Auditory cues modulate the short timescale dynamics of STN activity during stepping in Parkinson's disease.

BACKGROUND: Gait impairment has a major impact on quality of life in patients with Parkinson's disease (PD). It is believed that basal ganglia oscillatory activity at ß frequencies (15-30 Hz) may contribute to gait impairment, but the precise dynamics of this oscillatory activity during gait remain unclear. Additionally, auditory cues are known to lead to improvements in gait kinematics in PD. If the neurophysiological mechanisms of this cueing effect were better understood they could be leveraged to treat gait impairments using adaptive Deep Brain Stimulation (aDBS) technologies. OBJECTIVE: We aimed to characterize the dynamics of subthalamic nucleus (STN) oscillatory activity during stepping movements in PD and to establish the neurophysiological mechanisms by which auditory cues modulate gait. METHODS: We studied STN local field potentials (LFPs) in eight PD patients while they performed stepping movements. Hidden Markov Models (HMMs) were used to discover transient states of spectral activity that occurred during stepping with and without auditory cues. RESULTS: The occurrence of low and high ß bursts was suppressed during and after auditory cues. This manifested as a decrease in their fractional occupancy and state lifetimes. Interestingly, α transients showed the opposite effect, with fractional occupancy and state lifetimes increasing during and after auditory cues. CONCLUSIONS: We show that STN oscillatory activity in the α and ß frequency bands are differentially modulated by gait-promoting oscillatory cues. These findings suggest that the enhancement of α rhythms may be an approach for ameliorating gait impairments in PD.

A pilot study on a patient with refractory headache: Personalized deep brain stimulation through stereoelectroencephalography.

The integration of stereoelectroencephalography with therapeutic deep brain stimulation (DBS) holds immense promise as a viable approach for precise treatment of refractory disorders, yet it has not been explored in the domain of headache or pain management. Here, we implanted 14 electrodes in a patient with refractory migraine and integrated clinical assessment and electrophysiological data to investigate personalized targets for refractory headache treatment. Using statistical analyses and cross-validated machine-learning models, we identified high-frequency oscillations in the right nucleus accumbens as a critical headache-related biomarker. Through a systematic bipolar stimulation approach and blinded sham-controlled survey, combined with real-time electrophysiological data, we successfully identified the left dorsal anterior cingulate cortex as the optimal target for the best potential treatment. In this pilot study, the concept of the herein-proposed data-driven approach to optimizing precise and personalized treatment strategies for DBS may create a new frontier in the field of refractory headache and even pain disorders.

Quantifying Sharpness and Nonlinearity in Neonatal Seizure Dynamics.

The integration of multiple electrophysiological biomarkers is crucial for monitoring neonatal seizure dynamics. The present study aimed to characterize the temporal dynamics of neonatal seizures by analyzing intrinsic waveforms of epileptic electroencephalogram (EEG) signals. We proposed a complementary set of methods considering envelope power, focal sharpness changes, and nonlinear patterns of EEG signals of 79 neonates with seizures. Features derived from EEG signals were used as input to the machine learning classifier. All three characteristics were significantly elevated during seizure events, as agreed upon by all viewers (P < 0.0001). Envelope power was elevated in the entire seizure period, and the degree of nonlinearity rose at the termination of a seizure event. Epileptic sharpness effectively characterizes an entire seizure event, complementing the role of envelope power in identifying its onset. However, the degree of nonlinearity showed superior discriminability for the termination of a seizure event. The proposed computational methods for intrinsic sharp or nonlinear EEG patterns evolving during neonatal seizure could share some features with envelope power. Current findings may be helpful in developing strategies to improve neonatal seizure monitoring.

Spatiotemporal convolution sleep network based on graph attention mechanism with automatic feature extraction.

