Image Based Detection of Craniofacial Abnormalities using Feature Extraction by Classical Convolutional Neural Network

Saloni Agarwal, Rami Robert Hallac, Rashika Mishra, Chao Li, Ovidiu Daescu, Alex A Kane

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

The ubiquitous approach of transfer learning for feature extraction is harnessed for image based detection of two types of craniofacial abnormalities: pediatric cleft and craniosynostosis. In the current study, using features extracted from pre-Trained AlexNet activations, we train a multi class support vector machine (SVM) for cleft lip abnormality and developed a multi-view classifier using max voting for craniosynostosis anomaly detection. We achieved Area under the ROC curve (AUC) value of 0.95 for cleft abnormality and 0.84 for craniosynostosis.

Original languageEnglish (US)
Title of host publication2018 IEEE 8th International Conference on Computational Advances in Bio and Medical Sciences, ICCABS 2018
PublisherInstitute of Electrical and Electronics Engineers Inc.
Volume2018-October
ISBN (Electronic)9781538685204
DOIs
Publication statusPublished - Nov 20 2018
Event8th IEEE International Conference on Computational Advances in Bio and Medical Sciences, ICCABS 2018 - Las Vegas, United States
Duration: Oct 18 2018Oct 20 2018

Other

Other8th IEEE International Conference on Computational Advances in Bio and Medical Sciences, ICCABS 2018
CountryUnited States
CityLas Vegas
Period10/18/1810/20/18

    Fingerprint

Keywords

  • AlexNet
  • Craniofacial
  • Craniosynostosis
  • multiclass SVM
  • Pediatric Cleft
  • Transfer Learning

ASJC Scopus subject areas

  • Biomedical Engineering
  • Computational Theory and Mathematics

Cite this

Agarwal, S., Hallac, R. R., Mishra, R., Li, C., Daescu, O., & Kane, A. A. (2018). Image Based Detection of Craniofacial Abnormalities using Feature Extraction by Classical Convolutional Neural Network. In 2018 IEEE 8th International Conference on Computational Advances in Bio and Medical Sciences, ICCABS 2018 (Vol. 2018-October). [8541948] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICCABS.2018.8541948