RESUMEN

Background  Dental caries is one of the most prevalent conditions in dentistry worldwide. Early identification and classification of dental caries are essential for effective prevention and treatment. Panoramic dental radiographs are commonly used to screen for overall oral health, including dental caries and tooth anomalies. However, manual interpretation of these radiographs can be time-consuming and prone to human error. Therefore, an automated classification system could help streamline diagnostic workflows and provide timely insights for clinicians. Methods This article presents a deep learning-based, custom-built model for the binary classification of panoramic dental radiographs. The use of histogram equalization and filtering methods as preprocessing techniques effectively addresses issues related to irregular illumination and contrast in dental radiographs, enhancing overall image quality. By incorporating three separate panoramic dental radiograph datasets, the model benefits from a diverse dataset that improves its training and evaluation process across a wide range of caries and abnormalities. Results The dental radiograph analysis model is designed for binary classification to detect the presence of dental caries, restorations, and periapical region abnormalities, achieving accuracies of 97.01%, 81.63%, and 77.53%, respectively. Conclusions The proposed algorithm extracts discriminative features from dental radiographs, detecting subtle patterns indicative of tooth caries, restorations, and region-based abnormalities. Automating this classification could assist dentists in the early detection of caries and anomalies, aid in treatment planning, and enhance the monitoring of dental diseases, ultimately improving and promoting patients' oral healthcare.

RESUMEN

The integrity of the reconstructed human epidermis generated in vitro can be assessed using histological analyses combined with immunohistochemical staining of keratinocyte differentiation markers. Technical differences during the preparation and capture of stained images may influence the outcome of computational methods. Due to the specific nature of the analyzed material, no annotated datasets or dedicated methods are publicly available. Using a dataset with 598 unannotated images showing cross-sections of in vitro reconstructed human epidermis stained with DAB-based immunohistochemistry reaction to visualize four different keratinocyte differentiation marker proteins (filaggrin, keratin 10, Ki67, HSPA2) and counterstained with hematoxylin, we developed an unsupervised method for the detection and quantification of immunohistochemical staining. The pipeline consists of the following steps: (i) color normalization; (ii) color deconvolution; (iii) morphological operations; (iv) automatic image rotation; and (v) clustering. The most effective combination of methods includes (i) Reinhard's normalization; (ii) Ruifrok and Johnston color-deconvolution method; (iii) proposed image-rotation method based on boundary distribution of image intensity; and (iv) k-means clustering. The results of the work should enhance the performance of quantitative analyses of protein markers in reconstructed human epidermis samples and enable the comparison of their spatial distribution between different experimental conditions.

RESUMEN

Swahili corpus is a dataset generated by collecting written Kiswahili sentences from different sectors that deals with Kiswahili documents. Corpus of intended language is needed in Natural Language Processing (NLP) task to fit algorithm in order to understand that language before training the model. Swahili corpus dataset generated contained 1,693,228 sentences with 39,639,824 words and 871,452 unique words. Corpus exported in text file format with storage size of 168 MB. These sentences collected from different sources in different categories as follows: - Health (AFYA), Business and Industries (BIASHARA), Parliament (BUNGE), Religion (DINI), Education (ELIMU), News (HABARI), Agriculture (KILIMO), Social Media (MITANDAO), Non-Governmental Organizations (MASHIRIKA YA KIRAIA), Government (SERIKALI), Laws (SHERIA) and Politics (SIASA). This abstract outlines the systematic data collection process employed for the creation of a Swahili corpus derived from multiple public websites and reports. The compilation of this corpus involves a meticulous and comprehensive approach to ensure the representation of diverse linguistic contexts and topics relevant to the Swahili language. The data collection process commenced with the identification of suitable sources across various domains, including news articles, health publications, online forums, and Governmental public reports. Websites and platforms with publicly available Swahili content were systematically crawled and archived to capture a broad spectrum of linguistic expressions. Furthermore, special attention was given to reputable sources to maintain the authenticity of the corpus and linguistic richness. The inclusion of diverse sources ensures that the corpus reflects the linguistic nuances inherent in different contexts and registers within the Swahili language. Additionally, efforts were made to incorporate variations in domain dialects, acknowledging the linguistic diversity present in Swahili. The potential for reusing this Swahili corpus is vast. Researchers, linguists, and language enthusiasts can leverage the diverse and extensive dataset for a multitude of applications, including NLP tasks such as sentiment analysis, textual data clustering, classifications tasks and machine translation. The Corpus can serve as training data for developing and evaluating NLP algorithms, including part-of-speech tagging, and named entity recognition. Also, text mining techniques can be applied to corpus and enable researchers to extract valuable insights, identify patterns, and discover knowledge from large textual datasets.

