Poster | 6th Internet World Congress for Biomedical Sciences |
Gustavo Camps i Valls(1), Antonio J. Serrano López(2), Jesús Modia(3), José David Martín Guerrero(4), José Vicente Francés Víllora(5)
(1)Universitat de València - Burjassot, Valencia. Spain
(2)(3)Dpto. Electrónica. Universidad de Valencia - Burjasot. Spain
(4)G.P.D.S. Departament d´Enginyeria Electrònica. Universitat de València - Burjassot. Spain
(5)Dpto. Ingenieria Electrónica. University of Valencia - Burjassot. Spain
Contact address: |
Gustavo Camps i Valls Universitat de València C/ Dr. Moliner, 50 Burjassot, Valencia Valencia Spain gustavo.camps@uv.es |
[Health Informatics] |
[Medical Electronics & Engineering] |
The use of clinical images for diagnosis is one of the fields that is acquiring an extensive range of applicability: cancer identification, histology problems or bone damage prospecting are just a few examples.
But one of the main problems in the digital processing of clinical images is the enhancement of their quality through noise suppression or reduction, which is commonly known as "image cleaning". There is a great variety of techniques dedicated to this task: from the application of the "Olympic median", hybrid methods, artificial neural networks or adaptive filtering to mathematical morphology procedures, etc. Each one of them focuses depending on the image and the kind of noise. In the present work, an effective algorithm for locating uniform random noise and the consequent application of a Selective Median Filter (SMF) is presented once the identified noise (isolated points or sequences) has been located.
The results obtained with the SMF reveal how it is possible to detect with a high level of fidelity the noise intoxication and therefore removing it without loosing substantial image quality.
Visual outcomes as soon as different error function measures tell us that the cleaning process has improved significantly the anterior methods for the set of the standard images considered.
[Health Informatics] |
[Medical Electronics & Engineering] |