TITLE: A Probabilistic Neural Network Framework for Detection of Malignant Melanoma
AUTHORS: M. Hintz-Madsen, L.K. Hansen, J. Larsen and K.T. Drzewiecki
Department of Mathematical Modelling, Building 321
Technical University of Denmark, DK-2800 Lyngby, Denmark
emails: lkhansen,jl,thko@imm.dtu.dk
www: http://eivind.imm.dtu.dk
Dept. of Plastic Surgery S, National University Hospital,
Blegdamsvej 9, DK-2100 Copenhagen, Denmark
ABSTRACT:
The work reported in this chapter
concerns the classification of dermatoscopic images of skin
lesions. The overarching goals of the work are:
- Develop an objective and
cost-efficient tool for classification of skin lesions.
This involves extracting relevant information from dermatoscopic
images in the form of dermatoscopic features and designing reliable classifiers.
- Gain insight into the importance of
dermatoscopic features.
The importance of dermatoscopic features is still very much a matter of
research. Any additional insight into this area is desirable.
- Develop a probabilistic neural
classifier design framework.
In order to obtain reliable classification systems based on neural
networks, a principled probabilistic approach will be followed.
Hence, the work should be of interest to both the
dermatological and engineering communities.
To appear in edited CRC Press book.