Murilo Gazzola

Dr. Murilo Gazzola


Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

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

Representative Publications

View All Publications
1 January 2025 in Healthcare
Verified
Journal Article
Open access
PDF file
LBLarissa Cristina Berti
Larissa Cristina Berti
MGMarcelo Matheus Gauy
Marcelo Matheus Gauy
LSLuana Cristina Santos da Silva
Luana Cristina Santos da Silva
JRJulia Vasquez Valenci Rios
Julia Vasquez Valenci Rios
VMViviam Batista Morais
Viviam Batista Morais
MFMarcelo Finger
Marcelo Finger
Background/Objectives: The aim of this paper was to compare voice and speech characteristics between post-COVID-19 and control subjects. The hypothesis was that acoustic parameters of voice and speech may differentiate subjects infected by COVID-19 from control subjects. Additionally, we expected to observe the persistence of symptoms in women. Methods: In total, 134 subjects participated in the study, were selected for convenience and divided into two groups: 70 control subjects and 64 post-COVID-19 subjects, with an average time of 8.7 months after infection. The recordings were made using the SPIRA software (v.1.0.) on cell phones, based on three verbal tasks: sustained production of the vowel/a/, reading a sentence, and producing a rhyme. Acoustic analyses of speech and voice were carried out with the PRAAT software (v.4.3.18), based on the following parameters: total sentence duration, number of pauses, pause duration, f0, f0SD, jitter, shimmer, and harmonics-to-noise ratio (HNR). Results: Regarding the acoustic characteristics of speech, there were no differences between the groups or between the sexes. Regarding the acoustic characteristics of voice, jitter, shimmer, and HNR, significant differences between the groups were found. Differences between sexes were observed in the following frequency-related parameters: f0, f0SD, and jitter. Conclusions: Some acoustic characteristics of the patients’ voice may show a deteriorated condition even after exacerbation of the disease. These characteristics are compatible with some of the symptoms reported by post-COVID-19 subjects, such as the presence of tension and fatigue. These voice acoustic parameters could be used as biomarkers to screen voice disorders in long-COVID, using artificial intelligence (AI), accelerating the search for diagnosis by specialists.

10 January 2022 in Language Resources and Evaluation
Verified
Journal Article
Closed access
PDF file
MGMurilo Gazzola
Murilo Gazzola
SLSidney Leal
Sidney Leal
BPBreno Pedroni
Breno Pedroni
FRFabio Rocha
Fabio Rocha
SPSabine Pompéia
Sabine Pompéia
SASandra Maria Aluísio
Sandra Maria Aluísio

1 January 2024 in Radiologia Brasileira
Verified
Journal Article
Open access
PDF file
LSLislie Gabriela Santin
Lislie Gabriela Santin
HLHenrique Min Ho Lee
Henrique Min Ho Lee
VSViviane Mariano da Silva
Viviane Mariano da Silva
ECEllison Fernando Cardoso
Ellison Fernando Cardoso
MGMurilo Gazzola
Murilo Gazzola
Resumo Objetivo: Desenvolver uma aplicação de processamento de linguagem natural capaz de identificar automaticamente doenças cirúrgicas benignas da vesícula biliar a partir de laudos radiológicos. Materiais e Métodos: Desenvolvemos um classificador de texto para classificar laudos como contendo ou não doenças cirúrgicas benignas da vesícula biliar. Selecionamos aleatoriamente 1.200 laudos com descrição da vesícula biliar de nosso banco de dados, incluindo diferentes modalidades. Quatro radiologistas classificaram os laudos como doença benigna cirúrgica ou não. Duas arquiteturas de aprendizagem profunda foram treinadas para a classificação: a rede neural convolucional (convolutional neural network - CNN) e a memória longa de curto prazo bidirecional (bidirectional long short-term memory - BiLSTM). Para representar palavras de forma vetorial, os modelos incluíram uma representação Word2Vec, com dimensões variando de 300 a 1000. Os modelos foram treinados e avaliados por meio da divisão do conjunto de dados entre treinamento, validação e teste (80/10/10). Resultados: CNN e BiLSTM tiveram bom desempenho em ambos os espaços dimensionais. Relatamos para 300 e 1000 dimensões, respectivamente, as pontuações F1 de 0,95945 e 0,95302 para o modelo CNN e de 0,96732 e 0,96732 para a BiLSTM. Conclusão: Nossos modelos alcançaram alto desempenho, independentemente de diferentes arquiteturas e espaços dimensionais.

1 January 2024 in Radiologia Brasileira
Verified
Journal Article
Open access
PDF file
LSLislie Gabriela Santin
Lislie Gabriela Santin
HLHenrique Min Ho Lee
Henrique Min Ho Lee
VSViviane Mariano da Silva
Viviane Mariano da Silva
ECEllison Fernando Cardoso
Ellison Fernando Cardoso
MGMurilo Gazzola
Murilo Gazzola
Objective: To develop a natural language processing application capable of automatically identifying benign gallbladder diseases that require surgery, from radiology reports.