we also propose a whole matrix computation approach that takes advantage of the affine transformation matrix and uses matrix multiplication to calculate shifting parameters of similar sub-polygons. The proposed method improves hologram computation speed for the conventional full analytical approach. Optical experimental results are demonstrated which prove that the proposed method can effectively reconstruct three-dimensional scenes.

3D dynamic holographic display by modulating complex amplitude experimentally.

Complex amplitude modulation method is presented theoretically and performed experimentally for three-dimensional (3D) dynamic holographic display with reduced speckle using a single phase-only spatial light modulator. The determination of essential factors is discussed based on the basic principle and theory. The numerical simulations and optical experiments are performed, where the static and animated objects without refinement on the surfaces and without random initial phases are reconstructed successfully. The results indicate that this method can reduce the speckle in reconstructed images effectively; furthermore, it will not cause the internal structure in the reconstructed pixels. Since the complex amplitude modulation is based on the principle of phase-only hologram, it does not need the stringent alignment of pixels. This method can be used for high resolution imaging or measurement in various optical areas.

Analytical brightness compensation algorithm for traditional polygon-based method in computer-generated holography.

In three-dimensional (3D) holographic display, current brightness compensation algorithm of the traditional polygon-based method experimentally obtains the compensation factor, which depends on the fabrication process. In this paper, we proposed an analytical brightness compensation method discarding the influence of the fabrication. The surface property function with the flat power spectral density and the compensation factor obtained from the simplified relationship between the original and the rotated frequencies are used to analytically compensate the radiant energy of the tilted polygon. The optical reconstruction results show the proposed method could effectively compensate the brightness and ensure the further shading process. The proposed method separates the brightness compensation from the fabrication process, which is important for deepening the investigation of the hologram fabrication and achieving realistic 3D reconstruction.

Reducing the memory usage for effective computer-generated hologram calculation using compressed look-up table in full-color holographic display.

A fast algorithm with low memory usage is proposed to generate the hologram for full-color 3D display based on a compressed look-up table (C-LUT). The C-LUT is described and built to reduce the memory usage and speed up the calculation of the computer-generated hologram (CGH). Numerical simulations and optical experiments are performed to confirm this method, and several other algorithms are compared. The results show that the memory usage of the C-LUT is kept low when number of depth layers of the 3D object is increased, and the time for building the C-LUT is independent of the number of depth layers of the 3D object. The algorithm based on C-LUT is an efficient method for saving memory usage and calculation time, and it is expected that it could be used for realizing real-time and full-color 3D holographic display in the future.