BACKGROUND AND OBJECTIVE: Graph neural networks (GNNs) are widely used for automatic sleep staging. However, the majority of GNNs are based on spectral approaches, as far as we know, which heavily depend on the Laplacian eigenbasis determined by the graph structure with a large computing cost. METHODS: We introduced a non-spectral approach named graph attention networks v2 (GATv2) as the core of our network to extract spatial information (S-GATv2 in our work), which is more flexible and intuitive than the routined spectral method. Meanwhile, to resolve the issue of weak generalization of using traditional feature extraction, the multi-convolutional layers are implemented to automatically extract features. In this work, the proposed spatiotemporal convolution sleep network (ST-GATv2) consists of multi-convolution layers and a GATv2 block. Of note, the graph attention technique to the time domain was applied to construct temporal GATv2 (T-GATv2), which intends to capture the connection between two channels in the adjacent sleep stages. Besides, the modified function is further proposed to capture the hidden changing trend information by the difference in the feature's value of the two adjacent stages. RESULTS: In our experiment, we used the SS3 datasets in the MASS as our test datasets to compare with other advanced models. Our result reveals our model achieves the highest accuracy at 89.0 %. Besides, the proposed T-GATv2 block and modified function bring an approximate 0.5 % improvement in Kappa and F1-score. CONCLUSIONS: Our results support the potential of graph attention mechanisms and creative blocks (T-GATv2 and modified function) in sleep classification. We suggest the proposed ST-GATv2 model as an effective tool in sleep staging in either healthy or diseased states.

A novel dynamic cardiorespiratory coupling quantification method reveals the effect of aging on the autonomic nervous system.

Traditional cardiopulmonary coupling (CPC) based on the Fourier transform shares an inherent trade-off between temporal and frequency resolutions with fixed window designs. Therefore, a cross-wavelet cardiorespiratory coupling (CRC) method was developed to highlight interwave cardiorespiratory dynamics and applied to evaluate the age effect on the autonomic regulation of cardiorespiratory function. The cross-wavelet CRC visualization successfully reflected dynamic alignments between R-wave interval signal (RR intervals) and respiration. Strong and continuous CRC was shown if there was perfect temporal coordination between consecutive R waves and respiration, while CRC becomes weaker and intermittent without such coordination. Using real data collected on electrocardiogram (ECG) and respiratory signals, the heart rate variability (HRV) and CRC were calculated. Subsequently, comparisons were conducted between young and elderly individuals. Young individuals had significantly higher partial time and frequency HRV indices than elderly individuals, indicating stronger control of parasympathetic regulation. The overall coupling strength of the CRC of young individuals was higher than that of elderly individuals, especially in high-frequency power, which was significantly lower in the elderly group than in the young group, achieving better results than the HRV indices in terms of statistical significance. Further analyses of the time-frequency dynamics of CRC indices revealed that the coupling strength was consistently higher in the high-frequency (HF) band (0.15-0.4 Hz) in young participants compared to elderly individuals. The dynamic CRC between respiration and HRV indices was accessible by integrating the cross-wavelet spectrum and coherence. Young participants had a significantly higher level of CRC in the HF band, indicating that aging reduces vagus nerve modulation.

A Novel Sleep Staging Method Based on EEG and ECG Multimodal Features Combination.

Accurate sleep staging evaluates the quality of sleep, supporting the clinical diagnosis and intervention of sleep disorders and related diseases. Although previous attempts to classify sleep stages have achieved high classification performance, little attention has been paid to integrating the rich information in brain and heart dynamics during sleep for sleep staging. In this study, we propose a generalized EEG and ECG multimodal feature combination to classify sleep stages with high efficiency and accuracy. Briefly, a hybrid features combination in terms of multiscale entropy and intrinsic mode function are used to reflect nonlinear dynamics in multichannel EEGs, along with heart rate variability measures over time/frequency domains, and sample entropy across scales are applied for ECGs. For both the max-relevance and min-redundancy method and principal component analysis were used for dimensionality reduction. The selected features were classified by four traditional machine learning classifiers. Macro-F1 score, macro-geometric mean, and Cohen kappa value are adopted to evaluate the classification performance of each class in an imbalanced dataset. Experimental results show that EEG features contribute more to wake stage classification while ECG features contribute more to deep sleep stages. The proposed combination achieves the highest accuracy of 84.3% and the highest kappa value of 0.794 on the support vector machine in the ISRUC-S3 dataset, suggesting the proposed multimodal features combination is promising in accuracy and efficiency compared to other state-of-the-art methods.

Flexible Data Rate Allocation Using Non-Orthogonal Multiple Access (NOMA) in a Mode Division Multiplexing (MDM) Optical Power Splitter for System-on-Chip Networks.

