کنفرانس بین المللی فناوری اطلاعات و دانش

صفحه اصلی / چهاردهمین کنفرانس بین المللی فناوری اطلاعات و دانش

COVID-19 Image Retrieval Using Siamese Deep Neural Network and Hashing Technique

نویسندگان :

Farsad Zamani Boroujeni¹ Doryaneh Hossein Afshari² Fatemeh Mahmoodi³

1- دانشگاه آزاد اسلامی واحد علوم و تحقیقات 2- دانشگاه آزاد اسلامی اصفهان (خوراسگان) 3- دانشگاه آزاد اسلامی اصفهان (خوراسگان)

کلمات کلیدی :

Lung Image Retrieval،Siamese Neural Network،Hashing،Minimum Redundancy Maximum Relevance،Feature Selection

چکیده :

Medical images are crucial for early diagnosis and are increasingly important in modern medicine. This study presents a medical image retrieval system using a Siamese neural network with 13 layers. The system uses the Minimum Redundancy Maximum Relevance (mRMR) technique to extract deep features from the Siamese, and then uses binary hashing to retrieve similar images using Hamming distance. The experimental results show that the proposed method improves lung image retrieval by over 6% compared to previous methods, achieving an average accuracy of 93.83% and 92.73% in 5 and 10 retrieved images, respectively.