RESUMEN

BACKGROUND: Spectral data from multiple sources can be integrated into multi-block fusion chemometric models, such as sequentially orthogonalized partial-least squares (SO-PLS), to improve the prediction of sample quality features. Pre-processing techniques are often applied to mitigate extraneous variability, unrelated to the response variables. However, the selection of suitable pre-processing methods and identification of informative data blocks becomes increasingly complex and time-consuming when dealing with a large number of blocks. The problem addressed in this work is the efficient pre-processing, selection, and ordering of data blocks for targeted applications in SO-PLS. RESULTS: We introduce the PROSAC-SO-PLS methodology, which employs pre-processing ensembles with response-oriented sequential alternation calibration (PROSAC). This approach identifies the best pre-processed data blocks and their sequential order for specific SO-PLS applications. The method uses a stepwise forward selection strategy, facilitated by the rapid Gram-Schmidt process, to prioritize blocks based on their effectiveness in minimizing prediction error, as indicated by the lowest prediction residuals. To validate the efficacy of our approach, we showcase the outcomes of three empirical near-infrared (NIR) datasets. Comparative analyses were performed against partial-least-squares (PLS) regressions on single-block pre-processed datasets and a methodology relying solely on PROSAC. The PROSAC-SO-PLS approach consistently outperformed these methods, yielding significantly lower prediction errors. This has been evidenced by a reduction in the root-mean-squared error of prediction (RMSEP) ranging from 5 to 25 % across seven out of the eight response variables analyzed. SIGNIFICANCE: The PROSAC-SO-PLS methodology offers a versatile and efficient technique for ensemble pre-processing in NIR data modeling. It enables the use of SO-PLS minimizing concerns about pre-processing sequence or block order and effectively manages a large number of data blocks. This innovation significantly streamlines the data pre-processing and model-building processes, enhancing the accuracy and efficiency of chemometric models.

RESUMEN

The complexity of the cancer problem domain presents challenges not only to the medical analysis systems tasked with its analysis, but also to the users of such systems. While it is desirable to assist users in operating these medical analysis systems, prior groundwork is required before this can be achieved, such as recognising patterns in the way users create certain analyses within these systems. In this paper, we use machine learning algorithms to analyse user behaviour patterns and attempt to predict the next user interaction within the CARESS medical analysis system. Since an appropriate pre-processing scheme is essential for the performance of these algorithms, we propose the usage of a Natural Language Processing (NLP)- inspired approach to preserve some semantic cohesion of the mostly categorical features of these user interactions. Furthermore, we propose to use a sliding window that contains information about the latest user interactions in combination with Latent Dirichlet Allocation (LDA) to extract a latent topic from these last interactions and use it as additional input to the machine learning models. We compare this pre-processing scheme with other approaches that utilise one-hot encoding and feature hashing. The results of our experiments show that the sliding window LDA scheme is a promising solution, that performs better for our use case than the other evaluated pre-processing schemes. Overall, our results provide an important piece for further research and development in the area of assisting users in operating analysis systems in complex problem domains.