We put forward and demonstrate a silicon photonics (SiPh)-based mode division multiplexed (MDM) optical power splitter that supports transverse-electric (TE) single-mode, dual-mode, and triple-mode (i.e., TE0, TE1, and TE2). An optical power splitter is needed for optical signal distribution and routing in optical interconnects. However, a traditional optical splitter only divides the power of the input optical signal. This means the same data information is received at all the output ports of the optical splitter. The powers at different output ports may change depending on the splitting ratio of the optical splitter. The main contributions of our proposed optical splitter are: (i) Different data information is received at different output ports of the optical splitter via the utilization of NOMA. By adjusting the power ratios of different channels in the digital domain (i.e., via software control) at the Tx, different channel data information can be received at different output ports of the splitter. It can increase the flexibility of optical signal distribution and routing. (ii) Besides, the proposed optical splitter can support the fundamental TE0 mode and the higher modes TE1, TE2, etc. Supporting mode-division multiplexing and multi-mode operation are important for future optical interconnects since the number of port counts is limited by the chip size. This can significantly increase the capacity besides wavelength division multiplexing (WDM) and spatial division multiplexing (SDM). The integrated SiPh MDM optical power splitter consists of a mode up-conversion section implemented by asymmetric directional couplers (ADCs) and a Y-branch structure for MDM power distribution. Here, we also propose and discuss the use of the Genetic algorithm (GA) for the MDM optical power splitter parameter optimization. Finally, to provide adjustable data rates at different output ports after the MDM optical power splitter, non-orthogonal multiple access-orthogonal frequency division multiplexing (NOMA-OFDM) is also employed. Experimental results validate that, in three modes (TE0, TE1, and TE2), user-1 and user-2 achieve data rates of (user-1: greater than 22 Gbit/s; user-2: greater than 12 Gbit/s) and (user-1: greater than 12 Gbit/s; user-2: 24 Gbit/s), respectively, at power-ratio (PR) = 2.0 or 3.0. Each channel meets the hard-decision forward-error-correction (HD-FEC, i.e., BER = 3.8 × 10-3) threshold. The proposed method allows flexible data rate allocation for multiple users for optical interconnects and system-on-chip networks.

Bi-Directional Free-Space Visible Light Communication Supporting Multiple Moveable Clients Using Light Diffusing Optical Fiber.

In this work, we put forward and demonstrate a bi-direction free-space visible light communication (VLC) system supporting multiple moveable receivers (Rxs) using a light-diffusing optical fiber (LDOF). The downlink (DL) signal is launched from a head-end or central office (CO) far away to the LDOF at the client side via a free-space transmission. When the DL signal is launched to the LDOF, which acts as an optical antenna to re-transmit the DL signal to different moveable Rxs. The uplink (UL) signal is sent via the LDOF towards the CO. In a proof-of-concept demonstration, the LDOF is 100 cm long, and the free space VLC transmission between the CO and the LDOF is 100 cm. 210 Mbit/s DL and 850 Mbit/s UL transmissions meet the pre-forward-error-correction bit error rate (pre-FEC BER = 3.8 × 10-3) threshold.

Unmanned-aerial-vehicle based optical camera communication system using light-diffusing fiber and rolling-shutter image-sensor.

We put forward and demonstrate a light-diffusing fiber equipped unmanned-aerial-vehicle (UAV) to provide a large field-of-view (FOV) optical camera communication (OCC) system. The light-diffusing fiber can act as a bendable, lightweight, extended and large FOV light source for the UAV-assisted optical wireless communication (OWC). During UAV flying, the light-diffusing fiber light source could be tilted or bended; hence, offering large FOV as well as supporting large receiver (Rx) tilting angle are particularly important for the UAV-assisted OWC systems. To improve the transmission capacity of the OCC system, one method based on the camera shutter mechanism, which is known as rolling-shuttering is utilized. The rolling-shuttering method makes use of the feature of complementary-metal-oxide-semiconductor (CMOS) image sensor to extract signal pixel-row by pixel-row. The data rate can be significantly increased since the capture start time for each pixel-row is different. As the light-diffusing fiber is thin and occupies only a few pixels in the CMOS image frame, Long-short-term-memory neural-network (LSTM-NN) is used to enhance the rolling-shutter decoding. Experimental results show that the light-diffusing fiber can satisfactorily act as an "omnidirectional optical antenna" providing wide FOVs and 3.6 kbit/s can be achieved, accomplishing the pre-forward error correction bit-error-rate (pre-FEC BER = 3.8 × 10-3).

Cross-Frequency Coupling and Intelligent Neuromodulation.

