High-Speed Clearing and High-Resolution Staining for Analysis of Various Markers for Neurons and Vessels.

BACKGROUND: Tissue clearing enables deep imaging in various tissues by increasing the transparency of tissues, but there were limitations of immunostaining of the large-volume tissues such as the whole brain. METHODS: Here, we cleared and immune-stained whole mouse brain tissues using a novel clearing technique termed high-speed clearing and high-resolution staining (HCHS). We observed neural structures within the cleared brains using both a confocal microscope and a light-sheet fluorescence microscope (LSFM). The reconstructed 3D images were analyzed using a computational reconstruction algorithm. RESULTS: Various neural structures were well observed in three-dimensional (3D) images of the cleared brains from Gad-green fluorescent protein (GFP) mice and Thy 1-yellow fluorescent protein (YFP) mice. The intrinsic fluorescence signals of both transgenic mice were preserved after HCHS. In addition, large-scale 3D imaging of brains, immune-stained by the HCHS method using a mild detergent-based solution, allowed for the global topological analysis of several neuronal markers such as c-Fos, neuronal nuclear protein (NeuN), Microtubule-associated protein 2 (Map2), Tuj1, glial fibrillary acidic protein (GFAP), and tyrosine hydroxylase (TH) in various anatomical regions in the whole mouse brain tissues. Finally, through comparisons with various existing tissue clearing methodologies such as CUBIC, Visikol, and 3DISCO, it was confirmed that the HCHS methodology results in relatively less tissue deformation and higher fluorescence retention. CONCLUSION: In conclusion, the development of 3D imaging based on novel tissue-clearing techniques (HCHS) will enable detailed spatial analysis of neural and vascular networks present within the brain.

Artificial intelligence-based prescription of personalized scalp cosmetics improved the scalp condition: efficacy results from 100 participants.

Background: Scalp-related symptoms such as dandruff and itching are common with diverse underlying etiologies. We previously proposed a novel classification and scoring system for scalp conditions, called the scalp photographic index (SPI); it grades five scalp features using trichoscopic images with good reliability. However, it requires trained evaluators.Aim: To develop artificial intelligence (AI) algorithms for assessment of scalp conditions and to assess the feasibility of AI-based recommendations on personalized scalp cosmetics.Methods: Using EfficientNet, convolutional neural network (CNN) models (SPI-AI) ofeach scalp feature were established. 101,027 magnified scalp images graded according to the SPI scoring were used for training, validation, and testing the model Adults with scalp discomfort were prescribed shampoos and scalp serums personalized according to their SPI-AI-defined scalp types. Using the SPI, the scalp conditions were evaluated at baseline and at weeks 4, 8, and 12 of treatment.Results: The accuracies of the SPI-AI for dryness, oiliness, erythema, folliculitis, and dandruff were 91.3%, 90.5%, 89.6%, 87.3%, and 95.2%, respectively. Overall, 100 individuals completed the 4-week study; 43 of these participated in an extension study until week 12. The total SPI score decreased from 32.70 ± 7.40 at baseline to 15.97 ± 4.68 at week 4 (p < 0.001). The efficacy was maintained throughout 12 weeks.Conclusions: SPI-AI accurately assessed the scalp condition. AI-based prescription of tailored scalp cosmetics could significantly improve scalp health.

De novo design of protein minibinder agonists of TLR3.

Toll-like Receptor 3 (TLR3) is a pattern recognition receptor that initiates antiviral immune responses upon binding double-stranded RNA (dsRNA). Several nucleic acid-based TLR3 agonists have been explored clinically as vaccine adjuvants in cancer and infectious disease, but present substantial manufacturing and formulation challenges. Here, we use computational protein design to create novel miniproteins that bind to human TLR3 with nanomolar affinities. Cryo-EM structures of two minibinders in complex with TLR3 reveal that they bind the target as designed, although one partially unfolds due to steric competition with a nearby N-linked glycan. Multimeric forms of both minibinders induce NF-κB signaling in TLR3-expressing cell lines, demonstrating that they may have therapeutically relevant biological activity. Our work provides a foundation for the development of specific, stable, and easy-to-formulate protein-based agonists of TLRs and other pattern recognition receptors.

