Artificial intelligence-driven radiomics study in cancer:the role of feature engineering and modeling

Modern medicine is reliant on various medical imaging technologies for non-invasively observing patients’anatomy.However,the interpretation of medical images can be highly subjective and dependent on the expertise of clinicians.Moreover,some potentially useful quantitative information in medical images,especially that which is not visible to the naked eye,is often ignored during clinical practice.In contrast,radiomics performs high-throughput feature extraction from medical images,which enables quantitative analysis of medical images and prediction of various clinical endpoints.Studies have reported that radiomics exhibits promising performance in diagnosis and predicting treatment responses and prognosis,demonstrating its potential to be a non-invasive auxiliary tool for personalized medicine.However,radiomics remains in a developmental phase as numerous technical challenges have yet to be solved,especially in feature engineering and statistical modeling.In this review,we introduce the current utility of radiomics by summarizing research on its application in the diagnosis,prognosis,and prediction of treatment responses in patients with cancer.We focus on machine learning approaches,for feature extraction and selection during feature engineering and for imbalanced datasets and multi-modality fusion during statistical modeling.Furthermore,we introduce the stability,reproducibility,and interpretability of features,and the generalizability and interpretability of models.Finally,we offer possible solutions to current challenges in radiomics research.

Liver transplantation as an alternative for the treatment of neuroendocrine liver metastasis: Appraisal of the current evidence

Background:Liver transplantation(LT)for neuroendocrine liver metastases(NELM)is still in debate.Studies comparing LT with liver resection(LR)for NELM are scarce,as patient selection is heterogeneous and experience is limited.The goal of this review was to provide a critical analysis of the evidence on LT versus LR in the treatment of NELM.Data sources:A scoping literature search on LT and LR for NELM was performed with PubMed,including English articles up to March 2023.Results:International guidelines recommend LR for NELM in resectable,well-differentiated tumors in the absence of extrahepatic metastatic disease with superior results of LR compared to systemic or liver-directed therapies.Advanced liver surgery has extended resectability criteria whilst entailing increased perioperative risk and short disease-free survival.In highly selected patients(based on the Milan criteria)with unresectable NELM,oncologic results of LT are promising.Prognostic factors include tumor biology(G1/G2)and burden,waiting time for LT,patient age and extrahepatic spread.Based on low-level evi-dence,LT for low-grade NELM within the Milan criteria resulted in improved disease-free survival and overall survival compared to LR.The benefits of LT were lost in patients beyond the Milan NELM-criteria.Conclusions:With adherence to strict selection criteria especially tumor biology,LT for NELM is becoming a valuable option providing oncologic benefits compared to LR.Recent evidence suggests even stricter selection criteria with regard to tumor biology.

A phenology-based vegetation index for improving ratoon rice mapping using harmonized Landsat and Sentinel-2 data

Ratoon rice,which refers to a second harvest of rice obtained from the regenerated tillers originating from the stubble of the first harvested crop,plays an important role in both food security and agroecology while requiring minimal agricultural inputs.However,accurately identifying ratoon rice crops is challenging due to the similarity of its spectral features with other rice cropping systems(e.g.,double rice).Moreover,images with a high spatiotemporal resolution are essential since ratoon rice is generally cultivated in fragmented croplands within regions that frequently exhibit cloudy and rainy weather.In this study,taking Qichun County in Hubei Province,China as an example,we developed a new phenology-based ratoon rice vegetation index(PRVI)for the purpose of ratoon rice mapping at a 30 m spatial resolution using a robust time series generated from Harmonized Landsat and Sentinel-2(HLS)images.The PRVI that incorporated the red,near-infrared,and shortwave infrared 1 bands was developed based on the analysis of spectro-phenological separability and feature selection.Based on actual field samples,the performance of the PRVI for ratoon rice mapping was carefully evaluated by comparing it to several vegetation indices,including normalized difference vegetation index(NDVI),enhanced vegetation index(EVI)and land surface water index(LSWI).The results suggested that the PRVI could sufficiently capture the specific characteristics of ratoon rice,leading to a favorable separability between ratoon rice and other land cover types.Furthermore,the PRVI showed the best performance for identifying ratoon rice in the phenological phases characterized by grain filling and harvesting to tillering of the ratoon crop(GHS-TS2),indicating that only several images are required to obtain an accurate ratoon rice map.Finally,the PRVI performed better than NDVI,EVI,LSWI and their combination at the GHS-TS2 stages,with producer's accuracy and user's accuracy of 92.22 and 89.30%,respectively.These results demonstrate that the proposed PRVI based on HLS data can effectively identify ratoon rice in fragmented croplands at crucial phenological stages,which is promising for identifying the earliest timing of ratoon rice planting and can provide a fundamental dataset for crop management activities.

