Full-Scale Aggregated MobileUNet: An Improved U-Net Architecture for SAR Oil Spill Detection.

Oil spills are a major threat to marine and coastal environments. Their unique radar backscatter intensity can be captured by synthetic aperture radar (SAR), resulting in dark regions in the images. However, many marine phenomena can lead to erroneous detections of oil spills. In addition, SAR images of the ocean include multiple targets, such as sea surface, land, ships, and oil spills and their look-alikes. The training of a multi-category classifier will encounter significant challenges due to the inherent class imbalance. Addressing this issue requires extracting target features more effectively. In this study, a lightweight U-Net-based model, Full-Scale Aggregated MobileUNet (FA-MobileUNet), was proposed to improve the detection performance for oil spills using SAR images. First, a lightweight MobileNetv3 model was used as the backbone of the U-Net encoder for feature extraction. Next, atrous spatial pyramid pooling (ASPP) and a convolutional block attention module (CBAM) were used to improve the capacity of the network to extract multi-scale features and to increase the speed of module calculation. Finally, full-scale features from the encoder were aggregated to enhance the network's competence in extracting features. The proposed modified network enhanced the extraction and integration of features at different scales to improve the accuracy of detecting diverse marine targets. The experimental results showed that the mean intersection over union (mIoU) of the proposed model reached more than 80% for the detection of five types of marine targets including sea surface, land, ships, and oil spills and their look-alikes. In addition, the IoU of the proposed model reached 75.85 and 72.67% for oil spill and look-alike detection, which was 18.94% and 25.55% higher than that of the original U-Net model, respectively. Compared with other segmentation models, the proposed network can more accurately classify the black regions in SAR images into oil spills and their look-alikes. Furthermore, the detection performance and computational efficiency of the proposed model were also validated against other semantic segmentation models.

Inter-Multilevel Super-Orthogonal Space-Time Coding Scheme for Reliable ZigBee-Based IoMT Communications.

The Internet of Things (IoT) technology has revolutionized the healthcare industry by enabling a new paradigm for healthcare delivery. This paradigm is known as the Internet of Medical Things (IoMT). IoMT devices are typically connected via a wide range of wireless communication technologies, such as Bluetooth, radio-frequency identification (RFID), ZigBee, Wi-Fi, and cellular networks. The ZigBee protocol is considered to be an ideal protocol for IoMT communication due to its low cost, low power usage, easy implementation, and appropriate level of security. However, maintaining ZigBee's high reliability is a major challenge due to multi-path fading and interference from coexisting wireless networks. This has increased the demand for more efficient channel coding schemes that can achieve a more reliable transmission of vital patient data for ZigBee-based IoMT communications. To meet this demand, a novel coding scheme called inter-multilevel super-orthogonal space-time coding (IM-SOSTC) can be implemented by combining the multilevel coding and set partitioning of super-orthogonal space-time block codes based on the coding gain distance (CGD) criterion. The proposed IM-SOSTC utilizes a technique that provides inter-level dependency between adjacent multilevel coded blocks to facilitate high spectral efficiency, which has been compromised previously by the high coding gain due to the multilevel outer code. In this paper, the performance of IM-SOSTC is compared to other related schemes via a computer simulation that utilizes the quasi-static Rayleigh fading channel. The simulation results show that IM-SOSTC outperforms other related coding schemes and is capable of providing the optimal trade-off between coding gain and spectral efficiency whilst guaranteeing full diversity and low complexity.

Don't Ditch the Laptop Just Yet: A Direct Replication of Mueller and Oppenheimer's (2014) Study 1 Plus Mini Meta-Analyses Across Similar Studies.

In this direct replication of Mueller and Oppenheimer's (2014) Study 1, participants watched a lecture while taking notes with a laptop (n = 74) or longhand (n = 68). After a brief distraction and without the opportunity to study, they took a quiz. As in the original study, laptop participants took notes containing more words spoken verbatim by the lecturer and more words overall than did longhand participants. However, laptop participants did not perform better than longhand participants on the quiz. Exploratory meta-analyses of eight similar studies echoed this pattern. In addition, in both the original study and our replication, higher word count was associated with better quiz performance, and higher verbatim overlap was associated with worse quiz performance, but the latter finding was not robust in our replication. Overall, results do not support the idea that longhand note taking improves immediate learning via better encoding of information.

AspSnFR: A genetically encoded biosensor for real-time monitoring of aspartate in live cells.

Aspartate is crucial for nucleotide synthesis, ammonia detoxification, and maintaining redox balance via the malate-aspartate-shuttle (MAS). To disentangle these multiple roles of aspartate metabolism, tools are required that measure aspartate concentrations in real time and in live cells. We introduce AspSnFR, a genetically encoded green fluorescent biosensor for intracellular aspartate, engineered through displaying and screening biosensor libraries on mammalian cells. In live cells, AspSnFR is able to precisely and quantitatively measure cytosolic aspartate concentrations and dissect its production from glutamine. Combining high-content imaging of AspSnFR with pharmacological perturbations exposes differences in metabolic vulnerabilities of aspartate levels based on nutrient availability. Further, AspSnFR facilitates tracking of aspartate export from mitochondria through SLC25A12, the MAS' key transporter. We show that SLC25A12 is a rapidly responding and direct route to couple Ca2+ signaling with mitochondrial aspartate export. This establishes SLC25A12 as a crucial link between cellular signaling, mitochondrial respiration, and metabolism.

