karolayne Santos Azevedo

Mrs. karolayne Santos Azevedo

orcid

Research Focus & Metrics

Academic Career Length

6Years (2021-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
27 December 2024 in AI
Verified
Journal Article
Open access
PDF file
MDMatheus Dalmolin
Matheus Dalmolin
KAkarolayne Santos Azevedo
karolayne Santos Azevedo
LSLuisa Souza
Luisa Souza
CFCaroline Brunetto de Farias
Caroline Brunetto de Farias
MLMartina Lichtenfels
Martina Lichtenfels
Marcelo A. C. FernandesMarcelo A. C. Fernandes
Marcelo A. C. Fernandes
This study investigates the use of machine learning (ML) models combined with explainable artificial intelligence (XAI) techniques to identify the most influential genes in the classification of five recurrent cancer types in women: breast cancer (BRCA), lung adenocarcinoma (LUAD), thyroid cancer (THCA), ovarian cancer (OV), and colon adenocarcinoma (COAD). Gene expression data from RNA-seq, extracted from The Cancer Genome Atlas (TCGA), were used to train ML models, including decision trees (DTs), random forest (RF), and XGBoost (XGB), which achieved accuracies of 98.69%, 99.82%, and 99.37%, respectively. However, the challenges in this analysis included the high dimensionality of the dataset and the lack of transparency in the ML models. To mitigate these challenges, the SHAP (Shapley Additive Explanations) method was applied to generate a list of features, aiming to understand which characteristics influenced the models’ decision-making processes and, consequently, the prediction results for the five tumor types. The SHAP analysis identified 119, 80, and 10 genes for the RF, XGB, and DT models, respectively, totaling 209 genes, resulting in 172 unique genes. The new list, representing 0.8% of the original input features, is coherent and fully explainable, increasing confidence in the applied models. Additionally, the results suggest that the SHAP method can be effectively used as a feature selector in gene expression data. This approach not only enhances model transparency but also maintains high classification performance, highlighting its potential in identifying biologically relevant features that may serve as biomarkers for cancer diagnostics and treatment planning.

Verified
Journal Article
Open access
PDF file
LSLucileide Medeiros Dantas da Silva
Lucileide Medeiros Dantas da Silva
SSSérgio Natan Silva
Sérgio Natan Silva
LSLuisa Souza
Luisa Souza
KAkarolayne Santos Azevedo
karolayne Santos Azevedo
LGLuiz Affonso Affonso Guedes
Luiz Affonso Affonso Guedes
Marcelo A. C. FernandesMarcelo A. C. Fernandes
Marcelo A. C. Fernandes
The literature on ECG delineation algorithms has seen significant growth in recent decades. However, several challenges still need to be addressed. This work aims to propose a lightweight R-peak-detection algorithm that does not require pre-setting and performs classification on a sample-by-sample basis. The novelty of the proposed approach lies in the utilization of the typicality eccentricity detection anomaly (TEDA) algorithm for R-peak detection. The proposed method for R-peak detection consists of three phases. Firstly, the ECG signal is preprocessed by calculating the signal’s slope and applying filtering techniques. Next, the preprocessed signal is inputted into the TEDA algorithm for R-peak estimation. Finally, in the third and last step, the R-peak identification is carried out. To evaluate the effectiveness of the proposed technique, experiments were conducted on the MIT-BIH arrhythmia database (MIT-AD) for R-peak detection and validation. The results of the study demonstrated that the proposed evolutive algorithm achieved a sensitivity (Se in %), positive predictivity (+P in %), and accuracy (ACC in %) of 95.45%, 99.61%, and 95.09%, respectively, with a tolerance (TOL) of 100 milliseconds. One key advantage of the proposed technique is its low computational complexity, as it is based on a statistical framework calculated recursively. It employs the concepts of typicity and eccentricity to determine whether a given sample is normal or abnormal within the dataset. Unlike most traditional methods, it does not require signal buffering or windowing. Furthermore, the proposed technique employs simple decision rules rather than heuristic approaches, further contributing to its computational efficiency.

21 July 2024
Verified
Conference Paper
Open access
PDF file
CTClara Freire M. Teixeira
Clara Freire M. Teixeira
MOMaria Isabel M. Oliveira
Maria Isabel M. Oliveira
MSMaria Laura B. da Silva
Maria Laura B. da Silva
ACAlba Sandyra B. L. Campos
Alba Sandyra B. L. Campos
KAkarolayne Santos Azevedo
karolayne Santos Azevedo
Estudos apontam que meninas se interessam menos pelas áreas STEM (do inglês, Science, Technology, Engineering, and Mathematics), não por incapacidade cognitiva, mas por questões políticas, pela falta de estrutura educacional, social, cultural e principalmente devido a estereótipos. Este artigo apresenta ações do Projeto Ada, um projeto de extensão que visa promover a participação feminina nas áreas STEM. As atividades propostas incluem rodas de conversas, exibições de filmes temáticos e oficinas de desenvolvimento de jogos digitais com o objetivo de conscientizar sobre a baixa representatividade das mulheres nessas áreas.

1 January 2021
Verified
Conference Paper
Open access
KAkarolayne Santos Azevedo
karolayne Santos Azevedo
RBRaquel de Melo Barbosa
Raquel de Melo Barbosa
Marcelo A. C. FernandesMarcelo A. C. Fernandes
Marcelo A. C. Fernandes
Título: Aprendizagem Profunda Aplicada à Classificação do SARS-CoV-2 Autores: Karolayne Azevedo, Raquel Barbosa and Marcelo Fernandes. Resumo: Este artigo propõe uma técnica, baseada em aprendizado de máquina, que faz uso de uma rede neural convolucional (Convolutional Neural Network – CNN) Read More ...

29 September 2025
Verified
Conference Paper
Open access
PDF file
ASAlessandro S. Silva
Alessandro S. Silva
KAkarolayne Santos Azevedo
karolayne Santos Azevedo
Marcelo A. C. FernandesMarcelo A. C. Fernandes
Marcelo A. C. Fernandes
Este trabalho propõe uma abordagem supervisionada para a classificação de variantes do SARS-CoV-2 (VOCs) a partir de embeddings vetoriais derivados de substituições de aminoácidos. Utilizando o modelo pré-treinado all-MiniLM-L6-v2, foram gerados vetores de alta dimensionalidade que codificam mutações sem necessidade de alinhamento genômico. Esses embeddings alimentaram classificadores como SVM, Random Forest, XGBoost e k-NN, avaliados por validação cruzada e em um teste externo com quase 288 mil amostras. O XGBoost obteve os melhores resultados, com acurácia de 99,83%, F1 macro de 99,83% e logLoss de 0,0068, mantendo desempenho elevado mesmo em dados não vistos. Os resultados evidenciam que modelos baseados em árvores superam alternativas como SVM, especialmente na variante Gamma. A proposta se mostra robusta, precisa e escalável para aplicação em sistemas automatizados de vigilância genômica, fornecendo uma segunda perspectiva de análise complementar à abordagem clínica e laboratorial, e superando métodos tradicionais e híbridas da literatura recente.