Monitoring Surface Deformation Using Distributed Scatterers InSAR

In the past two decades,extensive and in-depth research has been conducted on Time Series InSAR technology with the advancement of high-performance SAR satellites and the accumulation of big SAR data.The introduction of distributed scatterers in Distributed Scatterers InSAR(DS-InSAR)has significantly expanded the application scenarios of InSAR geodetic measurement by increasing the number of measurement points.This study traces the history of DS-InSAR,presents the definition and characteristics of distributed scatterers,and focuses on exploring the relationships and distinctions among proposed algorithms in two crucial steps:statistically homogeneous pixel selection and phase optimization.Additionally,the latest research progress in this field is tracked and the possible development direction in the future is discussed.Through simulation experiments and two real InSAR case studies,the proposed algorithms are compared and verified,and the advantages of DS-InSAR in deformation measurement practice are demonstrated.This work not only offers insights into current trends and focal points for theoretical research on DS-InSAR but also provides practical cases and guidance for applied research.

Bird species present in urban parks are more colorful than urban avoiders:A test in the Argentinian Pampas

Bird plumage color has been assessed as a possible trait driving the presence of bird species in urban areas.Although some species can see the ultraviolet(UV) spectrum,the mentioned studies did not take into account UV reflectance when characterizing bird plumage.This study aimed to use a recent database of the colorfulness in passerines that incorporated the UV spectrum to compare bird colorfulness and other traits between urban parks and rural areas in Central-East Argentina.Birds in urban parks were surveyed in 51 parks in 6 cities during breeding and non-breeding seasons.A list of Passeriformes species from parks was created,and a list of urban avoider species was created from the bibliography.Species traits were body mass,clutch size,migratory status,nesting site,diet and habitat breadth,and plumage colorfulness.A total of 85 species were detected in the regional pool,of which 30 species were detected in urban parks.Bird species present in urban parks were more colorful than bird species only present in rural areas.In addition,bird presence in urban parks was positively related to their regional frequency and diet breadth.Moreover,urban presence was related to nesting on trees and buildings,whereas species not present in urban parks nested on the ground.The results obtained showed that bird color is significantly associated with presence of bird species in urban parks.

Insights into genetic diversity and phenotypic variations in domestic geese through comprehensive population and pan-genome analysis

