Predicting spatial gene expression from hematoxylin and eosin (H&E)-stained histological images is a central challenge in the field of computational pathology. Although many models leverage the spatial context from entire whole-slide images to boost performance, the predictive capability achievable from single, isolated tissue patches remains unexplored. Without this fundamental baseline, it is difficult to fairly measure the marginal benefit of incorporating spatial context. In this study, we established a lightweight, efficient, and reproducible baseline for single-patch gene expression prediction, thereby providing a standardized foundation for future research. We trained various convolution-based pretrained architectures, such as EfficientNet, ResNet, and DenseNet, using a fixed data split and seed on a curated human liver Visium dataset. To assess cross-tissue generalizability, we further validated the proposed baseline on an independent breast cancer spatial transcriptomics dataset. Through full fine-tuning and morphology-preserving augmentation, EfficientNet-B0 achieved a maximum Pearson correlation coefficient (PCC) of 0.310 for highly expressed genes on liver tissue, surpassing previous single-patch methods. Furthermore, the model predicted 50 genes with PCC ≥ 0.30 using only 5.3M parameters--2.5× more genes than a ResNet-50-based baseline despite using 5× fewer parameters-suggesting a favorable performance-efficiency tradeoff within this single-patch setting. On the independent breast cancer dataset, EfficientNet-B0 consistently achieved best or near-best performance across all gene sets under both single-donor and multi-donor evaluations, demonstrating stable generalization across tissue types. Subsequent biological analysis reveals that model prediction accuracy is primarily driven by the strength of spatial organization in H&E histology, measured by Moran's I spatial autocorrelation, rather than transcript abundance or chemical properties. Consequently, this work provides a reference for quantifying the contributions of spatial context, multiresolution integration, and architectural complexity in future spatial transcriptomics models.

Voice activity detection (VAD) involves discriminating speech segments from background noise and is a critical step in numerous speech‐related applications. However, distinguishing speech from noise based on the properties of noise is fallible, because it is difficult to predict and characterise the noise occurring in real life. In this study, the authors instead focus on the intrinsic characteristics of speech. The harmonic peaks of vowel sounds have higher energies than the other spectral components of speech and are the speech features most likely to survive in most cases of severe noise. Therefore, the energy differences between harmonic peaks and other spectral features show promise for enabling robust VAD. To exploit this feature, the harmonic peaks must be accurately located. For this purpose, this study proposes an efficient harmonic peak location detection (HPD) method. Based on extensive experiments conducted in the presence of various noise types and signal‐to‐noise ratios, we found that VAD with the proposed HPD approach outperforms existing VAD methods and does so with reasonable computational cost and higher robustness.

Increasing attention is being paid to soft errors due to their high incidence. Although various approaches have been developed to mitigate such soft errors in digital storage components such as latches and flip-flops, most of them can detect only a single soft error and correct it only in limited cases, with a high overhead of hardware resources. In this paper, we propose a new self-repairing digital system based on the state attractor-converging mechanism, inspired by the recovery process of a living cell. To implement the state attractor-converging mechanism, the proposed system introduces additional states other than a working state, such that any abnormal transition to an additional state caused by a soft error can be recovered immediately. Moreover, the proposed system employs a reconfiguration method that enables a system with an erroneous state to be recovered regardless of any working state of the system while having a similar hardware overhead compared with the existing systems. From simulation analysis and experimental verification, we show that the proposed system can detect and correct not only a single soft error without any restriction but also two simultaneous soft errors under a few conditions. Furthermore, the proposed system can recover a soft error occurring in the control hardware used for error detection and correction, which is not possible with any other existing approaches.

Self-repairing digital systems have recently emerged as the most promising alternative for fault-tolerant systems. However, such systems are still impractical in many cases, particularly due to the complex rerouting process that follows cell replacement. They lose efficiency when the circuit size increases, due to the extra hardware in addition to the functional circuit and the unutilization of normal operating hardware for fault recovery. In this paper, we propose a system inspired by endocrine cellular communication, which simplifies the rerouting process in two ways: 1) by lowering the hardware overhead along with the increasing size of the circuit and 2) by reducing the hardware unutilized for fault recovery while maintaining good fault-coverage. The proposed system is composed of a structural layer and a gene-control layer. The structural layer consists of novel modules and their interconnections. In each module of our system, the encoded data, called the genome, contains information about the function and the connection. Therefore, a faulty module can be replaced and the whole system's functions and connections are maintained by simply assigning the same encoded data to a spare (stem) module. In existing systems, a huge amount of hardware, such as a dynamic routing system, is required for such an operation. The gene-control layer determines the neighboring spare module in the structural layer to replace the faulty module without collision. We verified the proposed mechanism by implementing the system with a field-programmable gate array with the application of a digital clock whose status can be monitored with light-emitting-diodes. In comparison with existing methods, the proposed architecture and mechanism are efficient enough for application with real fault-tolerant systems dealing with harsh and remote environments, such as outer space or deep sea.

Self-repairing digital systems have received increasing attention as modern systems are getting more complex and fast. Currently available self-repairing architectures have, however, some limitations such as storage overhead required to prepare all possible rewiring strategies and temporal incorrectness caused by elongated repairing time. In this paper, we propose a novel self-repairing architecture for fast fault recovery with an efficient use of limited resources, which can be easily applied to real complex digital systems. The proposed architecture consists of three layers: a working layer, a control layer, and an interface layer. The working layer employs a hybrid scheme of using both redundant and empty cells with a newly devised self-test. This relieves the overhead of redundant cells required to be prepared in advance by considering every possible fault situation. In the control layer, an ordered assignment control is proposed. The order of working-priority of each processor that controls a normal cell in the working layer is predetermined. A faulty processor is detected by a majority decision among neighboring control processors and corrected by rearranging the order of working-priority. The interface layer connects an external PC for reprogramming. Through this fault recovery mechanism, the system can keep normal functioning under noisy environments. We implemented the proposed self-repairing architecture using an field-programmable gate array board with an application of a dot-matrix LED display and verified its robust operation. The proposed architecture can be widely used as a new platform for self-repairing systems.

MOTIVATION: Gene regulatory networks (GRNs) govern cellular differentiation processes and enable construction of multicellular organisms from single cells. Although such networks are complex, there must be evolutionary design principles that shape the network to its present form, gaining complexity from simple modules. RESULTS: To isolate particular design principles, we have computationally evolved random regulatory networks with a preference to result either in hysteresis (switching threshold depending on current state), or in multistationarity (having multiple steady states), two commonly observed dynamical features of GRNs related to differentiation processes. We have analyzed the resulting evolved networks and compared their structures and characteristics with real GRNs reported from experiments. CONCLUSION: We found that the artificially evolved networks have particular topologies and it was notable that these topologies share important features and similarities with the real GRNs, particularly in contrasting properties of positive and negative feedback loops. We conclude that the structures of real GRNs are consistent with selection to favor one or other of the dynamical features of multistationarity or hysteresis. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

In this paper, we describe the "soft remote control system" that provides easy control of home installed appliances to improve the inhabitant's comfort. This system is feasible to control various home appliances naturally by recognizing the user's hand gestures in the "intelligent sweet home". The user selects the device that he/her wants to control by pointing it with his/her hand. Then, the user can command operation of the desired function via a predefined hand gesture. The approach complements the inconvenience of the conventional remote controllers giving additional freedom to persons with movement deficits and people without disabilities.