Cross-frequency coupling (CFC) reflects (nonlinear) interactions between signals of different frequencies. Evidence from both patient and healthy participant studies suggests that CFC plays an essential role in neuronal computation, interregional interaction, and disease pathophysiology. The present review discusses methodological advances and challenges in the computation of CFC with particular emphasis on potential solutions to spurious coupling, inferring intrinsic rhythms in a targeted frequency band, and causal interferences. We specifically focus on the literature exploring CFC in the context of cognition/memory tasks, sleep, and neurological disorders, such as Alzheimer's disease, epilepsy, and Parkinson's disease. Furthermore, we highlight the implication of CFC in the context and for the optimization of invasive and noninvasive neuromodulation and rehabilitation. Mainly, CFC could support advancing the understanding of the neurophysiology of cognition and motor control, serve as a biomarker for disease symptoms, and leverage the optimization of therapeutic interventions, e.g., closed-loop brain stimulation. Despite the evident advantages of CFC as an investigative and translational tool in neuroscience, further methodological improvements are required to facilitate practical and correct use in cyborg and bionic systems in the field.

A novel measure of cardiopulmonary coupling during sleep based on the synchrosqueezing transform algorithm.

OBJECTIVE: This paper presents a novel method to quantify cardiopulmonary dynamics for automatic sleep apnea detection by integrating the synchrosqueezing transform (SST) algorithm with the standard cardiopulmonary coupling (CPC) method. METHODS: Simulated data were designed to validate the reliability of the proposed method, with varying levels of signal bandwidth and noise contamination. Real data were collected from the Physionet sleep apnea database, consisting of 70 single-lead ECGs with expert-labeled apnea annotations on a minute-by-minute basis. Three different signal processing techniques applied to sinus interbeat interval and respiratory time series include short-time Fourier transform, continuous Wavelet transform, and synchrosqueezing transform, respectively. Subsequently, the CPC index was computed to construct sleep spectrograms. Features derived from such spectrogram were used as input to five machine- learning-based classifiers including decision trees, support vector machines, k-nearest neighbors, etc. Results: The simulation results showed that the SST-CPC method is robust to both noise level and signal bandwidth, outperforming Fourier-based and Wavelet-based approaches. Meanwhile, the SST-CPC spectrogram exhibited relatively explicit temporal-frequency biomarkers compared with the rest. Furthermore, by integrating SST-CPC features with common-used heart rate and respiratory features, accuracies for per-minute apnea detection improved from 72% to 83%, validating the added value of CPC biomarkers in sleep apnea detection. CONCLUSION: The SST-CPC method improves the accuracy of automatic sleep apnea detection and presents comparable performances with those automated algorithms reported in the literature. SIGNIFICANCE: The proposed SST-CPC method enhances sleep diagnostic capabilities, and may serve as a complementary tool to the routine diagnosis of sleep respiratory events.

Variational Phase-Amplitude Coupling Characterizes Signatures of Anterior Cortex under Emotional Processing.

Emotion, an essential aspect in inferring human psychological states, is featured by entangled oscillators operating at multiple frequencies and montages. However, the dynamics of mutual interactions among rhythmic activities in EEGs under various emotional expressions are unclear. To this end, a novel method named variational phase-amplitude coupling is proposed to quantify the rhythmic nesting structure in EEGs under emotional processing. The proposed algorithm lies in variational mode decomposition, featured by its robustness to noise artifacts and its merit in avoiding the mode-mixing problem. This novel method reduces the risk of spurious coupling compared to that with ensemble empirical mode decomposition or iterative filter when evaluated by simulations. An atlas of cross-couplings in EEGs under eight emotional processing is established. Mainly, α activity in the anterior frontal region serves as a critical sign for neutral emotional state, whereas γ amplitude seems to be linked with both positive and negative emotional states. Moreover, for those γ-amplitude-related couplings under neutral emotional state, the frontal lobe is associated with lower phase-given frequencies while the central lobe is attached to higher ones. The γ-amplitude-related coupling in EEGs is a promising biomarker for recognizing mental states. We recommend our method as an effective tool in characterizing the entangled multifrequency rhythms in brain signals for emotion neuromodulation.

Design Consideration, Numerical and Experimental Analyses of Mode-Division-Multiplexed (MDM) Silicon Photonics Integrated Circuit with Sharp Bends.

