Skip to Top NavigationSkip to Top Navigation UNotesSkip to Utility NavigationSkip to Search
Mobile Menu
Bookmark and Share

CETA's Hassan Salehi Presents and Publishes Research Paper at International Conference


Posted 02/21/2018
Category: Accolades
Dr. Hassan S. Salehi

Dr. Hassan S. Salehi

The deep convolutional neural network architecture for early dental caries detection.

The deep convolutional neural network architecture for early dental caries detection.

Dr. Hassan S. Salehi, visiting assistant professor of electrical and computer engineering at the University of Hartford's College of Engineering, Technology, and Architecture (CETA) has presented and published a research article at the SPIE Photonics West International Conference, held in San Francisco CA, January 27-February 1, 2018. The paper, "Deep learning classifier with optical coherence tomography images for early dental caries detection," was written by Dr. Salehi along with Nima Karimian, PhD candidate at UCONN ECE; Dr. Mina Mahdian, assistant professor and program director at Stony Brook University School of Dental Medicine; Dr. Hisham Alnajjar, CETA Associate Dean; and Dr. Aditya Tadinada, assistant professor at University of Connecticut Health Center (UCHC).   
 
In this research project, Dr. Salehi has been leading the development of a novel approach combining Deep Convolutional Neural Networks (CNN) and Optical Coherence Tomography (OCT) imaging modality for classification of human oral tissues to detect early dental caries. In this paper, OCT images of oral tissues with various densities are input to a deep convolutional neural network classifier to determine variations in tissue densities resembling the demineralization process. The deep convolutional neural network automatically learns a hierarchy of increasingly complex features and a related classifier directly from training data sets. The initial convolutional neural network layer parameters are randomly selected. The convolutional neural network employs two convolutional and pooling layers to extract features and then classify each patch based on the probabilities from the SoftMax classification layer. Afterward, the neural network calculates the error between the classification result and the reference label, and then utilizes the backpropagation process to fine-tune all the layer parameters to minimize this error using batch gradient descent algorithm. The proposed technique is validated on ex vivo OCT images of human oral tissues (enamel, cortical bone, trabecular bone, muscular tissue, and fatty tissue), which attested to effectiveness of the proposed method.
 
SPIE Photonics West is the world's largest multidisciplinary event for photonics. Every year over 20,000 people come to hear the latest research and find the latest devices and systems driving technology markets including state-of-the art medical technologies, the Internet of things (IoT), smart manufacturing and “Industry 4.0,” autonomous vehicles, scientific research, communications, displays, and other solutions powered by photonics.
 
Dr. Hassan S. Salehi

Dr. Hassan S. Salehi

The deep convolutional neural network architecture for early dental caries detection.

The deep convolutional neural network architecture for early dental caries detection.