Correction: Alkhaleefah et al. Connected-SegNets: A Deep Learning Model for Breast Tumor Segmentation from X-ray Images. Cancers 2022, 14, 4030.

In the original publication [...].

Connected-SegNets: A Deep Learning Model for Breast Tumor Segmentation from X-ray Images.

Inspired by Connected-UNets, this study proposes a deep learning model, called Connected-SegNets, for breast tumor segmentation from X-ray images. In the proposed model, two SegNet architectures are connected with skip connections between their layers. Moreover, the cross-entropy loss function of the original SegNet has been replaced by the intersection over union (IoU) loss function in order to make the proposed model more robust against noise during the training process. As part of data preprocessing, a histogram equalization technique, called contrast limit adapt histogram equalization (CLAHE), is applied to all datasets to enhance the compressed regions and smooth the distribution of the pixels. Additionally, two image augmentation methods, namely rotation and flipping, are used to increase the amount of training data and to prevent overfitting. The proposed model has been evaluated on two publicly available datasets, specifically INbreast and the curated breast imaging subset of digital database for screening mammography (CBIS-DDSM). The proposed model has also been evaluated using a private dataset obtained from Cheng Hsin General Hospital in Taiwan. The experimental results show that the proposed Connected-SegNets model outperforms the state-of-the-art methods in terms of Dice score and IoU score. The proposed Connected-SegNets produces a maximum Dice score of 96.34% on the INbreast dataset, 92.86% on the CBIS-DDSM dataset, and 92.25% on the private dataset. Furthermore, the experimental results show that the proposed model achieves the highest IoU score of 91.21%, 87.34%, and 83.71% on INbreast, CBIS-DDSM, and the private dataset, respectively.

High-resolution chemical dissection of a model eukaryote reveals targets, pathways and gene functions.

Due to evolutionary conservation of biology, experimental knowledge captured from genetic studies in eukaryotic model organisms provides insight into human cellular pathways and ultimately physiology. Yeast chemogenomic profiling is a powerful approach for annotating cellular responses to small molecules. Using an optimized platform, we provide the relative sensitivities of the heterozygous and homozygous deletion collections for nearly 1800 biologically active compounds. The data quality enables unique insights into pathways that are sensitive and resistant to a given perturbation, as demonstrated with both known and novel compounds. We present examples of novel compounds that inhibit the therapeutically relevant fatty acid synthase and desaturase (Fas1p and Ole1p), and demonstrate how the individual profiles facilitate hypothesis-driven experiments to delineate compound mechanism of action. Importantly, the scale and diversity of tested compounds yields a dataset where the number of modulated pathways approaches saturation. This resource can be used to map novel biological connections, and also identify functions for unannotated genes. We validated hypotheses generated by global two-way hierarchical clustering of profiles for (i) novel compounds with a similar mechanism of action acting upon microtubules or vacuolar ATPases, and (ii) an un-annotated ORF, YIL060w, that plays a role in respiration in the mitochondria. Finally, we identify and characterize background mutations in the widely used yeast deletion collection which should improve the interpretation of past and future screens throughout the community. This comprehensive resource of cellular responses enables the expansion of our understanding of eukaryotic pathway biology.

Evidence for a functionally relevant rocaglamide binding site on the eIF4A-RNA complex.

Translation initiation is an emerging target in oncology and neurobiology indications. Naturally derived and synthetic rocaglamide scaffolds have been used to interrogate this pathway; however, there is uncertainty regarding their precise mechanism(s) of action. We exploited the genetic tractability of yeast to define the primary effect of both a natural and a synthetic rocaglamide in a cellular context and characterized the molecular target using biochemical studies and in silico modeling. Chemogenomic profiling and mutagenesis in yeast identified the eIF (eukaryotic Initiation Factor) 4A helicase homologue as the primary molecular target of rocaglamides and defined a discrete set of residues near the RNA binding motif that confer resistance to both compounds. Three of the eIF4A mutations were characterized regarding their functional consequences on activity and response to rocaglamide inhibition. These data support a model whereby rocaglamides stabilize an eIF4A-RNA interaction to either alter the level and/or impair the activity of the eIF4F complex. Furthermore, in silico modeling supports the annotation of a binding pocket delineated by the RNA substrate and the residues identified from our mutagenesis screen. As expected from the high degree of conservation of the eukaryotic translation pathway, these observations are consistent with previous observations in mammalian model systems. Importantly, we demonstrate that the chemically distinct silvestrol and synthetic rocaglamides share a common mechanism of action, which will be critical for optimization of physiologically stable derivatives. Finally, these data confirm the value of the rocaglamide scaffold for exploring the impact of translational modulation on disease.