Highest Kappa Coefficient Research Articles

Identifying the Areas at Risk of Huaico Occurrences in the Department of Lima, Peru

Because of local climate, a phenomenon called huaico occurs in the coastal regions of Peru, configured by an alluvial flow of surface runoff caused by precipitation and accompanied by the transport of solid particles. A total of 24% of the huaicos recorded in Peru from 2003 to 2019 were concentrated in the Department of Lima alone and affected 38,000 people. Thus, the aim of this study was to use Maxent to identify the areas at risk of huaicos in this department. To this end, a georeferenced database was created that included the locations of these events for modeling. We used variables suggested by Peru’s Geological, Mining, and Metallurgical Institute (INGEMMET)—geology, geomorphology, DEM, slope, and precipitation—which returned extremely high kappa coefficients. Approximately 42% of Lima’s area is likely to have a huaico occurrence. The most crucial variable for the models was the geomorphological classification characterized by the accumulation of mobilized material, as was the case in previous huaico models. In addition, the monthly approach should have been more effective at determining the differences in the precipitation levels. Thus, new models for the coastal departments of Peru using Maxent algorithms should take a new approach related to precipitation, although the use of Maxent proved satisfactory.

Efficient UAV-Based Automatic Classification of Cassava Fields Using K-Means and Spectral Trend Analysis

High-resolution images captured by Unmanned Aerial Vehicles (UAVs) play a vital role in precision agriculture, particularly in evaluating crop health and detecting weeds. However, the detailed pixel information in these images makes classification a time-consuming and resource-intensive process. Despite these challenges, UAV imagery is increasingly utilized for various agricultural classification tasks. This study introduces an automatic classification method designed to streamline the process, specifically targeting cassava plants, weeds, and soil classification. The approach combines K-means unsupervised classification with spectral trend-based labeling, significantly reducing the need for manual intervention. The method ensures reliable and accurate classification results by leveraging color indices derived from RGB data and applying mean-shift filtering parameters. Key findings reveal that the combination of the blue (B) channel, Visible Atmospherically Resistant Index (VARI), and color index (CI) with filtering parameters, including a spatial radius (sp) = 5 and a color radius (sr) = 10, effectively differentiates soil from vegetation. Notably, using the green (G) channel, excess red (ExR), and excess green (ExG) with filtering parameters (sp = 10, sr = 20) successfully distinguishes cassava from weeds. The classification maps generated by this method achieved high kappa coefficients of 0.96, with accuracy levels comparable to supervised methods like Random Forest classification. This technique offers significant reductions in processing time compared to traditional methods and does not require training data, making it adaptable to different cassava fields captured by various UAV-mounted optical sensors. Ultimately, the proposed classification process minimizes manual intervention by incorporating efficient pre-processing steps into the classification workflow, making it a valuable tool for precision agriculture.

Optimizing crop monitoring: mapping cultivation stages and types with sentinel-1/2 and random forest algorithm

ABSTRACT Crop-type mapping is essential for agricultural monitoring, improving crop yield models, and supporting sustainable land management. Studying crop phenology, which includes the timing from germination to maturity, is equally important for precise agricultural practices. However, maps detailing crop phenology are scarce. This study addresses this gap by mapping both crop types and cultivation time, offering valuable insights for agriculture applications. The research integrates Sentinel-1/2 data, NDVI phenology curves, and machine learning (ML) algorithms through Google Earth Engine to map crops in El-Behera Governorate, Egypt. By stratifying the region into different soil types, it accounts for various agro-environmental conditions. Field visits in July 2022 provided crop-type information, and Sentinel-2 NDVI curves were used to determine cultivation time. The random forest (RF) classifier, known for its accuracy, was employed, achieving an overall accuracy (OA) of 97.4% for crop type and 95.6% for cultivation time, with high Kappa coefficients. Feature importance analysis highlighted the significance of specific spectral bands, particularly in the Red Edge region, for classification accuracy. Hyperparameter tuning across different zones demonstrated the need for customized tuning. This study enhances understanding of agro-environmental variability’s impact on crop classification and shows the potential of the RF algorithm in advancing crop mapping, suggesting further studies on how shifts in cultivation time could affect crop yield.