Asunto(s)

RESUMEN

Lung cancer (LC) continues to be a leading death cause in China, primarily due to late diagnosis. This study aimed to evaluate the effectiveness of using plasma-based near-infrared spectroscopy (NIRS) for LC early diagnosis. A total of 171 plasma samples were collected, including 73 healthy controls (HC), 73 LC, and 25 benign lung tumors (B). NIRS was utilized to measure the spectra of samples. Pre-processing methods, including centering and scaling, standard normal variate, multiplicative scatter correction, Savitzky-Golay smoothing, Savitzky-Golay first derivative, and baseline correction were applied. Subsequently, 4 machine learning (ML) algorithms, including partial least squares (PLS), support vector machines (SVM), gradient boosting machine, and random forest, were utilized to develop diagnostic models using train set data. Then, the predictive performance of each model was evaluated using test set samples. The study was conducted in 5 comparisons as follows: LC and HC, LC and B, B and HC, the diseased group (D) and HC, as well as LC, B and HC. Among the 5 comparisons, SVM consistently generated the best performance with a certain pre-processing method, achieving overall accuracy of 1.0 (kappa: 1.0) in the comparisons of LC and HC, B and HC, as well as D and HC. Pre-processing was identified as a crucial step in developing ML models. Interestingly, PLS demonstrated remarkable stability and relatively high predictive performance across the 5 comparisons, even though it did not achieve the top results like SVM. However, none of these algorithms were able to effectively distinguish B from LC. These findings indicate that the combination of plasma-based NIRS with ML algorithms is a rapid, non-invasive, effective, and economical method for LC early diagnosis.

Asunto(s)

RESUMEN

Analysis of the change in groundwater used as a drinking and irrigation water source is of critical importance in terms of monitoring aquifers, planning water resources, energy production, combating climate change, and agricultural production. Therefore, it is necessary to model groundwater level (GWL) fluctuations to monitor and predict groundwater storage. Artificial intelligence-based models in water resource management have become prevalent due to their proven success in hydrological studies. This study proposed a hybrid model that combines the artificial neural network (ANN) and the artificial bee colony optimization (ABC) algorithm, along with the ensemble empirical mode decomposition (EEMD) and the local mean decomposition (LMD) techniques, to model groundwater levels in Erzurum province, Türkiye. GWL estimation results were evaluated with mean square error (MSE), coefficient of determination (R2), and residual sum of squares (RSS) and visually with violin, scatter, and time series plot. The study results indicated that the EEMD-ABC-ANN hybrid model was superior to other models in estimating GWL, with R2 values ranging from 0.91 to 0.99 and MSE values ranging from 0.004 to 0.07. It has also been revealed that promising GWL predictions can be made with previous GWL data.

Asunto(s)

RESUMEN

Understanding the origins of sediment within stream networks is critical to developing effective strategies to mitigate sediment delivery and soil erosion in larger drainage basins. Sediment fingerprinting is a widely accepted approach to identifying sediment sources; however, it typically relies on labor-intensive and costly chemical analyses. Recent studies have recognized diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) as a non-destructive, cost-effective, and efficient alternative for estimating sediment contributions from multiple sources. This study aimed to assess (i) the effects of different particle size fractions on DRIFTS and conservatism tests, (ii) the effects of spectral pre-processing on discriminating sub-catchment spatial sediment sources, (iii) the efficiency of partial least squares regression (PLSR) and support vector machine regression (SVMR) chemometric models across different spectral resolutions and particle size fractions, and (iv) the quantification of sub-catchment spatial sediment source contributions using chemometric models across different particle size fractions. DRIFTS analysis was performed on three particle size fractions (<38 µm, 38-63 µm, and 63-125 µm) using 54 sediment samples from three different sub-catchments and 26 target sediment samples from the Andajrood catchment in Iran. Results showed significant effects of particle size fractions on DRIFTS for both sub-catchment sediment sources and target sediment samples. Conservatism tests indicated that DRIFTS behave conservative for the majority of target sediment samples. Spectral pre-processing techniques including SNV + SGD1 and SGD1 effectively discriminated sources across all particle size fractions and spectral resolutions. However, the optimal combination of pre-processing, spectral resolution, and regression models varied between sub-fractions. Validated model estimates revealed that sub-catchment 1 consistently contributed the most sediment across all particle size fractions, followed by sub-catchments 3 and 2. These results highlight the effectiveness of DRIFTS as a rapid, cost-effective, and precise method for discriminating and apportioning sediment sources within spatial sub-catchments.

