Abstract BackgroundTo help clinicians provide timely treatment and delay disease progress, it is crucial to identifydementia patients during the mild cognitive impairment (MCI) stage and stratify these MCI patients into earlyand late MCI stages before they progress to alzheimer's disease (AD). In the process of diagnosing MCI andAD in living patients, brain scans are regularly collected using neuroimaging technologies such as computedtomography (CT), magnetic resonance imaging (MRI), or positron emission tomography (PET). These brainscans measure the volume and molecular activity within the brain resulting in a very promising avenue todiagnose patients early in a non-invasive manner.MethodsWe have developed an optimal transport based transfer learning model to discriminate betweenearly and late MCI. Combing this transfer learning model with bootstrap aggregation strategy, we overcomethe over-tting problem and improve model stability and prediction accuracy.ResultsWith the transfer learning methods that we have developed, we outperform the current state of theart MCI stage classication frameworks and show that it is crucial to leverage alzheimer's disease and normalcontrol subjects to accurately predict early and late stage cognitive impairment.ConclusionsOur method is the current state of the art based on benchmark comparisons. This method is anecessary technological stepping stone to widespread clinical usage of MRI based early detection of AD.
In this work, we propose and demonstrate a flexible capacitive tactile sensor array based on graphene served as electrodes. The sensor array consists of 3 × 3 units with 3 mm spatial resolution, similar to that of human skin. Each unit has three layers. The middle layer with microstructured PDMS served as an insulator is sandwiched by two perpendicular graphene-based electrodes. The size of each unit is 3 mm × 3 mm and the initial capacitance is about 0.2 pF. High sensitivities of 0.73 kPa[Formula: see text] between 0 and 1.2 kPa and 0.26 kPa[Formula: see text] between 1.2 and 2.5 kPa were achieved on the fabricated graphene pressure sensors. The proposed flexible pressure sensor array shows a great potential on the application of electric skin or 3D touch control.
A highly efficient composite sealing material was prepared using drilling cuttings as the base material and a binder, a coagulant, and other additives as auxiliaries. A four-factor, three-level orthogonal test was designed based on the response surface method (RSM), and a response surface regression model was constructed using compressive strength, fluidity, expansion rate, and setting time as performance indexes to analyze the effects of each factor on material performance and optimize the material proportion. The samples were prepared by simulating the grouting process, the permeability of the samples was measured, and the sealability of the material was verified by analyzing the material microscopic morphology. Results showed that the regression model had a high level of confidence and accuracy and could predict the test results accurately within the range of the test. The effects of the interaction between factors on material performance were also examined. The low permeability of the sealing material samples verified the material’s feasibility. Gradual optimization of material performance revealed that the optimal proportion was 52.6% drill cuttings, 44.3% binder, 0.6% coagulant promoter, and 2.5% expansive agent. Under these conditions, the error between the predicted and test values of each material property was less than 5%, and the comprehensive performance was superior. These findings verify the accuracy of RSM and its applicability to the optimization of material performance. This work provides reasonable theoretical guidance for the preparation of drilling cuttings composite (DC) materials in practical engineering.
Large resistive sensor arrays (RSAs) show great potential in tactile perception. However, the large number of sensors can result in great hardware overhead and bring difficulties for acquiring and processing mass data timely in transient measurement applications. This paper implements a field programmable gate array (FPGA)-based data processing system for a large RSA of 96 × 96, which shows good power consumption and high-speed wireless data update. For crosstalk-free measure, the zero potential method is improved with bus switches, leading to fewer operational amplifiers required and less negative power consumption. A real-time embedded data processing system is realized by FPGA for excellent parallel processing ability. A high-speed wireless transfer scheme with automatic regulated transfer size is proposed and realized by a wireless fidelity module, which allows timely data analysis at the remote end. Moreover, fault identification of RSAs fabricated by micro-electromechanical system technology is achieved. Tests carried out on a 32 × 32 RSA show that the total power consumption is 2209 mW, including 1261 mW of processors and 948 mW of readout circuits, corresponding to 2.15 mW/pixel. The total negative power consumption of 549 mW has been reduced by 50% compared with the zero potential method. The scanning speed is 400 fps, and the wireless transfer speed is up to 120 fps when the transceiver and receiver are 5 m apart.
Considering the hydrostatic pressure, the spontaneous and piezoelectric polarization, the dielectric mismatch, and 3D confinement of the electron and hole, the exciton states and interband optical transitions in [0001]-oriented wurtzite InxGa1−xN/GaN strained coupled quantum dot (QD) nanowire heterostructures (NWHETs) have been investigated by using the effective mass approximation, the simplified coherent potential approximation, and a variational approach. Our results show that the hydrostatic pressure, the strong built-in electric field (BEF), and the dielectric mismatch have a significant influence on the exciton states and interband optical transitions. The exciton binding energy increases almost linearly with the hydrostatic pressure for a given QD NWHET. The emission wavelength has a blue-shift (red-shift) if the hydrostatic pressure (QD height or the potential barrier thickness) increases. Our calculations also indicate that the radiative decay time has a quick increase with increasing of the QD height and the barrier thickness. The radiative decay time decreases if the hydrostatic pressure increases. The BEF (dielectric mismatch) dramatically decreases (increases) the exciton binding energy. The physical reason has been analyzed in depth.
Recent advances in decoding language from brain signals (EEG and MEG) have been significantly driven by pre-trained language models, leading to remarkable progress on publicly available non-invasive EEG/MEG datasets. However, previous works predominantly utilize teacher forcing during text generation, leading to significant performance drops without its use. A fundamental issue is the inability to establish a unified feature space correlating textual data with the corresponding evoked brain signals. Although some recent studies attempt to mitigate this gap using an audio-text pre-trained model, Whisper, which is favored for its signal input modality, they still largely overlook the inherent differences between audio signals and brain signals in directly applying Whisper to decode brain signals. To address these limitations, we propose a new multi-stage strategy for semantic brain signal decoding via vEctor-quantized speCtrogram reconstruction for WHisper-enhanced text generatiOn, termed BrainECHO. Specifically, BrainECHO successively conducts: 1) Discrete autoencoding of the audio spectrogram; 2) Brain-audio latent space alignment; and 3) Semantic text generation via Whisper finetuning. Through this autoencoding--alignment--finetuning process, BrainECHO outperforms state-of-the-art methods under the same data split settings on two widely accepted resources: the EEG dataset (Brennan) and the MEG dataset (GWilliams). The innovation of BrainECHO, coupled with its robustness and superiority at the sentence, session, and subject-independent levels across public datasets, underscores its significance for language-based brain-computer interfaces.
'/%3e%3cuse xlink:href='%23a' transform='rotate(23.036 .093 25.536)'/%3e%3cuse xlink:href='%23a' transform='rotate(45.87 1.273 16.18)'/%3e%3cuse xlink:href='%23a' transform='rotate(69.945 .996 12.078)'/%3e%3cuse xlink:href='%23a' transform='rotate(20.66 -19.689 31.932)'/%3e%3c/svg%3e)
