The setup should allow for good OCT examination conditions. Patients may require wheelchair access and additional help during the assessment for which there needs to be sufficient space. Dimmable light permits the OCT examination to be performed in a darkened room. The effect of stray light from an open door/entrance should be minimized by positioning the patient and examiner appropriately. Allow for sufficient space on the examiner’s side too to permit teaching and technical help with challenging situations
To enable assistance to patients through a third person, provide sufficient space around the patient’s place. Finally, the examination room should have wheelchair access.
For teaching purposes allow sufficient space behind the OCT machine on the examiner’s side.
The patient is in a vulnerable role. Their emotions may be influenced by the disease. Insecurity and anxiety may create challenges for a good assessment (Box 3.1). Reasons can be physical disabilities affecting mobility and communication but also bad experiences with past investigations. Building and maintaining trust is a good way to help overcome these issues.
The main task for the patient is to fixate a small visual target during the examination, which requires keeping the eye, head, and body still.
Patients suffering from multiple sclerosis may have symptoms that can complicate the OCT assessment. These problems should be recognized prior to the OCT assessment. Some symptoms may require additional assistance and protocol modifications, which are best discussed prior to starting the assessment.
Vision: inability to maintain visual fixation because of a central visual field defect, poor visual acuity, and/or nystagmus.
Hearing: inability to follow auditory directions during the examination.
Mobility: pathology affecting the cerebellar function and pyramidal and extrapyramidal systems may all make it difficult for a patient to sit comfortably throughout the OCT assessment and maintain a still head position.
Cognition: impaired cognitive function may not only limit what a patient understands about the assessment and the ability to follow directions but also influence the patient’s behavior during the OCT assessment itself. Patients may be easily distractible, look around, keep talking, grow fidgety, or tire out during the assessment. Loss of praxic skills and higher visual functions pose a challenge too.
Face, lid, and eye: anatomical constraints of the face may make it impossible for the OCT lens to be moved close enough the eye to obtain an OCT image. A ptosis of narrow palpebral fissure may require extra assistance. Makeup and long eyelashes may dirty the OCT lens. Small pupils (<2 mm) may require pharmacological pupil dilation; this may also be of help in some cases with opacities in the light pathway, such as cataract or floaters.
In general terms, the OCT assessment is relatively straightforward in patients suffering from multiple sclerosis compared to patients with movement disorders, dementia, rapidly progressive neurodegenerative disorders, or pathology causing impaired vision and blindness.
The role of the examiner(s) depends on the professional background. The examiner can be an OCT technician, physician, nurse, trainee, or student. A trainee will require teaching, both hands on and theoretical. An OCT technician may want to discuss medical questions potentially arising from certain images. A nurse or physician may require technical assistance. Therefore the position of the examiner should be clear to the patient.
The task of the examiner is to obtain high-quality OCT data, which requires giving directions.
The examiner should be able to communicate in lay terms what the OCT assessment includes (Box 3.2). The examiner should be able to answer the patient’s question: “Will my eye be touched?” The examiner needs to give the appropriate instructions to aid the patient in maintaining visual fixation and keeping the eye, head, and body still. The examiner needs to recognize when extra assistance is needed for the patient. Likewise, an examiner needs to know when to call for help with OCT image acquisition, handling, and storage.
As a general rule, always explain to the patient what you want to do before you do it. Like with other paramedical tests, be very cautious about what you say to a patient with regard to the images just obtained. Your words may cause distress and harm. The errors of overreporting are likely to be greater than the errors of underreporting. There is no harm in asking for more time to discuss the images with your colleagues first.
Common to all OCT devices on the market is that a light signal is focused through a lens to the retina. Next, light scattered back from the retina is captured by the same lens. Finally, the OCT image is composed with the aid of device-specific hardware and software. For brevity, this chapter takes examples predominantly derived from one SD-OCT device. There are differences between the many commercial devices, but the OCT market is developing so rapidly that it will not be possible to have an updated pragmatic chapter on all.