Concordance and Discordance Between Radiology Residents and Consultant Radiologist Interpretation Of CT Brain

OBJECTIVES The primary objective of this study is to assess the degree of concordance and discordance between the interpretations of computed tomography (CT) brain images by resident and consultant radiologists while emphasizing the critical significance of accurate image interpretation for informed clinical decision-making. METHODOLOGY The evaluation of radiology reports for CT Brain interpretation through a prospective analysis at the Radiology Department of Rehman Medical Institute over two years, from 1st October 2020 to 31st October 2022. A total of 198 patients who underwent cranial CT scans were interpreted by residents (R1, R2, R3, R4). Following this, the consultant radiologists reviewed the images and completed their reports. The reports of the residents and the consultant radiologists were then compared, and concordance was achieved when the residents’ reports were consistent with the final radiologist’s reports. The data collected were recorded in Microsoft Excel. The statistical analysis was performed using SPSS version 22 (IBM Corp., Armonk, NY), and the kappa coefficient was used to determine the level of agreement between residents and consultants.RESULTSAmong the 198 CT Head reports evaluated, 186 of them were in agreement with the final report of the consultant radiologist. Of the correctly diagnosed cases, R1 correctly diagnosed 46 cases, R2 correctly diagnosed 80 cases, R3 correctly diagnosed 54 cases, and R4 correctly diagnosed 6 cases. Our study achieved a percentage agreement of 93.93, with a Cohen's kappa coefficient of 0.8.CONCLUSIONThe overall concordance rate between residents and consultant radiologists was 93.93%, with a kappa coefficient 0.8. This high kappa coefficient indicates strong agreement between the two groups.

Monitoring seagrass meadows in Maputo Bay using integrated remote sensing techniques and machine learning

Seagrass meadows are one of the most productive and valuable ecosystems on the planet. Monitoring seagrass meadows is essential to understand how these habitats change, and to develop better management and conservation practices. This study integrated satellite imagery from Sentinel-2 and Unmanned Aerial Vehicles (UAV) using machine learning to provide a consistent classification approach for monitoring seagrass in Maputo Bay, southern Mozambique. Sentinel-2 imagery was used to map seagrass extent and changes in Maputo Bay. The UAV systems were used to map seagrass at species level and biomass. All three algorithms tested in the ArcGIS environment could detect seagrass with high producer accuracy and Kappa coefficient. The area of seagrass in Maputo Bay decreased by 33.4 % between 1991 and 2023, with a decreasing trend of 0.48 km2/yr. A zonation pattern was observed for Oceana serrulata and Zostera capensis from the UAV imagery. The small and narrow leaved species (Z. capensis) occurred in the intertidal zone replaced by the broadleaved species (O. serrulata) in the subtidal. The total average aboveground biomass was 33.2 kg dry weight for the mapped area. The results of this study will guide implementation of combined satellite and UAV imagery with machine learning techniques for seagrass monitoring and restoration in Mozambique.

COMPARATIVE EVALUATION OF ARTIFICIAL INTELLIGENCE AND DRUG INTERACTION TOOLS: A PERSPECTIVE WITH THE EXAMPLE OF CLOPIDOGREL

