MacPI - Macular Pigment Optical Density Measurement

Andrew O'Brien, Conor Leahy, and Chris Dainty

This page presents a technical description of MacPI. For a general overview and the latest news on MacPI, please click here.

Introduction

The macular pigment (MP) is located in the macular region of the retina (Figure 1). The characteristics of the MP suggests that this yellow collection of carotenoids plays a protective role, especially in the protection of the vital cone photoreceptors. The accurate measurement of the macular pigment optical density (MPOD) is deemed important due to its inferred link to the development of age related macular degeneration (AMD) (gu re 2). Pigment levels may be boosted by lifestyle modification (e.g. diet), thus blindness in later life may be prevented by implementing these procedures to provide additional retinal protection.

Figure 1. Location of the macular pigment on the retina.

Methods
This device utilises the known spectral characteristics of the MP in order to obtain a measurement of the MPOD. This is achieved by obtaining a green and blue image of the retina The choice and design of the various components in the device hardware are vital to producing retinal images of sufficient quality, in order to maximise the effectiveness of the software. The device is designed to image the retina through an undilated pupil (2.5mm), to measure and account for ocular scatter, deal with differential spectral absorption across the retina, measure and account for heterogeneous illumination, aid in alignment of the subject and return an usable, accurate representation of the MPOD distribution.

Figure 2. MPOD distribution calculated from reflectance values for green and blue images.

Each element of equation 1 must be calculated accurately from the acquired reflectance values on the detector. The true reflectance values can be altered due to ocular scatter, absorbing substances within the retina and non-uniform illumination of the retina. Scatter is measured by use of an appropriately designed mask in the hardware and an analysis of the image of this mask in software. Non-uniform illumination can be represented by a polynomial fit. The non-uniformity can then be isolated and corrected for in the original images. Peripheral reectances are calculated by implementing an entropy method. Pixel groups are analysed for their information content and rejected or accepted depending on this entropy value. This ensures that no blood vessels (lower reflectance value) will influence the calculation of the true peripheral reflectance value. Matched filter techniques are implemented to automatically locate various features on the retina (scatter struts and fovea centres). Images are aligned by implementing an efficient Wiener filter based algorithm. The results screen from this device is shown in Figure 4.

Results
The MPOD of Forty two subjects was measured using this device and using a subjective device. A strong correlation between measurements from the two devices was found (Fig 5). The primary difficulty with correlating the two devices, is the inherent differences which arise due to the two devices employing two different techniques to extract the MPOD.

Figure 3. Results screen.

An objective method to measure the MPOD is quicker and easier to perform. An objective method should also be more accurate than a subjective method if the various difficulties inherent with the objective reflectance technique are successfully identified and accounted for.

If you have any questions about MacPi or if you would like to receive more information, please send an e-mail to any of the following:

• Disclaimer
• Privacy
• Copyright
• Accessibility