Background Domestic goose breeds are descended from either the Swan goose(Anser cygnoides)or the Greylag goose(Anser anser),exhibiting variations in body size,reproductive performance,egg production,feather color,and other phenotypic traits.Constructing a pan-genome facilitates a thorough identification of genetic variations,thereby deepening our comprehension of the molecular mechanisms underlying genetic diversity and phenotypic variability.Results To comprehensively facilitate population genomic and pan-genomic analyses in geese,we embarked on the task of 659 geese whole genome resequencing data and compiling a database of 155 RNA-seq samples.By constructing the pan-genome for geese,we generated non-reference contigs totaling 612 Mb,unveiling a collection of 2,813 novel genes and pinpointing 15,567 core genes,1,324 softcore genes,2,734 shell genes,and 878 cloud genes in goose genomes.Furthermore,we detected an 81.97 Mb genomic region showing signs of genome selection,encompassing the TGFBR2 gene correlated with variations in body weight among geese.Genome-wide association studies utilizing single nucleotide polymorphisms(SNPs)and presence-absence variation revealed significant genomic associations with various goose meat quality,reproductive,and body composition traits.For instance,a gene encoding the SVEP1 protein was linked to carcass oblique length,and a distinct gene-CDS haplotype of the SVEP1 gene exhibited an association with carcass oblique length.Notably,the pan-genome analysis revealed enrichment of variable genes in the“hair follicle maturation”Gene Ontology term,potentially linked to the selection of feather-related traits in geese.A gene presence-absence variation analysis suggested a reduced frequency of genes associated with“regulation of heart contraction”in domesticated geese compared to their wild counterparts.Our study provided novel insights into gene expression features and functions by integrating gene expression patterns across multiple organs and tissues in geese and analyzing population variation.Conclusion This accomplishment originates from the discernment of a multitude of selection signals and candidate genes associated with a wide array of traits,thereby markedly enhancing our understanding of the processes underlying domestication and breeding in geese.Moreover,assembling the pan-genome for geese has yielded a comprehensive apprehension of the goose genome,establishing it as an indispensable asset poised to offer innovative viewpoints and make substantial contributions to future geese breeding initiatives.

Inbreeding and genetic load in a pair of sibling grouse species:Tetrastes sewersowi and T.bonasia

Genetic load and inbreeding are recognized as important factors to be considered in conservation programs.Elevated levels of both can increase the risk of population extinction by negatively impacting fitness-related characters in many species of plants and animals,including humans(inbreeding depression).Genomic tech-niques are increasingly used in measuring and understanding genetic load and inbreeding and their importance in evolution and conservation.We used whole genome resequencing data from two sibling grouse species in subarctic Eurasia to quantify both.We found a large range of inbreeding measured as FROH(fraction of runs of homozygosity)in individuals from different populations of Chinese Grouse(Tetrastes sewerzowi)and Hazel Grouse(T.bonasia).FROH estimated from genome-wide runs of homozygosity(ROH)ranged from 0.02 to 0.24 among Chinese Grouse populations and from 0.01 to 0.44 in Hazel Grouse.Individuals from a population of Chinese Grouse residing in the Qilian mountains and from the European populations of Hazel Grouse(including samples from Sweden,Germany and Northeast Poland)were the most inbred(FROH ranged from 0.10 to 0.23 and 0.11 to 0.44,respectively).These levels are comparable to other highly inbred populations of birds.Hazel Grouse from northern China and Chinese Grouse residing in the Qinghai-Tibetan Plateau showed relatively lower inbreeding levels.Comparisons of the ratio between deleterious missense mutations and synonymous mutations revealed higher levels in Chinese Grouse as compared to Hazel Grouse.These results are possibly explained by higher fixation rates,mutational melt down,in the range-restricted Chinese Grouse compared to the wide-ranging Hazel Grouse.However,when we compared the relatively more severe class of loss-of-function muta-tions,Hazel Grouse had slightly higher levels than Chinese Grouse,a result which may indicate that purifying selection(purging)has been more efficient in Chinese Grouse on this class of mutations.

Evaluating the performance of genomic selection on purebred population by incorporating crossbred data in pigs

Genomic selection(GS)has been widely used in livestock,which greatly accelerated the genetic progress of complex traits.The population size was one of the significant factors affecting the prediction accuracy,while it was limited by the purebred population.Compared to directly combining two uncorrelated purebred populations to extend the reference population size,it might be more meaningful to incorporate the correlated crossbreds into reference population for genomic prediction.In this study,we simulated purebred offspring(PAS and PBS)and crossbred offspring(CAB)base on real genotype data of two base purebred populations(PA and PB),to evaluate the performance of genomic selection on purebred while incorporating crossbred information.The results showed that selecting key crossbred individuals via maximizing the expected genetic relationship(REL)was better than the other methods(individuals closet or farthest to the purebred population,CP/FP)in term of the prediction accuracy.Furthermore,the prediction accuracy of reference populations combining PA and CAB was significantly better only based on PA,which was similar to combine PA and PAS.Moreover,the rank correlation between the multiple of the increased relationship(MIR)and reliability improvement was 0.60-0.70.But for individuals with low correlation(Cor(Pi,PA or B),the reliability improvement was significantly lower than other individuals.Our findings suggested that incorporating crossbred into purebred population could improve the performance of genetic prediction compared with using the purebred population only.The genetic relationship between purebred and crossbred population is a key factor determining the increased reliability while incorporating crossbred population in the genomic prediction on pure bred individuals.