Objective: The study aims to compare the ability of free artificial intelligence (AI) chatbots to detect drug interactions with freely available drug interaction tools, using clopidogrel as an example. Material and Method: The Lexicomp database was used as a reference to determine drug interactions with clopidogrel. ChatGPT-3.5 AI and Bing AI were selected as the free AI chatbots. Medscape Drug Interaction Checker, DrugBank Drug Interaction Checker and Epocrates Interaction Check were selected as free drug interaction tools. Accuracy score and comprehensiveness score were calculated for each drug interaction tool and AI chatbots. The kappa coefficient was calculated to assess inter-source agreement for interaction severity. Result and Discussion: The results most similar to those of Lexicomp were obtained from the DrugBank and the ChatGPT-3.5 AI chatbot. The ChatGPT-3.5 AI chatbot performed best, with 69 correct results and an accuracy score of 307. ChatGPT-3.5 AI has the highest overall score of 387 points for accuracy and comprehensiveness. In addition, the highest kappa coefficient with Lexicomp was found for ChatGPT-3.5 AI chatbot (0.201, fair agreement). However, some of the results obtained by ChatGPT-3.5 AI need to be improved as they are incorrect/inadequate. Therefore, information obtained using AI tools should not be used as a reference for clinical applications by healthcare professionals and patients should not change their treatment without consulting doctor.

Study of Diagnostic Accuracy: Fundus Photography vs. Optical Coherence Tomography

(1) Background: This study aimed to determine the diagnostic accuracy that optical coherence tomography (OCT) can add to fundus photography (FP) in assessing the condition of the retinal posterior pole. (2) Methods: We conducted two blocks of analysis: First, the posterior pole of each eye was examined using an FP non-mydriatic imaging device. Second, OCT was used in addition to FP. After consolidating the specific diagnostic criteria, the assessments were evaluated using two blinded independent investigation groups (by optometrists, and by ophthalmologists who were considered the gold standard). (3) Results: We calculated the diagnostic accuracy of FP compared to OCT and found that they had similar sensitivity. FP had a slightly higher specificity (p-value: 0.01), and OCT had a higher kappa coefficient with 0.50 (95% CI: 0.46–0.55) vs. 0.39 (95% CI: 0.34–0.45) for FP. (4) Conclusions: On the basis that the role of the optometrist in Spain is not to diagnose but to detect lesions and refer patients to an ophthalmologist, the results of this study support the use of OCT, which provided gradable images in almost all examined eyes (97.5%), compared to FP (73.5%). However, optometrists need a detailed and standardized guide in order to conduct evaluations according to the ophthalmologist’s criteria.

Predicting urban tomorrow: CA-Markov modeling and district evolution

The global population is experiencing exponential growth, resulting in a substantial increase in urbanization and subsequent urban expansion. This uncontrolled expansion, often termed urban sprawl, poses significant challenges to sustainable urban development. Understanding the importance of this concern, obtaining precise information about changes in Land Use/Land Cover (LULC) becomes crucial. This research examines the transformative processes of LULC and urban expansion over a 20-year period in two distinct study areas. For this purpose, built-up areas are first analysed using satellite-derived land surface temperature data, revealing temperature increase over time attributed to urban expansion. To project future trends using CA-Markov model, the suitability maps for each LULC class aggregated through the Multi Criteria Evaluation (MCE) method. Subsequently, the CA-Markov simulates the LULC maps for 2017 and 2018 for each study area, with a focus on the year 2050. Model calibrated by comparing the simulated maps with the actual maps in both study areas, and the reliability is affirmed by high Kappa coefficients (> 80%). Consequently, the study predicts LULC maps for 2050, revealing that both areas will experience a continued increase in built-up areas, a decrease in forested areas, and a relative stability in agricultural zones over the next 33 years.

Automatic Wake and Deep-Sleep Stage Classification Based on Wigner-Ville Distribution Using a Single Electroencephalogram Signal.

This research paper outlines a method for automatically classifying wakefulness and deep sleep stage (N3) based on the American Academy of Sleep Medicine (AASM) standards. The study employed a single-channel EEG signal, leveraging the Wigner-Ville Distribution (WVD) for time-frequency analysis to determine EEG energy per second in specific frequency bands (δ, θ, α, and entire band). Particle Swarm Optimization (PSO) was used to optimize thresholds for distinguishing between wakefulness and stage N3. This process aims to mimic a sleep technician's visual scoring but in an automated fashion, with features and thresholds extracted to classify epochs into correct sleep stages. The study's methodology was validated using overnight PSG recordings from 20 subjects, which were evaluated by a technician. The PSG setup followed the 10-20 standard system with varying sampling rates from different hospitals. Two baselines, T1 for the wake stage and T2 for the N3 stage, were calculated using PSO to ascertain the best thresholds, which were then used to classify EEG epochs. The results showed high sensitivity, accuracy, and kappa coefficient, indicating the effectiveness of the classification algorithm. They suggest that the proposed method can reliably determine sleep stages, being aligned closely with the AASM standards and offering an intuitive approach. The paper highlights the strengths of the proposed method over traditional classifiers and expresses the intentions to extend the algorithm to classify all sleep stages in the future.