RESUMEN

Neuromarketing is an emerging research field that aims to understand consumers' decision-making processes when choosing which product to buy. This information is highly sought after by businesses looking to improve their marketing strategies by understanding what leaves a positive or negative impression on consumers. It has the potential to revolutionize the marketing industry by enabling companies to offer engaging experiences, create more effective advertisements, avoid the wrong marketing strategies, and ultimately save millions of dollars for businesses. Therefore, good documentation is necessary to capture the current research situation in this vital sector. In this article, we present a systematic review of EEG-based Neuromarketing. We aim to shed light on the research trends, technical scopes, and potential opportunities in this field. We reviewed recent publications from valid databases and divided the popular research topics in Neuromarketing into five clusters to present the current research trend in this field. We also discuss the brain regions that are activated when making purchase decisions and their relevance to Neuromarketing applications. The article provides appropriate illustrations of marketing stimuli that can elicit authentic impressions from consumers' minds, the techniques used to process and analyze recorded brain data, and the current strategies employed to interpret the data. Finally, we offer recommendations to upcoming researchers to help them investigate the possibilities in this area more efficiently in the future.

RESUMEN

Electroencephalography (EEG) evaluation is an important step in the clinical diagnosis of brain death during the standard clinical procedure. The processing of the brain-death EEG signals acquisition always carried out in the Intensive Care Unit (ICU). The electromagnetic environmental noise and prescribed sedative may erroneously suggest cerebral electrical activity, thus effecting the presentation of EEG signals. In order to accurately and efficiently assist physicians in making correct judgments, this paper presents a band-pass filter and threshold rejection-based EEG signal pre-processing method and an EEG-based coma/brain-death classification system associated with One Dimensional Convolutional Neural Network (1D-CNN) model to classify informative brain activity features from real-world recorded clinical EEG data. The experimental result shows that our method is well performed in classify the coma patients and brain-death patients with the classification accuracy of 99.71%, F1-score of 99.71% and recall score of 99.51%, which means the proposed model is well performed in the coma/brain-death EEG signals classification task. This paper provides a more straightforward and effective method for pre-processing and classifying EEG signals from coma/brain-death patients, and demonstrates the validity and reliability of the method. Considering the specificity of the condition and the complexity of the EEG acquisition environment, it presents an effective method for pre-processing real-time EEG signals in clinical diagnoses and aiding the physicians in their diagnosis, with significant implications for the choice of signal pre-processing methods in the construction of practical brain-death identification systems.

