ADP e MLP como Métodos de Detecção Automática de Vias Aéreas em Imagens 2D
##manager.scheduler.building##: Palacio de Convenciones de La Habana
##manager.scheduler.room##: Sala 12
Fecha: 2011-05-20 09:00 – 10:30
Última modificación: 2011-04-29 02:09
Resumen
Several information present in medical images
can be obtained by means of computer vision. Lung diseases
are among the leading causes of death in the world. The main
diseases of this type are asthma, bronchiectasis, chronic bronchitis,
emphysema and COPD. In this sense is importance to
achieve and determine early diagnosis and correct to detect the
actual stages of lung diseases, in which there is a big influence
of the status of lung airways. Therefore, the identification
correct airway is a crucial step. This paper develops and
evaluates two algorithms that is based on a neural network
MLP (Multi Layer Perceptron) and another in ALD
(analysis of lung densities) to automatically detect the airways.
Results algorithm for each image are compared with the
airways marked by agreement in three medical experts.
By Finally, the results are presented for the automatic
location airway with success rates greater than 75%.
can be obtained by means of computer vision. Lung diseases
are among the leading causes of death in the world. The main
diseases of this type are asthma, bronchiectasis, chronic bronchitis,
emphysema and COPD. In this sense is importance to
achieve and determine early diagnosis and correct to detect the
actual stages of lung diseases, in which there is a big influence
of the status of lung airways. Therefore, the identification
correct airway is a crucial step. This paper develops and
evaluates two algorithms that is based on a neural network
MLP (Multi Layer Perceptron) and another in ALD
(analysis of lung densities) to automatically detect the airways.
Results algorithm for each image are compared with the
airways marked by agreement in three medical experts.
By Finally, the results are presented for the automatic
location airway with success rates greater than 75%.