Spatiotemporal changes of desertification degree in the Algerian green barrier over the last four decades (1984–2020)

Desertification is a major challenge for the sustainable development of arid and semiarid regions worldwide. The major novelty of this study lies in assessing, for the first time, the desertification degree changes in the Algerian green barrier using remote sensing and the Analytic Hierarchy Process (AHP) approach from 1984 through 2020. Four widely used remote sensing indices (LSA, NDVI, TGSI, and PDSI) were used in the Google Earth Engine (GEE) cloud computing platform as indicators of desertification in the study area. The results of the accuracy assessment of the generated desertification degree maps for 1984, 2000, and 2020 show high Kappa coefficient values exceeding 80%. Overall, the degree of desertification of the green barrier increased from 1984 to 2020. The area of medium degree has decreased from 15393.21 km2 (37.70%) to 9624.43 km2 (23.57%) between 1984 and 2020, respectively; whereas the area of high degree has increased from 23907.08 km2 (58.55%) to 30502.52 km2 (74.70%) for the same period. The combination of remote sensing and the AHP technique has proven to be effective in the spatiotemporal monitoring of the desertification degree particularly in data–scarce regions. Our findings emphasize the high expansion of desertification across the green barrier area, which information is essential for adopting adequate strategies for desertification control in the future.

Enhancing diabetic retinopathy detection through preprocessing and feature extraction with MGA-CSG algorithm

Anticipatory monitoring of diabetic retinal (DR) disease is crucial in preventing vision loss and blindness, making it a leading cause of worldwide vision impairment. In this study, we propose a novel technique, the Modified Generative Adversarial-based Crossover Salp Grasshopper (MGA-CSG) approach, for early prediction and accurate classification of diabetic retinal diseases using fundus image datasets. The proposed MGA-CSG approach combines deep learning with an optimization algorithm to achieve superior classification accuracy. We start with pre-processing of data to attenuate image variation, convert intensities, denoise, and enhance contrast in the fundus images. Next, a Convolutional Neural Network (CNN) is used for feature extraction, capturing essential information from the images. To address the limited amount of feature vectors generated by the CNN and improve learning uncertainty, we employ a Generative Adversarial Network (GAN) model to augment the feature vectors. It generates additional images, allowing the classifier to train with fewer input samples, leading to improved classification performance. One of the key novelties of our approach lies in the utilization of both labeled and unlabeled data samples, enhancing the GAN training performance. We replace the binary classifier in the GAN with a multiclass classifier, enabling separate discrimination of retinal disease classes. The MGA-CSG approach leverages the crossover Grasshopper Optimizer Algorithm (GOA) and Salp Swarm Algorithm (SSA) to optimize the model and avoid falling into local optimal solutions. This optimization process further enhances the accuracy of diabetic retinal disease classification. Experimental results demonstrate the superiority of the proposed MGA-CSG approach, achieving high accuracy, precision, recall, specificity, F1-measure, and Kappa coefficient in predicting glaucoma, diabetic retinopathy, cataract, macular edema, and myopia. The classification achieved impressive accuracy, precision, recall, specificity, F1-measure, and Kappa coefficient of 98.8%, 98.7%, 96.5%, 97.1%, 98.2%, and 95%, respectively. Overall, our novel MGA-CSG approach offers an efficient and accurate method for early screening and classification of diabetic retinal diseases. The combination of CNN-based feature extraction, GAN-based augmentation, and MGA-CSG optimization contributes to the model's high performance. This approach holds significant potential in improving healthcare management and providing timely interventions for patients with retinal diseases.