RESUMEN

Studies of intracranial EEG networks have been used to reveal seizure generators in patients with drug-resistant epilepsy. Intracranial EEG is implanted to capture the epileptic network, the collection of brain tissue that forms a substrate for seizures to start and spread. Interictal intracranial EEG measures brain activity at baseline, and networks computed during this state can reveal aberrant brain tissue without requiring seizure recordings. Intracranial EEG network analyses require choosing a reference and applying statistical measures of functional connectivity. Approaches to these technical choices vary widely across studies, and the impact of these technical choices on downstream analyses is poorly understood. Our objective was to examine the effects of different re-referencing and connectivity approaches on connectivity results and on the ability to lateralize the seizure onset zone in patients with drug-resistant epilepsy. We applied 48 pre-processing pipelines to a cohort of 125 patients with drug-resistant epilepsy recorded with interictal intracranial EEG across two epilepsy centres to generate intracranial EEG functional connectivity networks. Twenty-four functional connectivity measures across time and frequency domains were applied in combination with common average re-referencing or bipolar re-referencing. We applied an unsupervised clustering algorithm to identify groups of pre-processing pipelines. We subjected each pre-processing approach to three quality tests: (i) the introduction of spurious correlations; (ii) robustness to incomplete spatial sampling; and (iii) the ability to lateralize the clinician-defined seizure onset zone. Three groups of similar pre-processing pipelines emerged: common average re-referencing pipelines, bipolar re-referencing pipelines and relative entropy-based connectivity pipelines. Relative entropy and common average re-referencing networks were more robust to incomplete electrode sampling than bipolar re-referencing and other connectivity methods (Friedman test, Dunn-Sidák test P < 0.0001). Bipolar re-referencing reduced spurious correlations at non-adjacent channels better than common average re-referencing (Δ mean from machine ref = -0.36 versus -0.22) and worse in adjacent channels (Δ mean from machine ref = -0.14 versus -0.40). Relative entropy-based network measures lateralized the seizure onset hemisphere better than other measures in patients with temporal lobe epilepsy (Benjamini-Hochberg-corrected P < 0.05, Cohen's d: 0.60-0.76). Finally, we present an interface where users can rapidly evaluate intracranial EEG pre-processing choices to select the optimal pre-processing methods tailored to specific research questions. The choice of pre-processing methods affects downstream network analyses. Choosing a single method among highly correlated approaches can reduce redundancy in processing. Relative entropy outperforms other connectivity methods in multiple quality tests. We present a method and interface for researchers to optimize their pre-processing methods for deriving intracranial EEG brain networks.

RESUMEN

PURPOSE: To determine the significance of complex-valued inputs and complex-valued convolutions compared to real-valued inputs and real-valued convolutions in convolutional neural networks (CNNs) for frequency and phase correction (FPC) of GABA-edited magnetic resonance spectroscopy (MRS) data. METHODS: An ablation study using simulated data was performed to determine the most effective input (real or complex) and convolution type (real or complex) to predict frequency and phase shifts in GABA-edited MEGA-PRESS data using CNNs. The best CNN model was subsequently compared using both simulated and in vivo data to two recently proposed deep learning (DL) methods for FPC of GABA-edited MRS. All methods were trained using the same experimental setup and evaluated using the signal-to-noise ratio (SNR) and linewidth of the GABA peak, choline artifact, and by visually assessing the reconstructed final difference spectrum. Statistical significance was assessed using the Wilcoxon signed rank test. RESULTS: The ablation study showed that using complex values for the input represented by real and imaginary channels in our model input tensor, with complex convolutions was most effective for FPC. Overall, in the comparative study using simulated data, our CC-CNN model (that received complex-valued inputs with complex convolutions) outperformed the other models as evaluated by the mean absolute error. CONCLUSION: Our results indicate that the optimal CNN configuration for GABA-edited MRS FPC uses a complex-valued input and complex convolutions. Overall, this model outperformed existing DL models.

Asunto(s)

RESUMEN

With high-frequency data of nitrate (NO3-N) concentrations in waters becoming increasingly important for understanding of watershed system behaviors and ecosystem managements, the accurate and economic acquisition of high-frequency NO3-N concentration data has become a key point. This study attempted to use coupled deep learning neural networks and routine monitored data to predict hourly NO3-N concentrations in a river. The hourly NO3-N concentration at the outlet of the Oyster River watershed in New Hampshire, USA, was predicted through neural networks with a hybrid model architecture coupling the Convolutional Neural Networks and the Long Short-Term Memory model (CNN-LSTM). The routine monitored data (the river depth, water temperature, air temperature, precipitation, specific conductivity, pH and dissolved oxygen concentrations) for model training were collected from a nested high-frequency monitoring network, while the high-frequency NO3-N concentration data obtained at the outlet were not included as inputs. The whole dataset was separated into training, validation, and testing processes according to the ratio of 5:3:2, respectively. The hybrid CNN-LSTM model with different input lengths (1d, 3d, 7d, 15d, 30d) displayed comparable even better performance than other studies with lower frequencies, showing mean values of the Nash-Sutcliffe Efficiency 0.60-0.83. Models with shorter input lengths demonstrated both the higher modeling accuracy and stability. The water level, water temperature and pH values at monitoring sites were main controlling factors for forecasting performances. This study provided a new insight of using deep learning networks with a coupled architecture and routine monitored data for high-frequency riverine NO3-N concentration forecasting and suggestions about strategies about variable and input length selection during preprocessing of input data.

