Applications for content-based image retrieval (CBIR) are found in a wide range of industries, including e-commerce, multimedia, and healthcare. CBIR is essential for organising and obtaining visual data from massive databases. Traditional techniques frequently fail to extract high-level, relevant information from images, producing retrieval results that are not ideal. This research introduces a novel Convolutional Fine-Tuned Threshold Adaboost (CFTAB) approach that integrates deep learning and machine learning techniques to enhance CBIR performance. This dataset comprises image-based data collected from multiple sources. This image data were pre-processed using Adaptive Histogram Equalization (AHE). The features of localized image data were extracted using VGG16. For an efficient CBIR process, a novel CFTAB approach was introduced. It combines both deep and machine learning (ML) methods in the proposed architecture to improve the excellence of image search. To further improve performance, CFTAB incorporates an improved AB algorithm. This algorithm adjusts the threshold levels dynamically within a robust classifier to optimize training outcomes.

• Klíčová slova
• Content-based image retrieval (CBIR), Convolutional Fine-Tuned Threshold Adaboost (CFTAB), Deep and machine learning (DL, ML), High-level information, VGG-16,
• Publikační typ
• časopisecké články MeSH

In therapeutic diagnostics, early diagnosis and monitoring of heart disease is dependent on fast time-series MRI data processing. Robust encryption techniques are necessary to guarantee patient confidentiality. While deep learning (DL) algorithm have improved medical imaging, privacy and performance are still hard to balance. In this study, a novel approach for analyzing homomorphivally-encrypted (HE) time-series MRI data is introduced: The Multi-Faceted Long Short-Term Memory (MF-LSTM). This method includes privacy protection. The MF-LSTM architecture protects patient's privacy while accurately categorizing and forecasting cardiac disease, with accuracy (97.5%), precision (96.5%), recall (98.3%), and F1-score (97.4%). While segmentation methods help to improve interpretability by identifying important region in encrypted MRI images, Generalized Histogram Equalization (GHE) improves image quality. Extensive testing on selected dataset if encrypted time-series MRI images proves the method's stability and efficacy, outperforming previous approaches. The finding shows that the suggested technique can decode medical image to expose visual representation as well as sequential movement while protecting privacy and providing accurate medical image evaluation.

• Klíčová slova
• Encryption, Heart Disease, MRI Images, Multi-faceted long short-term memory (MF-LSTM),
• MeSH
• algoritmy MeSH
• deep learning MeSH
• důvěrnost informací MeSH
• lidé středního věku MeSH
• lidé MeSH
• magnetická rezonanční tomografie * metody   MeSH
• nemoci srdce * diagnostické zobrazování   MeSH
• neuronové sítě MeSH
• počítačové zpracování obrazu metody   MeSH
• soukromí * MeSH
• zabezpečení počítačových systémů MeSH
• Check Tag
• lidé středního věku MeSH
• lidé MeSH
• mužské pohlaví MeSH
• ženské pohlaví MeSH
• Publikační typ
• časopisecké články MeSH