Forecasting environmental water availability of lakes using temporal fusion transformer: case studies of China's two largest freshwater lakes.

Preserving lacustrine ecosystems is vital for sustainable watershed development, and forecasting the environmental water availability of lakes would support policymakers in developing sound management strategies. This study proposed a methodology that merges the lake water level prediction and environmental water availability evaluation. The temporal fusion transformer (TFT) model forecasted the lake water levels for the next 7 days by inputting the streamflow and lake water level data for the past 30 days. The environmental water availability was assessed by comparing the forecasted lake water levels with the environmental water requirements, resulting in adequate, regular, scarce, and severely scarce environmental water availability. The methodology was tested in two case studies: Poyang Lake and Dongting Lake, the two largest freshwater lakes in the Yangtze River Basin, China. The TFT model performed well in forecasting the lake water levels, as shown by the high coefficient of determination and finite root mean square error. The coefficients of determination exceeded 0.98 during the model training, validation, and test for both Poyang Lake and Dongting Lake, and the root mean square errors ranged from 0.06 to 0.46 m. The accurate prediction of lake water level promoted the precise forecasting of the environmental water availability with the high Kappa coefficient exceeding 0.90. Results indicated the rationality and effectiveness of integrating the lake water level prediction and environmental water availability evaluation. Future research includes the applicability of the TFT model to other lakes worldwide to test the proposed approach and investigate strategies to cope with environmental water scarcity.

Serum neurofilament light chain levels in patients with cognitive deficits and movement disorders: comparison of cerebrospinal and serum neurofilament light chain levels with other biomarkers.

Serum neurofilament light chain (S NfL) is a non-specific marker of neuronal damage, including Alzheimer's disease (AD). We aimed to verify the reference interval (RI) of serum NfL using a highly sensitive ELISA, and to estimate the optimal cut-off value for neuronal damage. Our second objective was to compare NfL in cerebrospinal fluid (CSF) and serum (S) with the routine neurodegeneration biomarkers used in AD, and to assess their concentrations relative to the degree of cognitive deficit. Samples from 124 healthy volunteers were used to estimate the S NfL RI. For the comparison study, we used CSF and S samples from 112 patients with cognitive disorders. Cognitive functions were assessed using the mini-mental state examination. ELISA assays were used to determine the CSF and S NfL levels, CSF β-amyloid peptide42 (Aβ42), CSF β-amyloid peptide40 (Aβ40), CSF total tau protein (tTau), CSF phosphorylated tau protein (pTau), and CSF alpha-synuclein (αS). The estimated RI of S NfL were 2.25-9.19 ng.L-1. The cut-off value of S NfL for assessing the degree of neuronal impairment was 10.5 ng.L-1. We found a moderate statistically significant correlation between S NfL and CSF Aβ42 in the group with movement disorders, without dementia (rs = 0.631; p = 0.016); between S NfL and CSF Aβ40 in the group with movement disorder plus dementia (rs = -0.750; p = 0.052); between S NfL and CSF tTau in the control group (rs = 0.689; p = 0.009); and between S NfL and CSF pTau in the control group (rs = 0.749; p = 0.003). The non-parametric Kruskal-Wallis test revealed statistically significant differences between S NfL, CSF NfL, CSF Aβ42, CSF tTau, and CSF pTau and diagnosis within groups. The highest kappa coefficients were found between the concentrations of S NfL and CSF NfL (κ = 0.480) and between CSF NfL and CSF tTau (κ = 0.351). Our results suggested that NfL and tTau in CSF of patients with cognitive decline could be replaced by the less-invasive determination of S NfL using a highly sensitive ELISA method. S NfL reflected the severity of cognitive deficits assessed by mini-mental state examination (MMSE). However, S NfL is not specific to AD and does not appear to be a suitable biomarker for early diagnosis of AD.