Experimental clearance rate and intraguild predation of jellyfish Cyanea nozakii

Cyanea nozakii,a common jellyfish distributed in offshore China,has a complex trophic relationship with other zooplankton groups.However,few studies have reported the predation rates and prey selection patterns of C.nozakii medusae on different prey items.Research is also lacking on the intraguild predation of Aurelia coerulea(another common bloom jellyfish in offshore China)by C.nozakii.To address the knowledge gaps,the clearance rates of C.nozakii for different prey items,including copepods(small<1000μm and large>1000μm),fish larvae,and gelatinous prey(hydromedusae,A.coerulea ephyrae,and chaetognaths),were measured.The influence of predator size on the clearance rate was also determined.Additionally,we examined the intraguild predation of C.nozakii on A.coerulea medusae.The clearance rates of C.nozakii varied widely with prey organisms,being independent of prey concentrations.Gelatinous organisms,except for chaetognaths,were captured with considerably high efficiency,followed by fish larvae and copepods,indicating the preferential prey selection of gelatinous organisms by C.nozakii.The clearance rate increased linearly with the cross-sectional area of C.nozakii.Body size in medusae may,to some extents,underpin their capacity to capture more prey by increasing the encounter rate and capture success through ontogeny.C.nozakii preyed voraciously on A.coerulea in high feeding efficiency,but the clearance rate decreased with increasing A.coerulea(as prey)size.This phenomenon of intraguild predation suggests a speculative hypothesis of potential population regulation of A.coerulea by C.nozakii.The information regarding the feeding ecology of C.nozakii reported in this study is important for understanding plankton dynamics in marine ecosystems with extensive occurrences of this jellyfish.

Analysis of environmental selection pressure of superoxide dismutase in deep-sea sea cucumber

Manganese superoxide dismutase(MnSOD)is an antioxidant that exists in mitochondria and can effectively remove superoxide anions in mitochondria.In a dark,high-pressure,and low-temperature deep-sea environment,MnSOD is essential for the survival of sea cucumbers.Six MnSODs were identified from the transcriptomes of deep and shallow-sea sea cucumbers.To explore their environmental adaptation mechanism,we conducted environmental selection pressure analysis through the branching site model of PAML software.We obtained night positive selection sites,and two of them were significant(97F→H,134K→V):97F→H located in a highly conservative characteristic sequence,and its polarity c hange might have a great impact on the function of MnSOD;134K→V had a change in piezophilic a bility,which might help MnSOD adapt to the environment of high hydrostatic pressure in the deepsea.To further study the effect of these two positive selection sites on MnSOD,we predicted the point mutations of F97H and K134V on shallow-sea sea cucumber by using MAESTROweb and PyMOL.Results show that 97F→H,134K→V might improve MnSOD’s efficiency of scavenging superoxide a nion and its ability to resist high hydrostatic pressure by moderately reducing its stability.The above results indicated that MnSODs of deep-sea sea cucumber adapted to deep-sea environments through their amino acid changes in polarity,piezophilic behavior,and local stability.This study revealed the correlation between MnSOD and extreme environment,and will help improve our understanding of the organism’s adaptation mechanisms in deep sea.