Due to the popularity of different high bandwidth applications, it is becoming increasingly difficult to satisfy the huge data capacity requirements, since the traditional electrical interconnects suffer significantly from limited bandwidth and huge power consumption. Silicon photonics (SiPh) is one of the important technologies for increasing interconnect capacity and decreasing power consumption. Mode-division multiplexing (MDM) allows signals to be transmitted simultaneously, at different modes, in a single waveguide. Wavelength-division multiplexing (WDM), non-orthogonal multiple access (NOMA) and orthogonal-frequency-division multiplexing (OFDM) can also be utilized to further increase the optical interconnect capacity. In SiPh integrated circuits, waveguide bends are usually inevitable. However, for an MDM system with a multimode bus waveguide, the modal fields will become asymmetric when the waveguide bend is sharp. This will introduce inter-mode coupling and inter-mode crosstalk. One simple approach to achieve sharp bends in multimode bus waveguide is to use a Euler curve. Although it has been reported in the literature that sharp bends based on a Euler curve allow high performance and low inter-mode crosstalk multimode transmissions, we discover, by simulation and experiment, that the transmission performance between two Euler bends is length dependent, particularly when the bends are sharp. We investigate the length dependency of the straight multimode bus waveguide between two Euler bends. High transmission performance can be achieved by a proper design of the waveguide length, width, and bend radius. By using the optimized MDM bus waveguide length with sharp Euler bends, proof-of-concept NOMA-OFDM experimental transmissions, supporting two MDM modes and two NOMA users, are performed.

Wideband and Channel Switchable Mode Division Multiplexing (MDM) Optical Power Divider Supporting 7.682 Tbit/s for On-Chip Optical Interconnects.

Silicon photonics (SiPh) are considered a promising technology for increasing interconnect speed and capacity while decreasing power consumption. Mode division multiplexing (MDM) enables signals to be transmitted in different orthogonal modes in a single waveguide core. Wideband MDM components simultaneously supporting wavelength division multiplexing (WDM) and orthogonal frequency-division multiplexing (OFDM) can significantly increase the transmission capacity for optical interconnects. In this work, we propose, fabricate and demonstrate a wideband and channel switchable MDM optical power divider on an SOI platform, supporting single, dual and triple modes. The switchable MDM power divider consists of two parts. The first part is a cascaded Mach-Zehnder interferometer (MZI) for switching the data from their original TE0, TE1 and TE2 modes to different modes among themselves. After the target modes are identified, mode up-conversion and Y-branch are utilized in the second part for the MDM power division. Here, 48 WDM wavelength channels carrying OFDM data are successfully switched and power divided. An aggregated capacity of 7.682 Tbit/s is achieved, satisfying the pre-forward error correction (pre-FEC) threshold (bit-error-rate, BER = 3.8 × 10-3). Although up to three MDM modes are presented in the proof-of-concept demonstration here, the proposed scheme can be scaled to higher order modes operation.

Brain Complexity Predicts Response to Adrenocorticotropic Hormone in Infantile Epileptic Spasms Syndrome: A Retrospective Study.

INTRODUCTION: Infantile epileptic spasms syndrome (IESS) is an age-specific and severe epileptic encephalopathy. Although adrenocorticotropic hormone (ACTH) is currently considered the preferred first-line treatment, it is not always effective and may cause side effects. Therefore, seeking a reliable biomarker to predict the treatment response could benefit clinicians in modifying treatment options. METHODS: In this study, the complexities of electroencephalogram (EEG) recordings from 15 control subjects and 40 patients with IESS before and after ACTH therapy were retrospectively reviewed using multiscale entropy (MSE). These 40 patients were divided into responders and nonresponders according to their responses to ACTH. RESULTS: The EEG complexities of the patients with IESS were significantly lower than those of the healthy controls. A favorable response to treatment showed increasing complexity in the γ band but exhibited a reduction in the ß/α-frequency band, and again significantly elevated in the δ band, wherein the latter was prominent in the parieto-occipital regions in particular. Greater reduction in complexity was significantly linked with poorer prognosis in general. Occipital EEG complexities in the γ band revealed optimized performance in recognizing response to the treatment, corresponding to the area under the receiver operating characteristic curves as 0.8621, while complexities of the δ band served as a fair predictor of unfavorable outcomes globally. CONCLUSION: We suggest that optimizing frequency-specific complexities over critical brain regions may be a promising strategy to facilitate predicting treatment response in IESS.

Optical Beam Steerable Visible Light Communication (VLC) System Supporting Multiple Users Using RGB and Orthogonal Frequency Division Multiplexed (OFDM) Non-Orthogonal Multiple Access (NOMA).

