Enhancing DICOM Image Compression Using Sheep Flock Optimization Algorithm with Modified Haar Wavelet Approach
Year: 2025
Type: article
Abstract: Digital Imaging and Communication in Medicine (DICOM) is a medical imaging file standard employed for storing large amounts of data, such as imaging procedures, patient data, and the image itself. With the increasing use of medical imaging in medical diagnoses, it is imperative to have a secure and rapid technique for sharing a considerable number of medical images among medical staff, and compression has often been a choice. Different compression approaches are utilized, including lossless techniques such as Run-Length Encoding (RLE) and lossy methods such as JPEG and JPEG2000. Lossless compression preserves each image's information, which makes it fit for healthcare data where fidelity is vital, like in radiology. At the same time, lossy compression is used to sacrifice some image details to accomplish high compression ratios, frequently employed in situations where slight degradation in quality is acceptable, like in telemedicine applications or while transmitting larger datasets over a limited bandwidth network. DICOM compression standard ensures compatibility and interoperability over dissimilar medical imaging modalities and systems. This article introduces a new DICOM Image Compression Using a Sheep Flock Optimization Algorithm with Modified Haar Wavelet (SFOA-MHW) approach. The presented SFOA-MHW technique uses SFOA to resolve the wavelet discontinuities that take place while performing image compression via thresholding. The SFOA-MHW technique converts the input images into sub-band details and approximation by using MHW, which then employs the threshold. Finally, the SFOA is used to select the threshold values. The SFOA-MHW technique aids in DICOM image compression by preserving fine details with a high compression ratio. The performance evaluation of the SFOA-MHW model is verified utilizing DICOM image sample sets. The experimental values highlighted that the SFOA-MHW technique gains better performance over other techniques in terms of distinct measures.
Authors: Navaneethan S, D.Kanchana, Rahul S G
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W4414851427