Prescreening of large-effect markers with multiple strategies improves the accuracy of genomic prediction

Presently,integrating multi-omics information into a prediction model has become a ameliorate strategy for genomic selection to improve genomic prediction accuracy.Here,we set the genomic and transcriptomic data as the training population data,using BSLMM,TWAS,and eQTL mapping to prescreen features according to |β_(b)|>0,top 1%of phenotypic variation explained(PVE),expression-associated single nucleotide polymorphisms(eSNPs),and egenes(false discovery rate(FDR)<0.01),where these loci were set as extra fixed effects(named GBLUP-Fix)and random effects(GFBLUP)to improve the prediction accuracy in the validation population,respectively.The results suggested that both GBLUP-Fix and GFBLUP models could improve the accuracy of longissimus dorsi muscle(LDM),water holding capacity(WHC),shear force(SF),and pH in Huaxi cattle on average from 2.14 to 8.69%,especially the improvement of GFBLUP-TWAS over GBLUP was 13.66%for SF.These methods also captured more genetic variance than GBLUP.Our study confirmed that multi-omics-assisted large-effects loci prescreening could improve the accuracyofgenomic prediction.

A Federated Learning Framework with Blockchain-Based Auditable Participant Selection

Federated learning is an important distributed model training technique in Internet of Things(IoT),in which participant selection is a key component that plays a role in improving training efficiency and model accuracy.This module enables a central server to select a subset of participants to performmodel training based on data and device information.By doing so,selected participants are rewarded and actively perform model training,while participants that are detrimental to training efficiency and model accuracy are excluded.However,in practice,participants may suspect that the central server may have miscalculated and thus not made the selection honestly.This lack of trustworthiness problem,which can demotivate participants,has received little attention.Another problem that has received little attention is the leakage of participants’private information during the selection process.We will therefore propose a federated learning framework with auditable participant selection.It supports smart contracts in selecting a set of suitable participants based on their training loss without compromising the privacy.Considering the possibility of malicious campaigning and impersonation of participants,the framework employs commitment schemes and zero-knowledge proofs to counteract these malicious behaviors.Finally,we analyze the security of the framework and conduct a series of experiments to demonstrate that the framework can effectively improve the efficiency of federated learning.

Rice Heat Tolerance Breeding: A Comprehensive Review and Forward Gaze

The yield potential of rice is seriously affected by heat stress due to climate change. Since rice is a staple food globally, it is imperative to develop heat-resistant rice varieties. Thus, a thorough understanding of the complex molecular mechanisms underlying heat tolerance and the impact of high temperatures on various critical stages of the crop is needed. Adoption of both conventional and innovative breeding strategies offers a long-term advantage over other methods, such as agronomic practices, to counter heat stress. In this review, we summarize the effects of heat stress, regulatory pathways for heat tolerance, phenotyping strategies, and various breeding methods available for developing heat-tolerant rice. We offer perspectives and knowledge to guide future research endeavors aimed at enhancing the ability of rice to withstand heat stress and ultimately benefit humanity.

Mapping of powdery mildew resistance genes transferred to common wheat from wild emmer wheat revealed three functional Pm60 haplotypes

Powdery mildew(PM),caused by Blumeria graminis f.sp.tritici(Bgt),is one of the destructive wheat diseases worldwide.Wild emmer wheat(Triticum turgidum ssp.dicoccoides,WEW),a tetraploid progenitor of common wheat,is a valuable genetic resource for wheat disease resistance breeding programs.We developed three hexaploid pre-breeding lines with PM resistance genes derived from three WEW accessions.These resistant pre-breeding lines were crossed with susceptible common wheat accessions.Segregations in the F2populations were 3 resistant:1 susceptible,suggesting a single dominant allele in each resistant parent.Mapping of the resistance gene in each line indicated a single locus on the long arm of chromosome 7A,at the approximate location of previously cloned Pm60 from T.urartu.Sanger sequencing revealed three different Pm60 haplotypes(Hap 3,Hap 5,and Hap 6).Co-segregating diagnostic markers were developed for identification and selection of each haplotype.The resistance function of each haplotype was verified by the virus-induced gene silencing(VIGS).Common wheat lines carrying each of these Pm60 haplotypes were resistant to most Bgt isolates and differences in the response arrays suggested allelic variation in response.

