ICEECE2012 Poster Presentations Adrenal cortex (113 abstracts)
Diagnosis of Cushing’s syndrome by automatic face classification using frontal and side-view photographs
R. Kosilek 1 , J. Schopohl 1 , M. Grünke 2 , C. Dimopoulou 1 , G. Stalla 3 , M. Reincke 1 , M. Günther 4 , R. Würtz 4 & H. Schneider 1
1Medizinische Klinik Innenstadt, Ludwig-Maximilians University, Munich, Germany; 2Medizinische Poliklinik, Campus Innenstadt, Ludwig-Maximilians University, Munich, Germany; 3Max-Planck-Instititut für Psychiatrie, Munich, Germany; 4Ruhr-Universität, Bochum, Germany.
Background: Cushing’s syndrome is a disease that presents with clear symptoms and causes considerable harm to the body if left untreated, yet often remains undiagnosed for prolonged periods of time. Face-classification software might recognize typical changes of the face and thus aid in diagnosing the disease early as we have previously shown in the classification of acromegaly.
Methods: Using a regular compact digital camera, we took frontal and side-view pictures of 21 female patients with Cushing’s syndrome (14 endogenous, seven iatrogenic) and of 42 age- and sex-matched controls (2:1 matching).
Nodes were then placed on disease-relevant structures of the face to analyze the pictures using computerized similarity analysis based on Gabor-jets and geometry functions. The leave-one-out cross-validation method was employed to classify subjects by the software.
Results: Using the same combination of Gabor-jets and geometry functions as in our previous publication, 80.1% of patients and 97.6% of controls were correctly classified by the software. This resulted in a total classification accuracy of 92.1%. If only frontal views and only one control person for each patient was included, the classification accuracy was 85.7 and 66.7% in patients and controls, respectively.
Conclusions: In this preliminary analysis we found a good classification accuracy of Cushing’s syndrome by face-classification software. By employing 2:1 matching and implementing side-view pictures into the classification process, we could substantially improve classification accuracy.
Declaration of interest: The authors declare that there is no conflict of interest that could be perceived as prejudicing the impartiality of the research project.
Funding: This research did not receive any specific grant from any funding agency in the public, commercial or not-for-profit sector.
Article tools
My recent searches
My recently viewed abstracts
Authors
Endocrine Abstracts
ISSN 1470-3947 (print) | ISSN 1479-6848 (online)
© Bioscientifica 2024 |
Privacy policy |
Cookie settings
BiosciAbstracts
Bioscientifica Abstracts is the gateway to a series of products that provide a permanent, citable record of abstracts for biomedical and life science conferences.
Find out more