Quantification of Optic Nerve Cross Sectional Area on MRI: A Novel Protocol using Fiji Software

This protocol aims to describe standardized methods for optic nerve cross sectional area assessment using MRI. We will describe image acquisition using widely available imaging sequence. We will also describe image analysis using the open source feature software.

One of the important uses of such protocol is in glaucoma patient assessment, where it will provide valuable data on optic nerve integrity. The image acquisition protocol will start by fully explaining the examination to the patient, and that the following instructions need to be followed. You will need to change into a special gown for imaging.

You need to make sure not to carry or wear any metallic material, this may include some kinds of face masks. Some implantable devices may not be compatible with MRI imaging, this may include pacemakers and insulin pumps. You must inform the physician of any metallic object left inside your body.

Due to the confined space of the MRI machine, the exam might not be suitable for patients with claustrophobia. The exam is expected to take 15 minutes where the patient is required to stay still. A proper consent form needs to be signed by the patient.

Prior to imaging, the physician will explain the steps to the patient and instruct them to focus on a straight target within the MRI machine, in order to obtain a properly oriented image. Further precautions are explained and safety apparatus, such as headphones and a panic button are provided to the patient to ensure a proper and comfortable exam. A coronal cut of the optic nerve will be taken orthogonal, i.e.

perpendicular to the nerve three millimeters posterior to the globe. Scout images in the transfer saw and oblique sagittal planes will be used to ensure optimal optic nerve direction and optic nerve globe junction positioning. The quality of gaze fixation can be accessed by CSF distribution around the optic nerve, where it should be uniformly distributed around the nerve with almost equal thickness on all sides.

Input the following parameters for image acquisition. At T2 weighted fat suppression sequence, TR, 3000 milliseconds. TE, 90 milliseconds.

Field of view, 16 by 16 centimeters. Matrix, 296 by 384. Slice thickness, three millimeters.

Slice gap, 0.3 millimeters. The final image analyzed was an oblique coronal image, three millimeters behind the globe. For the image analysis protocol, we will start by downloading the Fiji image processing package from imagej.net.

We will start by going to their website, selecting Downloads, Fiji, and then choosing the appropriate version for your device. After installing, launch the Fiji program. Upload the coronal image of the optic nerve to the program by selecting File, Open, and then selecting the coronal image you want.

Please note that you need to maintain image quality during transfer, as not doing that will result in unreliable image analysis results. Standardize the scale by specifying the number of pixels per a unit of length. This can be done by drawing a straight line on the map scale.

Then choose Set Scale from the Analyze menu bar. Specify the length of the line as appears on the map scale with the proper unit of length. Convert the image into gray scale using the Image menu, then choose Type, then 8-bit.

Quantifying the range of intensity of white matter pixels, choose the Lasso Selection tool by going to the Plugin menu, Segmentation, then Lasso selection, then select a sufficient area of white matter. To expand the area, hold down Shift and expand the white matter. Make sure you do not cover any gray matter during the process.

A total selected area of 1000 pixels is enough. You can make sure you have 1000 pixels selected by going to Analyze, and then Measure, to quantify the selected area. Open the Histogram tool from the Analyze menu, which shows the distribution of pixel intensity in the white matter area selected.

Click on the Live box to make sure that the histogram assesses the selected area. The graph on the histogram should show a normal distribution of intensity. Calculate the white matter intensity range as follows, low limit is mean intensity, minus three times standard deviation.

For our case, that's 94.372, minus three, multiplied by 7.085. The upper limit is mean intensity, plus three times standard deviation. For our case, it's 94.372, plus three times 7.085.

To open the Threshold tool, go to the Image menu, followed by Adjust, then select Threshold. Specify the range calculated from the previous step by selecting set, then typing in the upper and lower limit. In our case, it's 73 and 115.

Tick only Dark Background function and specify black and white annotations, B&W, from the dropdown list, then click Apply. The mask for white matter present within the optic disc will appear. Using the Lasso selection tool, which can be chosen from the Plugin menu, as explained above, or from the Quick Selection menu, select the black area, representing the optic disc.

Use the measure function from the Analyze menu bar to calculate the area of the optic disc measured in centimeters. In our case, it is 0.069. The cup to disc ratio for a 30 year old male patient presenting for an ophthalmology checkup exam was 0.8, which is suspicious and might be suggestive of glaucoma.

Upon performing optical coherence tomography for nerve fiber layer thickness, we found that the nerve thickness was within the normal limit for the age. The patient was planned for a head MRI, where a coronal cut for optic nerve assessment was ordered, as detailed in the previously mentioned protocol. We obtained the coronal MRI cut three millimeters behind the optic disc.

White matter mean intensity was 94.372, standard deviation 7.085, which resulted in a white matter intensity range of lower limit, 94.372 minus 21.255, equals 73.117. Upper limit, 94.372, plus 21.255, equals 115.627. Figure 2A shows the coronal image.

Figure 2B shows the coronal image after applying the white matter threshold using the calculated upper and lower limits. While figure 2C shows the optic nerve white matter for quantification. The cross-sectional area of the white matter of the left optic nerve was 6.9 millimeters squared, which is within the normal limits for his age.

We provided a detailed protocol for a standardized method for optic nerve assessment and quantification using MRI machines, which non-specialized medical doctors can execute with high reliability. We recommend adopting this protocol for studies on optic nerve assessment, which will enable comparison between study findings. Moreover, we recommend the use of this protocol when MRI images are ordered to exclude secondary causes of optic nerve neuropathy and glaucoma.