Senolytic therapeutics: An emerging treatment modality for osteoarthritis.

Osteoarthritis (OA), a chronic joint disease affecting millions of people aged over 65 years, is the main musculoskeletal cause of diminished joint mobility in the elderly. It is characterized by lingering pain and increasing deterioration of articular cartilage. Aging and accumulation of senescent cells (SCs) in the joints are frequently associated with OA. Apoptosis resistance; irreversible cell cycle arrest; increased p16INK4a expression, secretion of senescence-associated secretory phenotype factors, senescence-associated ß-galactosidase levels, secretion of extracellular vesicles, and levels of reactive oxygen and reactive nitrogen species; and mitochondrial dysregulation are some common changes in cellular senescence in joint tissues. Development of OA correlates with an increase in the density of SCs in joint tissues. Senescence-associated secretory phenotype has been linked to OA and cartilage breakdown. Senolytics and therapeutic pharmaceuticals are being focused upon for OA management. SCs can be selectively eliminated or killed by senolytics to halt the pathogenesis and progression of OA. Comprehensive understanding of how aging affects joint dysfunction will benefit OA patients. Here, we discuss age-related mechanisms associated with OA pathogenesis and senolytics as an emerging modality in the management of age-related SCs and pathogenesis of OA in preclinical and clinical studies.

Case Report of Efficacy of Skin Perfusion After Hyperbaric Oxygen Therapy Following Peripheral Tissue Injury due to Usage of Inotropes and Vasopressors.

Hyperbaric Oxygen Therapy (HBOT) has garnered significant attention as a therapeutic principle with potential benefits across a variety spectrum of medical conditions, ranging from wound healing and ischemic conditions to neurologic disorders and radiation-induced tissue damage. HBOT involves the administration of 100% oxygen at higher atmospheric pressures, leading to increased oxygen dissolved in bodily fluids and tissues. The elevated oxygen levels are proposed to facilitate tissue repair, reduce inflammation, and promote angiogenesis. This case report presents a compelling instance of the usefulness of HBOT in promoting skin perfusion and healing following peripheral tissue injury resulting from the administration of inotropic and vasopressor agents in septic shock patients.

SARS-CoV-2 infection elucidates unique features of pregnancy-specific immunity.

Pregnancy is a risk factor for increased severity of SARS-CoV-2 and other respiratory infections. The mechanisms underlying this risk have not been well-established, partly due to a limited understanding of how pregnancy shapes immune responses. To gain insight into the role of pregnancy in modulating immune responses at steady state and upon perturbation, we collected peripheral blood mononuclear cells (PBMC), plasma, and stool from 226 women, including 152 pregnant individuals (n = 96 with SARS-CoV-2 infection and n = 56 healthy controls) and 74 non-pregnant women (n = 55 with SARS-CoV-2 and n = 19 healthy controls). We found that SARS-CoV-2 infection was associated with altered T cell responses in pregnant compared to non-pregnant women. Differences included a lower percentage of memory T cells, a distinct clonal expansion of CD4-expressing CD8 + T cells, and the enhanced expression of T cell exhaustion markers, such as programmed cell death-1 (PD-1) and T cell immunoglobulin and mucin domain-3 (Tim-3), in pregnant women. We identified additional evidence of immune dysfunction in severely and critically ill pregnant women, including a lack of expected elevation in regulatory T cell (Treg) levels, diminished interferon responses, and profound suppression of monocyte function. Consistent with earlier data, we found maternal obesity was also associated with altered immune responses to SARS-CoV-2 infection, including enhanced production of inflammatory cytokines by T cells. Certain gut bacterial species were altered in pregnancy and upon SARS-CoV-2 infection in pregnant individuals compared to non-pregnant women. Shifts in cytokine and chemokine levels were also identified in the sera of pregnant individuals, most notably a robust increase of interleukin-27 (IL-27), a cytokine known to drive T cell exhaustion, in the pregnant uninfected control group compared to all non-pregnant groups. IL-27 levels were also significantly higher in uninfected pregnant controls compared to pregnant SARS-CoV-2-infected individuals. Using two different preclinical mouse models of inflammation-induced fetal demise and respiratory influenza viral infection, we found that enhanced IL-27 protects developing fetuses from maternal inflammation but renders adult female mice vulnerable to viral infection. These combined findings from human and murine studies reveal nuanced pregnancy-associated immune responses, suggesting mechanisms underlying the increased susceptibility of pregnant individuals to viral respiratory infections.