Toward a real validation of passive microwave snowmelt detection algorithms over the Antarctic Ice sheet

Antarctic Ice Sheet (AIS) surface snowmelt can profoundly impact climatological and hydrological processes, as well as energy and mass balances. Passive microwave remote sensing has proven highly effective in detecting snowmelt, and the accuracy of the algorithms is typically assessed using air temperature (Tair). Nonetheless, the occurrence of snowmelt cannot be solely determined by Tair being above the melting point, it can only provide evaluation of the algorithms rather than a real validation. This study is the first to conduct a reliable validation of six commonly used algorithms by using the melt rate data from a surface energy balance model, including the cross-polarized gradient ratio (XPGR) algorithm, the diurnal amplitude variation (DAV) algorithm, the Microwave Emission Model of Layered Snowpacks based (MEMLS-based) algorithm, the constant threshold algorithm (CTA), the Ashcraft and Long Algorithm (ALA), and the M + 30 K (M30 hereafter) algorithm. Results suggest that the M30 algorithm and MEMLS-based algorithm have overall better performances, with high overall accuracies (with a value of 95.81 % and 95.56 %) and high kappa coefficient (with a value of 0.69 and 0.68). Though Tair is not reliable for evaluating the algorithms, a threshold of Tair = − 0.8 ℃ can serve as an alternative criterion in the absence of melt rate data. Additionally, two algorithms with the highest accuracy were selected to investigate interannual snowmelt trends in the Antarctic and the sub-regions. The results show a significant increase in melt extent and density in the Wilkes and Adelie sub-region during the period 1979–2020.

Quantitative assessment of Land use/land cover changes in a developing region using machine learning algorithms: A case study in the Kurdistan Region, Iraq

The identification of land use/land cover (LULC) changes is important for monitoring, evaluating, and preserving natural resources. In the Kurdistan region, the utilization of remotely sensed data to assess the effectiveness of machine learning algorithms (MLAs) for LULC classification and change detection analysis has been limited. This study monitors and analyzes LULC changes in the study area from 1991 to 2021 using a quantitative approach with multi-temporal Landsat imagery. Five MLAs were applied: Support Vector Machine (SVM), Random Forest (RF), Artificial Neural Network (ANN), K-Nearest Neighbor (KNN), and Extreme Gradient Boosting (XGBoost). The results showed that the RF algorithm produced the most accurate maps of the three-decade study period, accompanied by a high kappa coefficient (0.93–0.97) compared with the SVM (0.91–0.95), ANN (0.91–0.96), KNN (0.92–0.96), and XGBoost (0.92–0.95) algorithms. Consequently, the RF classifier was implemented to categorize all obtainable satellite images. Socioeconomic changes throughout these transition periods were revealed by the change detection results. Rangeland and barren land areas decreased by 11.33 % (−402.03 km2) and 6.68 % (−236.8 km2), respectively. The transmission increases of 13.54 % (480.18 km2), 3.43 % (151.74 km2), and 0.71 % (25.22 km2) occurred in agricultural land, forest, and built-up areas, respectively. The outcomes of this study contribute significantly to LULC monitoring in developing regions, guiding stakeholders to identify vulnerable areas for better land use planning and sustainable environmental protection.

Predicting Future Land Use Utilizing Economic and Land Surface Parameters with ANN and Markov Chain Models

