Deep Learning-Based Brain Computed Tomography Image Classification with Hyperparameter Optimization through Transfer Learning for Stroke.

Brain computed tomography (CT) is commonly used for evaluating the cerebral condition, but immediately and accurately interpreting emergent brain CT images is tedious, even for skilled neuroradiologists. Deep learning networks are commonly employed for medical image analysis because they enable efficient computer-aided diagnosis. This study proposed the use of convolutional neural network (CNN)-based deep learning models for efficient classification of strokes based on unenhanced brain CT image findings into normal, hemorrhage, infarction, and other categories. The included CNN models were CNN-2, VGG-16, and ResNet-50, all of which were pretrained through transfer learning with various data sizes, mini-batch sizes, and optimizers. Their performance in classifying unenhanced brain CT images was tested thereafter. This performance was then compared with the outcomes in other studies on deep learning-based hemorrhagic or ischemic stroke diagnoses. The results revealed that among our CNN-2, VGG-16, and ResNet-50 analyzed by considering several hyperparameters and environments, the CNN-2 and ResNet-50 outperformed the VGG-16, with an accuracy of 0.9872; however, ResNet-50 required a longer time to present the outcome than did the other networks. Moreover, our models performed much better than those reported previously. In conclusion, after appropriate hyperparameter optimization, our deep learning-based models can be applied to clinical scenarios where neurologist or radiologist may need to verify whether their patients have a hemorrhage stroke, an infarction, and any other symptom.

Functional abnormalities in the cortical processing of sound complexity and musical consonance in schizophrenia: evidence from an evoked potential study.

BACKGROUND: Previous studies have demonstrated functional and structural temporal lobe abnormalities located close to the auditory cortical regions in schizophrenia. The goal of this study was to determine whether functional abnormalities exist in the cortical processing of musical sound in schizophrenia. METHODS: Twelve schizophrenic patients and twelve age- and sex-matched healthy controls were recruited, and participants listened to a random sequence of two kinds of sonic entities, intervals (tritones and perfect fifths) and chords (atonal chords, diminished chords, and major triads), of varying degrees of complexity and consonance. The perception of musical sound was investigated by the auditory evoked potentials technique. RESULTS: Our results showed that schizophrenic patients exhibited significant reductions in the amplitudes of the N1 and P2 components elicited by musical stimuli, to which consonant sounds contributed more significantly than dissonant sounds. Schizophrenic patients could not perceive the dissimilarity between interval and chord stimuli based on the evoked potentials responses as compared with the healthy controls. CONCLUSION: This study provided electrophysiological evidence of functional abnormalities in the cortical processing of sound complexity and music consonance in schizophrenia. The preliminary findings warrant further investigations for the underlying mechanisms.

Single ZnO nanowire-PZT optothermal field effect transistors.

A new type of pyroelectric field effect transistor based on a composite consisting of single zinc oxide nanowire and lead zirconate titanate (ZnO NW-PZT) has been developed. Under infrared (IR) laser illumination, the transconductance of the ZnO NW can be modulated by optothermal gating. The drain current can be increased or decreased by IR illumination depending on the polarization orientation of the Pb(Zr(0.3)Ti(0.7))O(3) (PZT) substrate. Furthermore, by combining the photocurrent behavior in the UV range and the optothermal gating effect in the IR range, the wide spectrum of response of current by light offers a variety of opportunities for nanoscale optoelectronic devices.

MOS photodetectors based on Au-nanorod doped graphene electrodes.

By using Au-nanorod (Au-NR) doped graphene as a transparent conducting electrode, Si-based metal-oxide-semiconductor (MOS) photodetectors (PDs) exhibit high external quantum efficiency (EQE) and fast response time. It is found that upon adding Au-NRs to the graphene, a significant increase in EQE is observed for both planar and Si-nanotip (Si-NT) MOS PDs. The planar Si-based MOS PDs reveal a notable photoresponse with an EQE of 49% at the peak wavelength of 530 nm under zero bias and an EQE of 66% at the peak wavelength of 600 nm under - 0.4 V bias. For the Si-NTs MOS PD, it exhibits a relatively high EQE of 71% under - 4 V bias due to the effect of light trapping arising from the nature of the Si-NT array.

Direct evidence of type II band alignment in nanoscale P3HT/CdSe heterostructures.

Due to inherent advantages of both constituent materials, organic/inorganic hybrid composites have attracted increasing attention. One of the fundamental issues needed to be resolved is their band alignment, which governs most of the electrical and optical properties. Here, we report the investigation of optical transition in poly(3-hexylthiophene) (P3HT)/CdSe nano-composites (NCs). It is found that the relaxation dynamics of photo-carriers in NCs is dominated by charge separation effects. Based on the band bending effect and the quantum confinement energy of electrons in the conduction band of CdSe quantum dots, we provide direct evidence of type II band alignment in P3HT/CdSe NCs. The establishment of a type II transition in NCs is very useful for the future design of efficient optoelectronic devices based on conjugated polymer/semiconductor hybrid systems.

Efficient light harvesting by photon downconversion and light trapping in hybrid ZnS nanoparticles/Si nanotips solar cells.

A hybrid colloidal ZnS nanoparticles/Si nanotips p-n active layer has been demonstrated to have promising potential for efficient solar spectrum utilization in crystalline silicon-based solar cells. The hybrid solar cell shows an enhancement of 20% in the short-circuit current and approximately 10% in power conversion efficiency compared to its counterpart without integrating ZnS nanoparticles. The enhancement has been investigated by external quantum efficiency, photoluminescence excitation spectrum, photoluminescence, and reflectance to distinct the role of ZnS quantum dots for light harvesting. It is concluded that ZnS nanoparticles not only act as frequency downconversion centers in the ultraviolet region but also serve as antireflection coating for light trapping in the measured spectral regime. Our approach is ready to be extended to many other material systems for the creation of highly efficient photovoltaic devices.

Giant white and blue light emission from Al(2)O(3) and ZnO nanocomposites.

A new and general approach enabling us to amplify not only the bandgap emission of ZnO nanorods but also the defect emission of Al(2)O(3) is proposed. The light intensity of the band edge emission of ZnO nanorods can be improved by as much as 19 times after the decoration of Al(2)O(3) layers. Moreover, white light emission arising from Al(2)O(3) defects in ZnO/Al(2)O(3) nanostructures also shows a large enhancement factor of 12 times. Our new strategy offers an alternative possibility to create strong white and blue light-emitting devices.