Improving Accuracy of Lung Nodule Classification Using Deep Learning with Focal Loss

Tran, Giang Son; Nghiem, Thi Phuong; Nguyen, Van Thi; Luong, Chi Mai; Burie, Jean-Christophe

doi:10.1155/2019/5156416

Cited by 130 publications

(55 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Para validar sus resultados [18] utiliza sensibilidad, especificidad, precisión y curva ROC. En cambio [16,17,19] calculan precisión, especificidad y sensibilidad. Por otra parte [10] calcula sensibilidad, especificidad y curva ROC.…”

Section: Validación De Resultadosunclassified

“…Son et al [19] proponen un método de aprendizaje profundo para mejorar la precisión de clasificación de nódulos pulmonares en la TC. Utilizan una red neuronal convolucional profunda 2D de 15 capas para la extracción y clasificación de nódulos y no nódulos.…”

Section: Revisión De La Literaturaunclassified

“…Acierto Clasificación de nódulos [10] Redes neuronales convoluci-94.59 % Benignos y malignos onales multivista 86.09 % Benigno, maligno primario y maligno metastásico [11] Red neuronal convolucional 81.41 % Benignos y malignos 3D y potenciación del gradiente [12] Redes neuronales y softmax 95 % Benignos y malignos [13] Red [17] Red neuronal backpropaga-91 % Benignos y malignos tion utilizando pendiente de gradiente con velocidad de aprendizaje variable [18] SVM 82 % De 2mm a 30mm de diámetro [19] Red neuronal convolucional 97.2 % Nódulos y no nódulos 2D y función de pérdida focal…”

Section: Referencia Método Utilizadounclassified

“…Mediante el aprendizaje automático se pueden detectar y predecir patrones [4], ha mostrado ser efectivo en aplicaciones delámbito médico, tales como: detección temprana de enfermedades, planificación quirúrgica, análisis de patologías, oncología y radiología. El aprendizaje automático es una de las disciplinas de la inteligencia artificial, que emplea estadística, probabilidad y técnicas de optimización, para comprender reglas complicadas y características de alta dimensión; en el tema de cáncer de pulmón se ha aplicado para detectarlo [5], clasificar la gravedad del cáncer [6], predecir la supervivencia de pacientes [7], clasificar el cáncer de pulmón [8,9] y clasificar nódulos pulmonares [10][11][12][13][14][15][16][17][18][19].…”

Section: Introductionunclassified

See 3 more Smart Citations

Estado del arte de clasificación de nódulos pulmonares utilizando Aprendizaje Automático

Mejía¹,

Biswal²,

Sánchez-DelaCruz³

2019

RCS

View full text Add to dashboard Cite

show abstract

Section: Validación De Resultadosunclassified

Section: Revisión De La Literaturaunclassified

Section: Referencia Método Utilizadounclassified

Section: Introductionunclassified

See 2 more Smart Citations

Estado del arte de clasificación de nódulos pulmonares utilizando Aprendizaje Automático

Mejía¹,

Biswal²,

Sánchez-DelaCruz³

2019

RCS

View full text Add to dashboard Cite

show abstract

“…Through the results, we presented some of the most fundamental and recent applications in the medical health system and also identified some of the challenges and opportunities of Deep Learning techniques. 26 Preprocessing methods such as wavelet denoisation were used to extract the precise contours of different tissues such as skull, cerebrospinal fluid (CSF), gray matter (GM) and white matter (WM) in five sets of images of the head MRI. Just as in (E6), a good precision was identified in the DL model, also highlighting the reduction of processing time compared to manual and semi-automatic segmentation.…”

Section: Multimodal Imaging 2019mentioning

confidence: 99%

Análise do deep learning em cuidados de saúde

Machado

Emmerick

Silva

et al. 2019

Rev enferm UFPE on line.

