Explainable Deep Learning Reproduces a 'Professional Eye' on the Diagnosis of Internal Disorders in Persimmon Fruit.

Recent rapid progress in deep neural network techniques has allowed recognition and classification of various objects, often exceeding the performance of the human eye. In plant biology and crop sciences, some deep neural network frameworks have been applied mainly for effective and rapid phenotyping. In this study, beyond simple optimizations of phenotyping, we propose an application of deep neural networks to make an image-based internal disorder diagnosis that is hard even for experts, and to visualize the reasons behind each diagnosis to provide biological interpretations. Here, we exemplified classification of calyx-end cracking in persimmon fruit by using five convolutional neural network models with various layer structures and examined potential analytical options involved in the diagnostic qualities. With 3,173 visible RGB images from the fruit apex side, the neural networks successfully made the binary classification of each degree of disorder, with up to 90% accuracy. Furthermore, feature visualizations, such as Grad-CAM and LRP, visualize the regions of the image that contribute to the diagnosis. They suggest that specific patterns of color unevenness, such as in the fruit peripheral area, can be indexes of calyx-end cracking. These results not only provided novel insights into indexes of fruit internal disorders but also proposed the potential applicability of deep neural networks in plant biology.

Novel neurotrophic phenylbutenoids from Indonesian ginger Bangle, Zingiber purpureum.

Two new curcuminoids 1 and 2, and a new phenylbutenoid dimer 3, were isolated from Bangle (Zingiber purpureum). Their structures were determined on the basis of comprehensive spectroscopic data and their biogenetic pathway. Compounds 1 and 2 are the first example of curcumin coupled with phenylbutenoid. Compounds 1 and 2 promoted neurite outgrowth of NGF-mediated PC12 cells at concentrations ranging from 1 to 10 µM. In addition, compound 1 was found to accelerate the prevention of Aß42 aggregation.

Updating ``Medical radar signal dataset for non-contact respiration and heart rate measurement" with expanded data on laboratory rats under isoflurane anesthesia.

The dataset presented in this article is an update of the dataset provided by K. Edanami and G. Sun entitled "Medical Radar Signal Dataset for Non-Contact Respiration and Heart Rate Measurement" [1]. The new dataset includes radar signals and reference laser measurements from experiments conducted on anesthetized rats. The rats were placed in a prone position, and isoflurane was administered in varying concentrations to maintain anesthesia. A 24 GHz radar and laser sensor were positioned above the rats to capture the necessary data. The dataset contains time-stamped radar I and Q channel signals as well as laser measurements. Additionally, MATLAB code for signal visualization and FFT (fast Fourier transform)-based respiration rate estimation is provided. This comprehensive dataset and accompanying MATLAB code facilitate the advancement of non-invasive respiration measurement techniques in small animals, with potential applications in biomedical research.

Magnetic and Transport Properties of New Dual-Phase High-Entropy Alloy FeRhIrPdPt.

High-entropy alloys (HEAs) are broadly explored from the perspective of mechanical, corrosion-resistance, catalytic, structural, superconducting, magnetic properties, and so on. In magnetic HEAs, 3d transition metals or rare-earth elements are well-studied compositional elements. We researched a magnetic HEA containing Fe combined with 4d and 5d transition metals, which has not been well investigated, and found a new dual-phase face-centered-cubic (fcc) HEA FeRhIrPdPt. The structural, magnetic, and transport properties were evaluated by assuming that FeRhIrPdPt is a mixture of FeRh4, FeIr4, FePd4, and FePt4, all with the fcc structure. The dual-phase is composed of a Rh- and Ir-rich main phase and a Pd- and Pt-rich minor one. FeRh4 and FeIr4 show spin freezings at low temperatures, while FePd4 and FePt4 are ferromagnetic. Two magnetic features can characterize FeRhIrPdPt. One is the canonical spin-glass transition at 90 K, and the other is a ferromagnetic correlation that appears below 300 K. The main and minor phases were responsible for the spin-glass transition and the ferromagnetic correlation below 300 K, respectively.

Hand motion recognition based on forearm deformation measured with a distance sensor array.

Studies of upper limb motion analysis using surface electromyogram (sEMG) signals measured from the forearm plays an important role in various applications, such as human interfaces for controlling robotic exoskeletons, prosthetic hands, and evaluation of body functions. Though the sEMG signals have a lot of information about the activities of the muscles, the signals do not have the activities of the deep layer muscles. We focused on forearm deformation, since hand motion brings the muscles, tendons, and skeletons under the skin. The reason why we focus is that we believe the forearm deformation delivers information about the activities of deep layer muscles. In this paper, we propose a hand motion recognition method based on the forearm deformation measured with a distance sensor array. The method uses the support vector machine. Our method achieved a mean accuracy of 92.6% for seven hand motions. Because the accuracy of the pronation and the supination are high, the distance sensor array has the potential to estimate the activities of deep layer muscles.