All one needs to do to obtain a good OCT image is to place the lens correctly. Practically, this is similar to the handling of a slit lamp. If you have never used a slit lamp or ophthalmoscope, think of the opening scene from a James Bond movie (Fig. 3.2).
A mnemonic aid for correct placement of the OCT light beam is the gun barrel sequence that features in almost every 007 film. The OCT light beam should enter the center of the pupil
Which OCT Do You Want to Own?
The OCT market is rapidly expanding. Both commercially developed and in-house-developed devices contribute to the data available. Access to in-house devices will remain exclusive to a few research centers. In contrast, the commercial sector has developed an impressive “shopping list,” which was recently reviewed by Fiona Costello . Device-specific features for spectral-domain and swept-source OCT are reproduced and updated from reference . At the time of writing, most literature on OCT in multiple sclerosis and optic neuritis is based on data from the Cirrus and Spectralis devices:
Cirrus (Zeiss)™ combines advanced software features that enable good-quality scan acquisition. The FastTrac ™ option reduces eye motion artifacts. The Guided Progression Analysis™ permits for longitudinal comparison of retinal layer thickness changes. The FoveaFinder™ ensures that the Early Treatment Diabetic Retinopathy Study (ETDRS) grid is centered on the fovea. The AutoCenter™ centers the ring scans on the optic nerve head. Automated seven-layer segmentation is possible. Finally, HD-OCT technology enables sharing of data with networked review stations.
The SPECTRALIS (Heidelberg Engineering) is a multimodality device combining the advantages of a confocal scanning laser ophthalmoscope (cSLO) with the benefits of achieving cross-sectional images of the retina with a spectral domain-based optical coherence tomography (OCT) module. Besides the standard infrared (IR)-based fundus image, the SPECTRALIS offers up to six different imaging modalities in the highest stage of expansion. The Multicolor™ feature combines three wavelengths (IR, green, blue) to generate a true detailed color fundus image, while the autofluorescence (Bluepeak™) allows the structural assessment of the retina as well as gathering metabolic information of the lipofuscin distribution in the retinal pigment epithelium (RPE), which is of relevance for ophthalmologic disease. Next, there is an option for fluorescence angiography (FA) and indocyanine green angiography (ICGA) in combination with a noncontact ultra-wide-field (UWF) lens peripheral angiography. The anterior segment (ASM) lens extends the capability to image the cornea, sclera, and chamber angles. An important feature is the active eye-tracking system (TruTrack™), which compensates involuntary eye movements and blinking at time of scanning without need for post-processing. This technique allows a precise point-to-point registration of the fundus image and OCT scan. Moreover, the eye-tracking system is the base for the rescan (AutoRescan™) function, which uses the baseline exam to place the follow-up exam on precisely the same position on the retina before acquiring the follow-up scan. This enables reliable longitudinal data. Recent software developments not only correct to circumvent rotatory problems (e.g., head tilt) but also adjust the scans individually on the retina. The so-called anatomic positioning system (APS) uses the center of the fovea and the Bruch membrane opening (BMO) as distinct reference points in the retina. Taking head tilt and cyclotorsion into account during acquisition allows for an individual classification of the scans. Finally there is the option for 11-layer segmentation on B-scans individually and as a batch as relevant for clinical studies.
IVue (Optovue). Designed to be easy and fast, the machine includes a foot switch and touch screen. There is also the option for imaging of the anterior segment including corneal thickness measurements. Color-coded retinal layer segmentation and mapping includes the GCL, GCC, and the option for 3D “en face” visualization.
Angiovue (Optovue). During the writing of this chapter, the first OCT device on the market to permit for 3D visualization of the retinal vasculature became available. This technology does not permit investigation for leakage, and the potential role for optic neuritis and multiple sclerosis remains to be seen. Possibly diseases like Susac syndrome or anterior ischemic optic neuropathy (AION) and paracentral acute middle maculopathy (PAMM) are of greater interest.
3D OCT 1000/2000 (Topcon). Longitudinal image acquisition is possible, enabling comparison over time. Both 3D and 2D OCT viewing options can be combined with the fundus image (FastMap™ software). There is the option to manually pinpoint the location of an OCT image in the fundus image (Pin-Point™). Images are viewed through the EyeRoute® Image Management System.