In order to achieve high-capacity visible light communication (VLC), five dimensions in physics, including frequency, time, quadrature modulation, space, and polarization can be utilized. Orthogonality should be maintained in order to reduce the crosstalk among different dimensions. In this work, we illustrate a high-capacity 21.01 Gbit/s optical beam steerable VLC system with vibration mitigation based on orthogonal frequency division multiplexed (OFDM) non-orthogonal multiple access (NOMA) signals using red, green, and blue (RGB) laser-diodes (LDs). The OFDM-NOMA can increase the spectral efficiency of VLC signal by allowing high overlapping of different data channel spectra in the power domain to maximize the bandwidth utilization. In the NOMA scheme, different data channels are digitally multiplexed using different levels of power with superposition coding at the transmitter (Tx). Successive interference cancellation (SIC) is then utilized at the receiver (Rx) to retrieve different power multiplexed data channels. The total data rates (i.e., Data 1 and Data 2) achieved by the R/G/B OFDM-NOMA channels are 8.07, 6.62, and 6.32 Gbit/s, respectively, achieving an aggregated data rate of 21.01 Gbit/s. The corresponding average signal-to-noise ratios (SNRs) of Data 1 in the R, G, and B channels are 9.05, 9.18 and 8.94 dB, respectively, while that of Data 2 in the R, G, and B channels are 14.92, 14.29, and 13.80 dB, respectively.

3D Visible Light-Based Indoor Positioning System Using Two-Stage Neural Network (TSNN) and Received Intensity Selective Enhancement (RISE) to Alleviate Light Non-Overlap Zones.

The high precision three-dimensional (3D) visible light-based indoor positioning (VLIP) systems have gained much attention recently for people or robot navigation, access tracking, etc. In this work, we put forward and present the first demonstration, up to the authors' knowledge, of a 3D VLIP system utilizing a two-stage neural network (TSNN) model. The positioning performance would degrade when the distance between the light emitting diode (LED) plane and the receiver (Rx) plane increases; however, because of the finite LED field-of-view (FOV), light non-overlap zones are created. These light non-overlap zones will produce high positioning error particularly for the 3D VLIP systems. Here, we also propose and demonstrate the Received-Intensity-Selective-Enhancement scheme, known as RISE, to alleviate the light non-overlap zones in the VLIP system. In a practical test-room with dimensions of 200 × 150 × 300 cm3, the experimental results show that the mean errors in the training and testing data sets are reduced by 54.1% and 27.9% when using the TSNN model with RISE in the z-direction, and they are reduced by 39.1% and 37.8% in the xy-direction, respectively, when comparing that with using a one stage NN model only. At the cumulative distribution function (CDF) P90, the TSNN model with RISE can reduce the errors by 36.78% when compared with that in the one stage NN model.

Using lighting design tool to simplify the visible light positioning plan and reduce the deep learning loading.

We put forward and transform the commercially available lighting design software into an indoor visible light positioning (VLP) design tool. The proposed scheme can work well with different deep learning methods for reducing the loading of training data set collection. The indoor VLP models under evaluation include second order regression, fully-connected neural-network (FC-NN), and convolutional neural-network (CNN). Experimental results show that the similar positioning accuracy can be obtained when the indoor VLP models are trained with experimentally acquired data set or trained with software obtained data set. Hence, the proposed method can reduce the training loading for the indoor VLP.

Optical camera communication (OCC) using a laser-diode coupled optical-diffusing fiber (ODF) and rolling shutter image sensor.

We demonstrate an optical-camera-communication (OCC) system utilizing a laser-diode (LD) coupled optical-diffusing-fiber (ODF) transmitter (Tx) and rolling-shutter based image sensor receiver (Rx). The ODF is a glass optical fiber produced for decorative lighting or embedded into small areas where bulky optical sources cannot fit. Besides, decoding the high data rate rolling-shutter pattern from the thin ODF Tx is very challenging. Here, we propose and experimentally demonstrate the pixel-row-per-bit based neural-network (PPB-NN) to decode the rolling-shutter-pattern emitted by the thin ODF Tx. The proposed PPB-NN algorithm is discussed. The proposed PPB-NN method can satisfy the pre-forward error correction (FEC) BER at data rate of 3,300 bit/s at a transmission distance of 35 cm. Theoretical analysis of the maximum ODF Tx angle is also discussed; and our experimental values agree with our theoretical results.