Application of Pyramidal Decomposition to Improve Digital Radiography Image Quality

Abstract

As an effort to create innovation in the world of radiography, it is necessary to develop technology in software. This effort is to improve image quality by using pyramidal decomposition. This digital image decomposition is referred to as pyramid decomposition. The original image is decomposed into several frequency bands, repeatedly divided into high-pass components and low-pass components. The high-pass component is set aside while the low-pass image is subjected to subsequent division. This creates a kind of "3D" stack of image layers. Each layer is at a lower frequency and therefore fuzzier. This processing was pioneered by Philips Healthcare as UNIK (Unified Image Quality Enhancement), and by Agfa as MUSICA (Multi-Scale Image Contrast Amplification) with various innovations. The test image uses digital radiography images resulting from innovation from 14bit RAW digital conversion into JPG format. Image quality is calculated using Mean Square Error (MSE) and Peak Signal Noise Ratio (PSNR). The pyramidal decomposition application succeeded in improving the quality of digital radiography images with an average MSE reduction value of 0.018 and an average PSNR increase of 22.114 dB. Visually, there is a constant increase in contrast and detail, so it can be applied in the medical field.

Keywords

• pyramidal decomposition, digital radiography, mse, psnr