DRI OCT Atlantis (Topcon). A very quick swept-source OCT with follow-up function. Has the advantage of invisible scan lines, which makes it easier for the patient to focus on the target. A large 12 mm-wide screen for scanning. Automated 7-layer segmentation.
OCT/SLO (Optos). It is possible to combine SLO and retinal tracking before, during, and after acquisition of the OCT scan. Registered follow-up is possible with the SLO “Lock and Track” function. There is the option to align 3D topographies to the SLO image, which permits correction for artifacts due to rotation and shift. Longitudinal comparisons, including retinal thickness, is possible with aid of the “Auto – Compare” feature. All images can be viewed remotely using the “Viewer Software.”
Copernicus HR (Optopol Technology). Longitudinal image comparison possible. Option for 3D visualization and volume maps. There is a disk damage likelihood scale (DDLS) for the optic nerve, which is based on the rim/disk (r/d) ratio and optic nerve size. Imaging of the anterior segment is possible at a 3 μm resolution. Remote viewing of images from a central database is possible.
Canon OCT HS-100. Longitudinal imaging is possible using SLO tracking of retinal images. Option for enhanced depth imaging of the choroid. Option to import retinal camera images, which then can be aligned and overlayed with the SLO image. Ten-layer retinal segmentation.
RS-3000 (Nidek). Longitudinal imaging using multifunctional follow software is possible for this SD-OCT device, which enables averaging up to 120 images. Option for selectable OCT sensitivity aids improved visualization with a range of ocular pathologies. Eye-tracking capabilities including torsion. Reports can be customized.
Infrared (IR) and (Confocal) Scanning Laser Ophthalmoscopy (c)SLO
There are two options: an IR camera or a scanning laser ophthalmoscope (SLO). The SLO scans the image line per line, similar to how old-fashioned television screens used to build up an image. In most devices this is a bit slower than the IR camera (about 16 images per second compared to 24 images per second), but scanning speeds can be increased. The speed of both IR and (c)SLO enables video recording. One advantage of the SLO and confocal SLO (cSLO) images is the high resolution. Taken together, the image quality of (c)SLO is better than for IR. In (c)SLO there is a small depth of focus and scattered light is better suppressed. Therefore patients find (c)SLO more comfortable (less bright light exposure). There are good 3D imaging capabilities. Finally, imaging with smaller pupil sizes becomes possible.
IR is used by the following OCT devices: Stratus, 3D OCT 2000, iVue, and Copernicus HR. (c)SLO is used by Spectralis, Cirrus, DR 1, RS-3000, OCT SLO, and Canon OCT HS-100.
Pupil Size Requirements
A preference in a neurological clinic is not to dilate the pupil pharmacologically. In most patients it will be possible to obtain good-quality OCT scans without the need to dilate the pupil. The pupil size should be measured at the same light level at which the OCT images will be acquired. At time of writing, the following pupil size requirements apply:
≥2 mm: Spectralis, Cirrus
≥2.5 mm: 3D OCT, DR 1, RS-3000
≥3 mm: iVue, OCT SLO, Copernicus HR, Cannon OCT HS-100
≥3.2 mm: Stratus
The axial resolution of OCT is related to the bandwidth of light. A finer axial resolution is achieved with a wider span of wavelengths in the light. The current light source range is 820–1050 nm.