Design of 2-Pyridone Fluorophores for Brighter Emissions at Longer Wavelengths.

The recent discovery of blue fluorophores with high quantum yields based on pyridone structures inspired the development of new low-molecular-weight fluorophores with bright emissions at tunable wavelengths, which are highly attractive for various applications. In this study, we propose a rational design strategy for 2-pyridone-based fluorophores with bright emissions at long wavelengths. With a detailed understanding of the positional substitution effects on each carbon atom of the 2-pyridone core, we developed a bright blue fluorophore (λabs =377 nm; λem =433 nm; ϵ=13,200 M-1 cm-1 ; ϕF =88 %) through C3 -aryl and C4 -ester substitutions followed by cyclization. Furthermore, by applying the intramolecular charge transfer (ICT) principle, we invented a bright green fluorophore through C3 - and C4 -diester and C6 -aryl substitutions. The ICT fluorophore based on the pyridone structure shows large molar absorptivity (ϵ=20,100 M-1 cm-1 ), longer emission wavelength (λem =539 nm), high emission quantum yield (ϕF =74 %), and large Stokes shift (Δv=5720 cm-1 ), which are comparable to those of practical fluorescent probes.

A syndromic neurodevelopmental disorder caused by rare variants in PPFIA3.

PPFIA3 encodes the protein-tyrosine phosphatase, receptor-type, F-polypeptide-interacting-protein-alpha-3 (PPFIA3), which is a member of the LAR-protein-tyrosine phosphatase-interacting-protein (liprin) family involved in synapse formation and function, synaptic vesicle transport, and presynaptic active zone assembly. The protein structure and function are evolutionarily well conserved, but human diseases related to PPFIA3 dysfunction are not yet reported in OMIM. Here, we report 20 individuals with rare PPFIA3 variants (19 heterozygous and 1 compound heterozygous) presenting with developmental delay, intellectual disability, hypotonia, dysmorphisms, microcephaly or macrocephaly, autistic features, and epilepsy with reduced penetrance. Seventeen unique PPFIA3 variants were detected in 18 families. To determine the pathogenicity of PPFIA3 variants in vivo, we generated transgenic fruit flies producing either human wild-type (WT) PPFIA3 or five missense variants using GAL4-UAS targeted gene expression systems. In the fly overexpression assays, we found that the PPFIA3 variants in the region encoding the N-terminal coiled-coil domain exhibited stronger phenotypes compared to those affecting the C-terminal region. In the loss-of-function fly assay, we show that the homozygous loss of fly Liprin-α leads to embryonic lethality. This lethality is partially rescued by the expression of human PPFIA3 WT, suggesting human PPFIA3 function is partially conserved in the fly. However, two of the tested variants failed to rescue the lethality at the larval stage and one variant failed to rescue lethality at the adult stage. Altogether, the human and fruit fly data reveal that the rare PPFIA3 variants are dominant-negative loss-of-function alleles that perturb multiple developmental processes and synapse formation.

Effects of a Free Electronic Book on Systemic Anatomy with Schematics on Student Learning / Efectos de un Libro Electrónico Gratuito sobre Anatomía Sistémica con Esquemas sobre el Aprendizaje de los Estudiantes

SUMMARY: For students in schools of nursing, health sciences, and premed, a systemic anatomy textbook with minimized contents, schematics, and mnemonics may be helpful for learning an otherwise often unappealing subject. In this study, we assess the educational effect of such a textbook. Schematic drawings, anatomy comics, and easily readable text were generated for the chapters of the book (e.g., skeletal system, articular system). The book was presented without charge via a webpage (anatomy.co.kr). Nursing students who were exposed to the book and those who were not exposed were compared; a survey was administered to those who were exposed. The students who read the presented textbook were more knowledgeable than those who used other textbooks. Hours spent reading the presented textbook and scores of fill-in-the-blank questions were positively correlated. In general, the students replied that the presented textbook was helpful for learning systemic anatomy. The systemic anatomy textbook accompanies preexisting textbooks in regional anatomy, neuroanatomy, and the histology, all of which are written by the same authors. We suggest anatomy instructors generate their own books with unique style to enrich the student learning process.