The main aim of this study is to comprehensively analyze the dynamics of land use and land cover (LULC) changes in the Bathinda region of Punjab, India, encompassing historical, current, and future trends. To forecast future LULC, the Cellular Automaton–Markov Chain (CA) based on artificial neural network (ANN) concepts was used using cartographic variables such as environmental, economic, and cultural. For segmenting LULC, the study used a combination of ML models, such as support vector machine (SVM) and Maximum Likelihood Classifier (MLC). The study is empirical in nature, and it employs quantitative analyses to shed light on LULC variations through time. The result indicates that the barren land is expected to shrink from 55.2 km2 in 1990 to 5.6 km2 in 2050, signifying better land management or increasing human activity. Vegetative expanses, on the other hand, are expected to rise from 81.3 km2 in 1990 to 205.6 km2 in 2050, reflecting a balance between urbanization and ecological conservation. Agricultural fields are expected to increase from 2597.4 km2 in 1990 to 2859.6 km2 in 2020 before stabilizing at 2898.4 km2 in 2050. Water landscapes are expected to shrink from 13.4 km2 in 1990 to 5.6 km2 in 2050, providing possible issues for water resources. Wetland regions are expected to decrease, thus complicating irrigation and groundwater reservoir sustainability. These findings are confirmed by strong statistical indices, with this study’s high kappa coefficients of Kno (0.97), Kstandard (0.95), and Klocation (0.97) indicating a reasonable level of accuracy in CA prediction. From the result of the F1 score, a significant issue was found in MLC for segmenting vegetation, and the issue was resolved in SVM classification. The findings of this study can be used to inform land use policy and plans for sustainable development in the region and beyond.

Advancing prostate cancer detection: a comparative analysis of PCLDA-SVM and PCLDA-KNN classifiers for enhanced diagnostic accuracy

This investigation aimed to assess the effectiveness of different classification models in diagnosing prostate cancer using a screening dataset obtained from the National Cancer Institute’s Cancer Data Access System. The dataset was first reduced using the PCLDA method, which combines Principal Component Analysis and Linear Discriminant Analysis. Two classifiers, Support Vector Machine (SVM) and k-Nearest Neighbour (KNN), were then applied to compare their performance. The results showed that the PCLDA-SVM model achieved an impressive accuracy rate of 97.99%, with a precision of 0.92, sensitivity of 92.83%, specificity of 97.65%, and F1 score of 0.93. Additionally, it demonstrated a low error rate of 0.016 and a Matthews Correlation Coefficient (MCC) and Kappa coefficient of 0.946. On the other hand, the PCLDA-KNN model also performed well, achieving an accuracy of 97.8%, precision of 0.93, sensitivity of 93.39%, specificity of 97.86%, an F1 score of 0.92, a high MCC and Kappa coefficient of 0.98, and an error rate of 0.006. In conclusion, the PCLDA-SVM method exhibited improved efficacy in diagnosing prostate cancer compared to the PCLDA-KNN model. Both models, however, showed promising results, suggesting the potential of these classifiers in prostate cancer diagnosis.

Exploring Effective Detection and Spatial Pattern of Prickly Pear Cactus (Opuntia Genus) from Airborne Imagery before and after Prescribed Fires in the Edwards Plateau

Over the past century, prickly pear (PP) cactus (e.g., genus Opuntia; subgenus Platyopuntia) has increased on semi-arid rangelands. Effective detection of cacti abundance and spatial pattern is challenging due to the inherent heterogeneity of rangeland landscapes. In this study, high-resolution multispectral imageries (0.21 m) were used to test object-based (OB) feature extraction, random forest (RF) machine learning, and spectral endmember (n-D) classification methods to map PP and evaluate its spatial pattern. We trained and tested classification methods using field-collected GPS location, plant cover, and spectrometry from 288 2 m radius polygons before a prescribed burn and 480 samples after the burn within a 69.2-ha burn unit. The most accurate classification method was then used to map PP distribution and quantify abundance before and after fire. As a case study, we assessed the spatial pattern of mapped PP cover, considering topoedaphic setting and burn conditions. The results showed that the endmember classification method, spectral angle mapper (SAM), outperformed the RF and OB classifications with higher kappa coefficients (KC) (0.93 vs. 0.82 and 0.23, respectively) and overall accuracies (OA) (0.96 vs. 0.91 and 0.49) from pre-fire imagery. KC and OA metrics of post-fire imagery were lower, but rankings among classification methods were similar. SAM classifications revealed that fire reduced PP abundance by 46.5%, but reductions varied by soil type, with deeper soils having greater decreases (61%). Kolmogorov-Smirnov tests indicated significant changes before and after fire in the frequency distribution of PP cover within deeper soils (D = 0.64, p = 0.02). A two-way ANOVA revealed that the interaction of season (pre- vs. post-fire) and soils significantly (p < 0.00001) influenced the spatial pattern of PP patches. Fire also reduced the size and shape of PP patches depending on the topoedaphic settings. This study provides an innovative and effective approach for integrating field data collection, remote sensing, and endmember classification methods to map prickly pear and assess the effects of prescribed fire on prickly pear spatial patterns. Accurate mapping of PP can aid in the design and implementation of spatially explicit rangeland management strategies, such as fire, that can help reduce and mitigate the ecological and economic impacts of prickly pear expansion.