Integrating artificial intelligence and high-throughput phenotyping for crop improvement

Crop improvement is crucial for addressing the global challenges of food security and sustainable agriculture.Recent advancements in high-throughput phenotyping(HTP)technologies and artificial intelligence(AI)have revolutionized the field,enabling rapid and accurate assessment of crop traits on a large scale.The integration of AI and machine learning algorithms with HTP data has unlocked new opportunities for crop improvement.AI algorithms can analyze and interpret large datasets,and extract meaningful patterns and correlations between phenotypic traits and genetic factors.These technologies have the potential to revolutionize plant breeding programs by providing breeders with efficient and accurate tools for trait selection,thereby reducing the time and cost required for variety development.However,further research and collaboration are needed to overcome the existing challenges and fully unlock the power of HTP and AI in crop improvement.By leveraging AI algorithms,researchers can efficiently analyze phenotypic data,uncover complex patterns,and establish predictive models that enable precise trait selection and crop breeding.The aim of this review is to explore the transformative potential of integrating HTP and AI in crop improvement.This review will encompass an in-depth analysis of recent advances and applications,highlighting the numerous benefits and challenges associated with HTP and AI.

Hyperspectral Image Based Interpretable Feature Clustering Algorithm

Hyperspectral imagery encompasses spectral and spatial dimensions,reflecting the material properties of objects.Its application proves crucial in search and rescue,concealed target identification,and crop growth analysis.Clustering is an important method of hyperspectral analysis.The vast data volume of hyperspectral imagery,coupled with redundant information,poses significant challenges in swiftly and accurately extracting features for subsequent analysis.The current hyperspectral feature clustering methods,which are mostly studied from space or spectrum,do not have strong interpretability,resulting in poor comprehensibility of the algorithm.So,this research introduces a feature clustering algorithm for hyperspectral imagery from an interpretability perspective.It commences with a simulated perception process,proposing an interpretable band selection algorithm to reduce data dimensions.Following this,amulti-dimensional clustering algorithm,rooted in fuzzy and kernel clustering,is developed to highlight intra-class similarities and inter-class differences.An optimized P systemis then introduced to enhance computational efficiency.This system coordinates all cells within a mapping space to compute optimal cluster centers,facilitating parallel computation.This approach diminishes sensitivity to initial cluster centers and augments global search capabilities,thus preventing entrapment in local minima and enhancing clustering performance.Experiments conducted on 300 datasets,comprising both real and simulated data.The results show that the average accuracy(ACC)of the proposed algorithm is 0.86 and the combination measure(CM)is 0.81.

Non-Targeted Metabolomics Reveals the Metabolic Alterations in Response to Artificial Selective Breeding in the Fast-Growing Strains of Pacific Oyster

Pacific oyster(Crassostrea gigas)is one of the most important mollusks cultured all around the world.Selective breeding programs of Pacific oysters in China is initiated since 2006 and developed the genetically improved strain with fast-growing trait.However,little is known about the metabolic signatures of the fast-growing trait.In the present study,the non-targeted metabolomics was performed to analyze the metabolic signatures of adductor muscle tissue in one-year old Pacific oysters from fast-growing strain and the wild population.A total of 7767 and 10174 valid peaks were extracted and quantified in ESI^(+)and ESI^(−)modes,resulting in 399 and 381 annotated metabolites,respectively.PCA and OPLS-DA revealed that considerable separation among samples from fastgrowing strain and wild population,suggesting the differences in metabolic signatures.Meanwhile,81 significantly different metabolites(SDMs)were identified in the comparisons between fast-growing strain and wild population,based on the strict thresholds.It was found that there were highly correlation and conserved coordination among these SDMs.KEGG enrichment analysis indicated that the SDMs were tightly related to pantothenate and CoA biosynthesis,steroid hormone biosynthesis,riboflavin metabolism,and arginine and proline metabolism.Of them,the CoA biosynthesis and metabolism,affected by pantetheine and pantothenic acid,might be important for the growth of Pacific oysters under artificial selective breeding.The study provides the comprehensive views of metabolic signatures in response to artificially selective breeding,and is helpful to better understand the molecular mechanism of fastgrowing traits in Pacific oysters.