Para los estudiantes de las escuelas de enfermería, ciencias de la salud y premedicina, un libro de texto de anatomía sistémica con contenidos, esquemas y mnemónicos minimizados puede ser útil para aprender un tema que de otro modo sería poco atractivo. En este estudio, evaluamos el efecto educativo de dicho libro de texto. Se generaron dibujos esquemáticos, cómics de anatomía y texto de fácil lectura para los capítulos del libro (por ejemplo, sistema esquelético, sistema articular). El libro se presentó sin costo a través de una página web (anatomy.co.kr). Se compararon los estudiantes de enfermería que estuvieron expuestos al libro y los que no estuvieron expuestos. Se administró una encuesta a quienes estuvieron expuestos. Los estudiantes que leyeron el libro de texto presentado tenían más conocimientos que aquellos que usaron otros libros de texto. Las horas dedicadas a leer el libro de texto presentado y las decenas de preguntas para completar espacios en blanco se correlacionaron positivamente. En general, los estudiantes respondieron que el libro de texto presentado fue útil para aprender anatomía sistémica. El libro de texto de anatomía sistémica acompaña a los libros de texto preexistentes de anatomía regional, neuroanatomía e histología, todos escritos por los mismos autores. Sugerimos que los instructores de anatomía generen sus propios libros con un estilo único para enriquecer el proceso de aprendizaje de los estudiantes.

An iterative noisy annotation correction model for robust plant disease detection.

Previous work on plant disease detection demonstrated that object detectors generally suffer from degraded training data, and annotations with noise may cause the training task to fail. Well-annotated datasets are therefore crucial to build a robust detector. However, a good label set generally requires much expert knowledge and meticulous work, which is expensive and time-consuming. This paper aims to learn robust feature representations with inaccurate bounding boxes, thereby reducing the model requirements for annotation quality. Specifically, we analyze the distribution of noisy annotations in the real world. A teacher-student learning paradigm is proposed to correct inaccurate bounding boxes. The teacher model is used to rectify the degraded bounding boxes, and the student model extracts more robust feature representations from the corrected bounding boxes. Furthermore, the method can be easily generalized to semi-supervised learning paradigms and auto-labeling techniques. Experimental results show that applying our method to the Faster-RCNN detector achieves a 26% performance improvement on the noisy dataset. Besides, our method achieves approximately 75% of the performance of a fully supervised object detector when 1% of the labels are available. Overall, this work provides a robust solution to real-world location noise. It alleviates the challenges posed by noisy data to precision agriculture, optimizes data labeling technology, and encourages practitioners to further investigate plant disease detection and intelligent agriculture at a lower cost. The code will be released at https://github.com/JiuqingDong/TS_OAMIL-for-Plant-disease-detection.

Structural basis for ligand recognition and signaling of hydroxy-carboxylic acid receptor 2.

Hydroxycarboxylic acid receptors (HCAR1, HCAR2, and HCAR3) transduce Gi/o signaling upon biding to molecules such as lactic acid, butyric acid and 3-hydroxyoctanoic acid, which are associated with lipolytic and atherogenic activity, and neuroinflammation. Although many reports have elucidated the function of HCAR2 and its potential as a therapeutic target for treating not only dyslipidemia but also neuroimmune disorders such as multiple sclerosis and Parkinson's disease, the structural basis of ligand recognition and ligand-induced Gi-coupling remains unclear. Here we report three cryo-EM structures of the human HCAR2-Gi signaling complex, each bound with different ligands: niacin, acipimox or GSK256073. All three agonists are held in a deep pocket lined by residues that are not conserved in HCAR1 and HCAR3. A distinct hairpin loop at the HCAR2 N-terminus and extra-cellular loop 2 (ECL2) completely enclose the ligand. These structures also reveal the agonist-induced conformational changes propagated to the G-protein-coupling interface during activation. Collectively, the structures presented here are expected to help in the design of ligands specific for HCAR2, leading to new drugs for the treatment of various diseases such as dyslipidemia and inflammation.