Asunto(s)

RESUMEN

As the number of electronic gadgets in our daily lives is increasing and most of them require some kind of human interaction, this demands innovative, convenient input methods. There are limitations to state-of-the-art (SotA) ultrasound-based hand gesture recognition (HGR) systems in terms of robustness and accuracy. This research presents a novel machine learning (ML)-based end-to-end solution for hand gesture recognition with low-cost micro-electromechanical (MEMS) system ultrasonic transducers. In contrast to prior methods, our ML model processes the raw echo samples directly instead of using pre-processed data. Consequently, the processing flow presented in this work leaves it to the ML model to extract the important information from the echo data. The success of this approach is demonstrated as follows. Four MEMS ultrasonic transducers are placed in three different geometrical arrangements. For each arrangement, different types of ML models are optimized and benchmarked on datasets acquired with the presented custom hardware (HW): convolutional neural networks (CNNs), gated recurrent units (GRUs), long short-term memory (LSTM), vision transformer (ViT), and cross-attention multi-scale vision transformer (CrossViT). The three last-mentioned ML models reached more than 88% accuracy. The most important innovation described in this research paper is that we were able to demonstrate that little pre-processing is necessary to obtain high accuracy in ultrasonic HGR for several arrangements of cost-effective and low-power MEMS ultrasonic transducer arrays. Even the computationally intensive Fourier transform can be omitted. The presented approach is further compared to HGR systems using other sensor types such as vision, WiFi, radar, and state-of-the-art ultrasound-based HGR systems. Direct processing of the sensor signals by a compact model makes ultrasonic hand gesture recognition a true low-cost and power-efficient input method.

Asunto(s)

RESUMEN

Lung diseases are the third-leading cause of mortality in the world. Due to compromised lung function, respiratory difficulties, and physiological complications, lung disease brought on by toxic substances, pollution, infections, or smoking results in millions of deaths every year. Chest X-ray images pose a challenge for classification due to their visual similarity, leading to confusion among radiologists. To imitate those issues, we created an automated system with a large data hub that contains 17 datasets of chest X-ray images for a total of 71,096, and we aim to classify ten different disease classes. For combining various resources, our large datasets contain noise and annotations, class imbalances, data redundancy, etc. We conducted several image pre-processing techniques to eliminate noise and artifacts from images, such as resizing, de-annotation, CLAHE, and filtering. The elastic deformation augmentation technique also generates a balanced dataset. Then, we developed DeepChestGNN, a novel medical image classification model utilizing a deep convolutional neural network (DCNN) to extract 100 significant deep features indicative of various lung diseases. This model, incorporating Batch Normalization, MaxPooling, and Dropout layers, achieved a remarkable 99.74% accuracy in extensive trials. By combining graph neural networks (GNNs) with feedforward layers, the architecture is very flexible when it comes to working with graph data for accurate lung disease classification. This study highlights the significant impact of combining advanced research with clinical application potential in diagnosing lung diseases, providing an optimal framework for precise and efficient disease identification and classification.

