Here some basic concepts of Statistical Shape Analysis are introduced and applied to a specific problem: automatic recognition and classification of cells coming from tumor tissues, from their nuclear profiles. The technique here described, which is commonly used for the description of the mean geometrical characteristics of families of random objects and their statistical analysis, is proposed as an alternative (or in addition) to the study of an asymptotic fractal model for the contour of nuclei.
|Titolo:||Statistical shape analysis applied to automatic recognition of tumor cells|
MICHELETTI, ALESSANDRA (Primo)
|Parole Chiave:||Shape analysis; landmarks; discriminant analysis|
|Settore Scientifico Disciplinare:||Settore MAT/06 - Probabilita' e Statistica Matematica|
|Data di pubblicazione:||2005|
|Tipologia:||Book Part (author)|
|Appare nelle tipologie:||03 - Contributo in volume|