BRD4/nuclear PD-L1/RelB circuit is involved in the stemness of breast cancer cells.

BACKGROUND: Breast cancer (BC) is the most common cancer diagnosed in women worldwide. BC stem cells (BCSCs) have been known to be involved in the carcinogenesis of the breast and contribute to therapeutic resistance. The programmed death-ligand 1 (PD-L1) expression of BC correlated with a poor prognosis. Immunotherapies that target PD-L1 have great potential and have been successful when applied to cancer treatment. However, whether PD-L1 regulates BCSC formation is unknown. METHODS: BCSCs were enriched by serum-free suspension culture. The properties of BCSCs were examined by mammosphere formation assay, CD44+/Cd24-, aldehyde dehydrogenase (ALDH) assay, CSC marker analysis, and mammosphere growth assay. To elucidate the functions of bromodomain-containing protein 4 (BRD4), nuclear PD-L1, and RelB proteins in the stemness of BCSCs, mammosphere formation was examined using BRD4 inhibitor and degrader, PD-L1 degrader, and RelB inhibitor. The antitumor function of 3',4',7,8-tetrahydroxyflavone (THF), a specific BRD4 inhibitor, was studied through in vivo tumor model and mouse studies, and the protein levels of c-Myc, PD-L1, and RelB were examined in tumor model under THF treatment. RESULTS: BRD4 was upregulated in breast CSCs and regulates the stemness of BCs. The downregulation of BRD4 using BRD4 PROTAC, ARV-825, and BRD4 inhibitor, (+)-JQ1, inhibits mammosphere formation and reduces the levels of breast CSC markers (CD44+/CD24- and ALDH1), stem cell marker genes, and mammosphere growth. BRD4 inhibitor (JQ1) and degrader (ARV825) downregulate membrane and nuclear fractions of PD-L1 through the inhibition of PD-L1 transcript levels. The knockdown of PD-L1 inhibits mammosphere formation. Verteporfin, a PD-L1 degrader, inhibits the transcripts and protein levels of PD-L1 and downregulates the transcript and protein levels of RelB. Calcitriol, a RelB inhibitor, and the knockdown of RelB using si-RelB regulate mammosphere formation through interleukin-6 (IL-6) expression. THF is a natural product and a potent selective BRD4 inhibitor, inhibits mammosphere formation, and reduces the levels of CD44+/CD24- and mammosphere growth by downregulating c-Myc, PD-L1, and RelB. 3',4',7,8-THF shows tumoricidal activity and increased levels of CD3+CD4+ and CD3+CD8+ T-cells in the tumor and tumor-draining lymph nodes (TDLNs) in the murine tumor model using 4T1 and MC38 cells. CONCLUSIONS: The results show the first evidence of the essential role of the BRD4/nuclear PD-L1/RelB axis in breast CSC formation. The nuclear PD-L1 regulates RelB, and the RelB/p65 complex induces IL6 and breast CSC formation. Targeting nuclear PD-L1 represents a potential and novel tool for immunotherapies of intractable BC. Video Abstract.

Improving Known-Unknown Cattle's Face Recognition for Smart Livestock Farm Management.