Assessment of Soybean Lodging Using UAV Imagery and Machine Learning.

Plant lodging is one of the most essential phenotypes for soybean breeding programs. Soybean lodging is conventionally evaluated visually by breeders, which is time-consuming and subject to human errors. This study aimed to investigate the potential of unmanned aerial vehicle (UAV)-based imagery and machine learning in assessing the lodging conditions of soybean breeding lines. A UAV imaging system equipped with an RGB (red-green-blue) camera was used to collect the imagery data of 1266 four-row plots in a soybean breeding field at the reproductive stage. Soybean lodging scores were visually assessed by experienced breeders, and the scores were grouped into four classes, i.e., non-lodging, moderate lodging, high lodging, and severe lodging. UAV images were stitched to build orthomosaics, and soybean plots were segmented using a grid method. Twelve image features were extracted from the collected images to assess the lodging scores of each breeding line. Four models, i.e., extreme gradient boosting (XGBoost), random forest (RF), K-nearest neighbor (KNN) and artificial neural network (ANN), were evaluated to classify soybean lodging classes. Five data preprocessing methods were used to treat the imbalanced dataset to improve classification accuracy. Results indicate that the preprocessing method SMOTE-ENN consistently performs well for all four (XGBoost, RF, KNN, and ANN) classifiers, achieving the highest overall accuracy (OA), lowest misclassification, higher F1-score, and higher Kappa coefficient. This suggests that Synthetic Minority Oversampling-Edited Nearest Neighbor (SMOTE-ENN) may be a good preprocessing method for using unbalanced datasets and the classification task. Furthermore, an overall accuracy of 96% was obtained using the SMOTE-ENN dataset and ANN classifier. The study indicated that an imagery-based classification model could be implemented in a breeding program to differentiate soybean lodging phenotype and classify lodging scores effectively.

The Similarities and Differences between Liquid-based and Conventional Methods in Evaluation of Thyroid Fine Needle Aspiration Cytologies.

Thyroid nodules are one of the most frequent medical issues in endocrinology in our country and around the world. The appropriate evaluation of the nodule is critical in the management of patients. Fine needle aspiration cytology (FNAC) is one of the most accurate tools for evaluating these nodules. Conventional and liquid-based (LB) methods are available for thyroid FNAC. In this paper, we aim to determine the best cytological method for the evaluation of thyroid fine needle aspiration materials. Fine needle aspiration (FNA) materials obtained and prepared by conventional and LB methods from 310 patients were evaluated. These slides were compared in terms of 12 cytological parameters in a three-tiered system, and the kappa coefficient was calculated. Two slides were prepared from FNA samples, and the rest of the material was left in protective solution for LB cytology. Surepath was used as a LB technique. Conventional slides (CS) were left to dry in the air and stained with May Grunwald Giemsa. Slides prepared by two different techniques were compared in 12 cytomorphological parameters, and Kappa coefficients and correlations were calculated. Kappa statistics. In comparison of LB and CS, the highest kappa coefficient was in chromatin texture (ĸ:0.738) and inflammatory cells (ĸ: 0.482). On other parameters, fair or poor agreement was observed. Although there are some superiorities to LB cytology, co-application of these two methods is more favorable in thyroid fine needle aspiration.