View full text Add to dashboard Cite

Objective: to analyze the applicability and benefits of Deep Learning in health care. Method: this is a descriptive study, reflective analysis, with articles from 2014 to 2019, published in English and peer-reviewed, in the CAPES Journal Portal, with the search equation ("Deep Learning "AND (" Health Care "OR Health-care OR Healthcare)). The results were presented in figure form followed by descriptive analysis. Results: it is revealed that 15 articles describe the applicability of Deep Learning in the health care area. This article analyzed the use of Deep Learning in different areas related to health care, highlighting the benefits found by the authors of those selected through the literature review. Conclusion: it is suggested the use of Deep Learning in health care in view of the benefits identified in the articles selected as: the prediction of disease stages; precise identification of pathological mutations and support to doctors and nurses in their daily activities. Descriptors: Benefits; Deep Learning; Health Care; Diseases; Physicians; Nurses. RESUMOObjetivo: analisar a aplicabilidade e os benefícios do Deep Learning na área de cuidados de saúde. Método: trata-se de um estudo descritivo, tipo análise reflexiva, com consulta a artigos entre os anos de 2014 a 2019, publicados em inglês e revisado por pares, no Portal de Periódicos da CAPES, com a equação de busca ("Deep Learning" AND ("Health Care" OR Health-care OR Healthcare)). Apresentaram-se os resultados em forma de figura seguida da análise descritiva. Resultados: revela-se que 15 artigos descrevem a aplicabilidade do Deep Learning na área de cuidados de saúde. Analisou-se, por este artigo, o emprego do Deep Learning em diferentes áreas referentes aos cuidados de saúde, destacando os benefícios encontrados pelos autores dos selecionados por meio da revisão de literatura. Conclusão: sugere-se o emprego do Deep Learning na área de cuidados de saúde diante dos benefícios identificados nos artigos selecionados como: a previsão dos estágios das doenças; a identificação precisa de mutações patológicas e o suporte aos médicos e aos enfermeiros em suas atividades diárias. Descritores: Benefícios; Deep Learning; Cuidados de Saúde; Doenças; Médicos; Enfermeiros. RESUMENObjetivo: analizar la aplicabilidad y los beneficios del Deep Learning en la atención médica. Método: se trata de un estudio descriptivo, tipo análisis reflexivo, con artículos de 2014 a 2019, publicados en inglés y revisados por pares, en el Portal de la revista CAPES, con la ecuación de búsqueda ("Deep Learning" Y ("Health Care" O Health-care O Healthcare)). Los resultados se presentaron en forma de figura seguida de un análisis descriptivo. Resultados: se revela que 15 artículos describen la aplicabilidad del Deep Learning en el área de la atención médica. Este artículo analizó el uso del Deep Learning en diferentes áreas relacionadas con la atención de la salud, destacando los beneficios encontrados por los autores de los seleccionados a través de la revisión de la literatura. Conclusión: s...

show abstract

Lung and colon cancer classification using hybrid principle component analysis network‐extreme learning machine

Raju

Rao

2022

Concurrency and Computation

View full text Add to dashboard Cite

Lung and colon cancers are dangerous diseases that can grow in organs and create a negative impact on human life in certain cases. The histological detection of such malignancies is one of the most critical parts of optimal treatment. As a result, the important objective of this article is to create an effective computerized diagnosis system for identifying adenocarcinomas of the colon as well as, adenocarcinomas and squamous cell carcinomas of the lungs using digital histopathology images and the combination of deep and machine learning techniques. For this, an effective optimized hybrid deep and machine learning framework is developed. This framework consists of two stages. In the first stage, the features of lung and colon images are extracted by principle component analysis network. Then the effective classification is conducted based on extreme learning machine (ELM) with the rider optimization algorithm which classifies lung and colon cancer into five types. The empirical investigation shows that the classification results on the benchmark LC25000 dataset have improved significantly. The use of this model will aid medical professionals in the development of an automatic and reliable system for detecting various forms of lung and colon cancers.

show abstract

Improving Accuracy of Lung Nodule Classification Using Deep Learning with Focal Loss

Cited by 130 publications

References 21 publications

Estado del arte de clasificación de nódulos pulmonares utilizando Aprendizaje Automático

Estado del arte de clasificación de nódulos pulmonares utilizando Aprendizaje Automático

Análise do deep learning em cuidados de saúde

Lung and colon cancer classification using hybrid principle component analysis network‐extreme learning machine

Contact Info

Product

Resources

About