Accurate identification of individual cattle is of paramount importance in precision livestock farming, enabling the monitoring of cattle behavior, disease prevention, and enhanced animal welfare. Unlike human faces, the faces of most Hanwoo cattle, a native breed of Korea, exhibit significant similarities and have the same body color, posing a substantial challenge in accurately distinguishing between individual cattle. In this study, we sought to extend the closed-set scope (only including identifying known individuals) to a more-adaptable open-set recognition scenario (identifying both known and unknown individuals) termed Cattle's Face Open-Set Recognition (CFOSR). By integrating open-set techniques to enhance the closed-set accuracy, the proposed method simultaneously addresses the open-set scenario. In CFOSR, the objective is to develop a trained model capable of accurately identifying known individuals, while effectively handling unknown or novel individuals, even in cases where the model has been trained solely on known individuals. To address this challenge, we propose a novel approach that integrates Adversarial Reciprocal Points Learning (ARPL), a state-of-the-art open-set recognition method, with the effectiveness of Additive Margin Softmax loss (AM-Softmax). ARPL was leveraged to mitigate the overlap between spaces of known and unknown or unregistered cattle. At the same time, AM-Softmax was chosen over the conventional Cross-Entropy loss (CE) to classify known individuals. The empirical results obtained from a real-world dataset demonstrated the effectiveness of the ARPL and AM-Softmax techniques in achieving both intra-class compactness and inter-class separability. Notably, the results of the open-set recognition and closed-set recognition validated the superior performance of our proposed method compared to existing algorithms. To be more precise, our method achieved an AUROC of 91.84 and an OSCR of 87.85 in the context of open-set recognition on a complex dataset. Simultaneously, it demonstrated an accuracy of 94.46 for closed-set recognition. We believe that our study provides a novel vision to improve the classification accuracy of the closed set. Simultaneously, it holds the potential to significantly contribute to herd monitoring and inventory management, especially in scenarios involving the presence of unknown or novel cattle.

A longer Achilles tendon moment arm length is not associated with superior hopping performance.

Variability in musculoskeletal and lower leg structure has the potential to influence hopping height. Achilles tendon moment arm length and plantarflexor muscle strength can influence ankle joint torque development and, consequently, hopping performance. While most studies have examined the connection of the Achilles tendon moment arm with hopping performance including the resting length, in this study we attempted to explore how the changes in Achilles tendon moment arm are related to hopping performance. Therefore, the purpose of this study was to test for correlations between foot and lower leg muscle structure parameters (i.e., muscle mass, volume, cross-sectional area and Achilles tendon moment arm length) and hopping height performance in relation to changes in Achilles tendon moment arm length. Eighteen participants (10 males 8 female) performed repetitive bilateral hopping on a force platform while sagittal plane kinematics of the lower leg were recorded. Additionally, maximal isometric plantarflexion was measured. To obtain structural parameters of the lower leg, the right lower leg of each participant was scanned with magnetic resonance imaging. The cross-sectional areas of the Achilles tendon, soleus, lateral and medial gastrocnemius were measured, while muscle volumes, muscle mass, and Achilles tendon moment arm length were calculated. Contrary to our initial assumption, longer Achilles tendon moment arm did not result in superior hopping performance. Interestingly, neither maximal isometric plantarflexion force nor muscle size correlated with repetitive bilateral hopping performance. We can assume that the mechanical characteristics of the tendon and the effective utilization of the stored strain energy in the tendon may play a more important role in repetitive hopping than the structural parameters of the lower leg.

Fatty Liver Change in Older Adults as an Important Risk Factor for Type 2 Diabetes: A Nationwide Cohort Study.

OBJECTIVE: To examine the association between changes in fatty liver disease (FLD) over time and the risk of type 2 diabetes in elderly individuals with prediabetes. METHODS: A total of 156,984 elderly individuals with prediabetes who underwent national health screening in 2009 and 2011 were followed up through December 31, 2019. The FLD status was defined as a change in the fatty liver index. Prediabetes was defined as impaired fasting glucose levels at baseline. Multivariable Cox proportional hazards regression was used to calculate the hazard ratio and CIs for type 2 diabetes according to the changes in FLD. RESULTS: During a median of 8.35 years of follow-up, type 2 diabetes developed in 29,422 (18.7%) elderly individuals with prediabetes. Multivariable adjusted hazard ratio of type 2 diabetes according to FLD change was 2.22 (95% CI, 2.11 to 2.34) in individuals with persistent FLD compared with those who have never had FLD. Although overall weight loss of 5% or more was associated with a 7% lower risk of type 2 diabetes in total participants, fatty liver status was important. Even with weight loss, those with a history of fatty liver-resolved FLD, new FLD, or persistent FLD-had an increased risk of type 2 diabetes. The risk of type 2 diabetes did not increase in individuals with sustained FLD-free status, regardless of weight change. CONCLUSION: The presence and change of FLD are important factors for the development of type 2 diabetes in elderly individuals with prediabetes.