The suitability assessment on site selection for bottom seeding scallop culture based on analytic hierarchy process

Scallop culture is an important way of bottom-seeding marine ranching,which is of great significance to improve the current situation of fishery resources.However,there are some problems in site-selection evaluation of marine ranching,such as imperfect criteria system,complex structure,untargeted criteria quantification,etc.In addition,no site-selection evaluation method of bottom-seeding culture areas for scallops is available.Therefore,we established a hierarchy structure model according to the analytic hierarchy process(AHP)theory,in which social,physical,chemical,and biological environments are used as main criteria,and marine functional zonation,water depth,current,water temperature,salinity,substrate type,water quality,sediment quality,red tide,phytoplankton,and zooplankton are used as sub-criteria,on which a multi-parameter evaluation system is set up.Meanwhile,the dualism method,assignment method,and membership function method were used to quantify sub-criteria,and a quantitative evaluation for the entire criteria was added,including the evaluation and analysis of two types of unsuitable environmental situations.By overall consideration in scallop yield,quality,and marine ranching construction objectives,the weight of the main criteria could be determined.Five grades in the suitability corresponding to the evaluation result were divided,and the Python language was used to create an evaluation system for efficient calculation and intuitive presentation of the evaluation outcome.Eight marine cases were simulated based on existing survey data,and the results prove that the method is feasible for evaluating and analyzing the site selection of bottom-seeding culture areas for scallops under various environmental situations.The proposed evaluation method can be promoted for the site selection of bottom-seeding marine ranching.This study provided theoretical and methodological references for the site selection evaluation of other types of marine ranching.

A Support Data-Based Core-Set Selection Method for Signal Recognition

In recent years,deep learning-based signal recognition technology has gained attention and emerged as an important approach for safeguarding the electromagnetic environment.However,training deep learning-based classifiers on large signal datasets with redundant samples requires significant memory and high costs.This paper proposes a support databased core-set selection method(SD)for signal recognition,aiming to screen a representative subset that approximates the large signal dataset.Specifically,this subset can be identified by employing the labeled information during the early stages of model training,as some training samples are labeled as supporting data frequently.This support data is crucial for model training and can be found using a border sample selector.Simulation results demonstrate that the SD method minimizes the impact on model recognition performance while reducing the dataset size,and outperforms five other state-of-the-art core-set selection methods when the fraction of training sample kept is less than or equal to 0.3 on the RML2016.04C dataset or 0.5 on the RML22 dataset.The SD method is particularly helpful for signal recognition tasks with limited memory and computing resources.

Kentish Plover(Charadrius alexandrinus) and Little Tern(Sternula albifrons) prefer shells for nesting:A field experiment

Shorebird populations are declining worldwide,mainly due to human disturbances and loss of coastal wetlands.However,supratidal habitats as saltpans could play a role in buffering human impact.Saltpans have shown to be important as feeding or breeding sites of some shorebird species.A potential conservation strategy to increase shorebird populations in saltpans is to manipulate the cues that birds use to select optimal breeding habitat.Here it is hypothesized that shorebirds are attracted to bivalve shells due to the advantages they offer.Following this hypothesis,we supplemented a restored saltpan in 2019 and 2021 with bivalve shells,expecting an increase in the number of breeding birds’ nests.More than 75% of Kentish Plover(Charadrius alexandrinus) and Little Tern(Sternula albifrons) nests were found in patches with shells in both years.The best model for both species indicates that the presence of shells is the factor that most correlates with the location of nests.The probability of choosing one place over another to settle their nest increases in areas with an abundance of shells,double in the case of the Kentish Plover and triple in the case of the Little Tern.The result of this study may constitute a valuable tool for attracting birds to restored saltpans and could contribute to the success of expensive restoration projects where time is usually a constraint.