Asunto(s)

RESUMEN

In this study, in order to realize the sharing of the near-infrared analysis model of holocellulose between three spectral instruments of the same type, 84 pulp samples and their content of holocellulose were taken as the research objects. The effects of 10 pre-processing methods, such as 1st derivative (D1st), 2nd derivative (D2nd), multiplicative scatter correction (MSC), standard normal variable transformation (SNV), autoscaling, normalization, mean centering and pairwise combination, on the transfer effect of the stable wavelength selected by screening wavelengths with consistent and stable signals (SWCSS) were discussed. The results showed that the model established by the wavelength selected by the SWCSS algorithm after the autoscaling pre-processing method had the best analysis effect on the two target samples. Root mean square error of prediction (RMSEP) decreased from 2.4769 and 2.3119 before the model transfer to 1.2563 and 1.2384, respectively. Compared with the full-spectrum model, the value of AIC decreased from 3209.83 to 942.82. Therefore, the autoscaling pre-processing method combined with SWCSS algorithm can significantly improve the accuracy and efficiency of model transfer and provide help for the application of SWCSS algorithm in the rapid determination of pulp properties by near-infrared spectroscopy (NIRS).

RESUMEN

This study worked with Chunghwa Telecom to collect data from 17 rooftop solar photovoltaic plants installed on top of office buildings, warehouses, and computer rooms in northern, central and southern Taiwan from January 2021 to June 2023. A data pre-processing method combining linear regression and K Nearest Neighbor (k-NN) was proposed to estimate missing values for weather and power generation data. Outliers were processed using historical data and parameters highly correlated with power generation volumes were used to train an artificial intelligence (AI) model. To verify the reliability of this data pre-processing method, this study developed multilayer perceptron (MLP) and long short-term memory (LSTM) models to make short-term and medium-term power generation forecasts for the 17 solar photovoltaic plants. Study results showed that the proposed data pre-processing method reduced normalized root mean square error (nRMSE) for short- and medium-term forecasts in the MLP model by 17.47% and 11.06%, respectively, and also reduced the nRMSE for short- and medium-term forecasts in the LSTM model by 20.20% and 8.03%, respectively.

RESUMEN

This study introduces "Specialis Revelio," a sophisticated text pre-processing module aimed at enhancing the detection of disguised toxic content in online communications. Through a blend of conventional and novel pre-processing methods, this module significantly improves the accuracy of existing toxic text detection tools, addressing the challenge of content that is deliberately altered to evade standard detection methods.•Integration with Existing Systems: "Specialis Revelio" is designed to augment popular toxic text classifiers, enhancing their ability to detect and filter toxic content more effectively.•Innovative Pre-processing Methods: The module combines traditional pre-processing steps like lowercasing and stemming with advanced strategies, including the handling of adversarial examples and typo correction, to reveal concealed toxicity.•Validation through Comparative Study: Its effectiveness was validated via a comparative analysis against widely used APIs, demonstrating a marked improvement in the detection of various toxic text indicators.

RESUMEN

Air pollution stands as a significant modern-day challenge impacting life quality, the environment, and the economy. It comprises various pollutants like gases, particulate matter, biological molecules, and more, stemming from sources such as vehicle emissions, industrial operations, agriculture, and natural events. Nitrogen dioxide (NO2), among these harmful gases, is notably prevalent in densely populated urban regions. Given its adverse effects on health and the environment, accurate monitoring of NO2 levels becomes imperative for devising effective risk mitigation strategies. However, the precise measurement of NO2 poses challenges as it traditionally relies on costly and bulky equipment. This has prompted the development of more affordable alternatives, although their reliability is often questionable. The aim of this article is to introduce a groundbreaking method for precisely calibrating cost-effective NO2 sensors. This technique involves statistical preprocessing of low-cost sensor readings, aligning their distribution with reference data. Central to this calibration is an artificial neural network (ANN) surrogate designed to predict sensor correction coefficients. It utilizes environmental variables (temperature, humidity, atmospheric pressure), cross-references auxiliary NO2 sensors, and incorporates short time series of previous readings from the primary sensor. These methods are complemented by global data scaling. Demonstrated using a custom-designed cost-effective monitoring platform and high-precision public reference station data collected over 5 months, every component of our calibration framework proves crucial, contributing to its exceptional accuracy (with a correlation coefficient near 0.95 concerning the reference data and an RMSE below 2.4 µg/m3). This level of performance positions the calibrated sensor as a viable, cost-effective alternative to traditional monitoring approaches.