A New Deep Learning-based Dynamic Paradigm Towards Open-World Plant Disease Detection.

Plant disease detection has made significant strides thanks to the emergence of deep learning. However, existing methods have been limited to closed-set and static learning settings, where models are trained using a specific dataset. This confinement restricts the model's adaptability when encountering samples from unseen disease categories. Additionally, there is a challenge of knowledge degradation for these static learning settings, as the acquisition of new knowledge tends to overwrite the old when learning new categories. To overcome these limitations, this study introduces a novel paradigm for plant disease detection called open-world setting. Our approach can infer disease categories that have never been seen during the model training phase and gradually learn these unseen diseases through dynamic knowledge updates in the next training phase. Specifically, we utilize a well-trained unknown-aware region proposal network to generate pseudo-labels for unknown diseases during training and employ a class-agnostic classifier to enhance the recall rate for unknown diseases. Besides, we employ a sample replay strategy to maintain recognition ability for previously learned classes. Extensive experimental evaluation and ablation studies investigate the efficacy of our method in detecting old and unknown classes. Remarkably, our method demonstrates robust generalization ability even in cross-species disease detection experiments. Overall, this open-world and dynamically updated detection method shows promising potential to become the future paradigm for plant disease detection. We discuss open issues including classification and localization, and propose promising approaches to address them. We encourage further research in the community to tackle the crucial challenges in open-world plant disease detection. The code will be released at https://github.com/JiuqingDong/OWPDD.

Embracing limited and imperfect training datasets: opportunities and challenges in plant disease recognition using deep learning.

Recent advancements in deep learning have brought significant improvements to plant disease recognition. However, achieving satisfactory performance often requires high-quality training datasets, which are challenging and expensive to collect. Consequently, the practical application of current deep learning-based methods in real-world scenarios is hindered by the scarcity of high-quality datasets. In this paper, we argue that embracing poor datasets is viable and aims to explicitly define the challenges associated with using these datasets. To delve into this topic, we analyze the characteristics of high-quality datasets, namely, large-scale images and desired annotation, and contrast them with the limited and imperfect nature of poor datasets. Challenges arise when the training datasets deviate from these characteristics. To provide a comprehensive understanding, we propose a novel and informative taxonomy that categorizes these challenges. Furthermore, we offer a brief overview of existing studies and approaches that address these challenges. We point out that our paper sheds light on the importance of embracing poor datasets, enhances the understanding of the associated challenges, and contributes to the ambitious objective of deploying deep learning in real-world applications. To facilitate the progress, we finally describe several outstanding questions and point out potential future directions. Although our primary focus is on plant disease recognition, we emphasize that the principles of embracing and analyzing poor datasets are applicable to a wider range of domains, including agriculture. Our project is public available at https://github.com/xml94/EmbracingLimitedImperfectTrainingDatasets.

Local refinement mechanism for improved plant leaf segmentation in cluttered backgrounds.

Plant phenotyping is a critical field in agriculture, aiming to understand crop growth under specific conditions. Recent research uses images to describe plant characteristics by detecting visual information within organs such as leaves, flowers, stems, and fruits. However, processing data in real field conditions, with challenges such as image blurring and occlusion, requires improvement. This paper proposes a deep learning-based approach for leaf instance segmentation with a local refinement mechanism to enhance performance in cluttered backgrounds. The refinement mechanism employs Gaussian low-pass and High-boost filters to enhance target instances and can be applied to the training or testing dataset. An instance segmentation architecture generates segmented masks and detected areas, facilitating the derivation of phenotypic information, such as leaf count and size. Experimental results on a tomato leaf dataset demonstrate the system's accuracy in segmenting target leaves despite complex backgrounds. The investigation of the refinement mechanism with different kernel sizes reveals that larger kernel sizes benefit the system's ability to generate more leaf instances when using a High-boost filter, while prediction performance decays with larger Gaussian low-pass filter kernel sizes. This research addresses challenges in real greenhouse scenarios and enables automatic recognition of phenotypic data for smart agriculture. The proposed approach has the potential to enhance agricultural practices, ultimately leading to improved crop yields and productivity.