820 nm: Stratus
830 nm: OCT SLO
840 nm: Cirrus, 3D OCT 2000, iVue
850 nm: Copernicus
855 nm: Canon OCT HS-100
870 nm: Spectralis
880 nm: RS-3000
1050 nm: DR 1
For all devices the axial resolution (3–10 μm) is better than the transverse resolution (12–20 μm). The reported axial resolution is:
3 μm: Canon OCT HS-100, Copernicus HR
5 μm: Cirrus, iVue
6 μm: 3D OCT 2000
7 μm: Spectralis, RS-3000
8 μm: DR 1
10 μm: OCT SLO, Stratus
The number of A-scan obtained per second determines the scanning speed. High scanning speed reduces the likelihood of motion artifacts. Scanning speed will become more and more important with the advance of Doppler OCT. The exquisite images of the retinal vasculature obtained by, for example, the Angiovue is only possible due to the high scanning speed of 70,000 A-scans/s. At time of writing, scanning speeds are:
400 A-scans/s: Stratus
25,000 A-scans/s: iVue
27,000 A-scans/s: Cirrus
40,000 A-scans/s: Spectralis
50,000 A-scans/s: 3D OCT 2000
52,000 A-scans/s: Copernicus HR
53,000 A-scans/s: RS-3000
70,000 A-scans/s: Angiovue, Canon OCT HS-100
100,000 A-scans/s: DR 1
Basic OCT Protocol
As a minimal requirement, one needs to capture (1) the area where all axons leave the eye, the optic nerve head, and (2) the area most relevant for our vision, the macula. A basic OCT protocol therefore comprises a volume scan of the macula and optic nerve head. Some patients with MS will have difficulties maintaining the visual fixation needed to acquire a good-quality volume scan. A volume scan may take too long for these patients. Therefore if speed matters, a ring scan around the optic nerve head may be used instead of the volume scan. The basic OCT scan protocol, which should be possible in most patients with MS, is summarized in Fig. 3.3.
Basic OCT protocol for patients with MS. (a) Ring scan of the optic nerve head and (b) volume scan of the macula
Start OCT Scan
This section was written for the Spectralis device (Fig. 3.4) and reproduced from reference . For other devices the reader is referred to the device-specific instruction manuals.
Heidelberg Spectralis OCT
First, prepare the software by entering the patient details into the database. In case of a follow-up examination, a patient record will already have been created and you will only need to specify the OCT operator taking the image.
Second, prepare the hardware. (1) Pull the OCT camera head back, (2) clean the parts coming in contact with the patient (forehead and chin rest), (3) remove the lens cap and wipe the lens if not clean, (4) adjust the height of the table to a comfortable position for the patient, (5) adjust the height of the patient’s head so that the marker indicating the canthus (Fig. 3.4) comes to the patient’s eye, and (6) ask if the patient is ready to go or if there are any questions. The patient should already know that the eye will not be touched.
Third, start the OCT acquisition window. There will be a short device calibration period before you can click on the yellow square to the bottom right of the screen to start imaging. Now the cSLO image will appear in the window to the left of the screen. A few seconds later, the OCT signal will be seen in the window to the right of the screen.
Fourth, repeat explaining to the patient that a fixation target will appear. This is a blue light. Some patients will see a series of vertically displaced blue lights for optical reasons. In this case ask them to look at the brightest one, which typically is the bottom one. Some patients will not be able to see the internal fixation target. If this is the case, (1) use the external fixation target consisting of a lamp, (2) use your finger (or alternative) as a target, or (3) give verbal instructions where to look (right, left, up, down). Be aware that good visual fixation will help to get quick and high-quality OCT images. Patients are permitted to blink and move eyes in between for their own comfort. A teary dry eye is no good for OCT imaging.
Fifth, in the acquisition window select “Axonal.” The advantage of the axonal setting is that it was developed specifically for neurological conditions. The OCT B-scan lines are orientated vertically, which permits capture of all axons cross-sectionally.
Now you are ready to choose from the “Preset” buttons. For a basic protocol, you will only need the macular volume scan and ring scan. It is your choice which one you start with. My personal preference is to start with the macular volume scan because the patient finds it easier to fixate.
The Macular Volume Scan
First, click on one of the macular volume scans from the preset buttons. Scans with a low ART1 (e.g., 9) are quicker than scans with a high ART (e.g., 40–100) and better suited to detect small microcysts but less suited for high-quality retinal layer segmentation. A high ART takes longer and may be less suited for patients with difficulties to maintain visual fixation or to sit comfortably for long in the required position.
Second, use the “007 technique” (Fig. 3.2) on the cSLO to get the retina in focus. The OCT image will appear once you have advanced the OCT camera head close enough to the patient’s eye. If the OCT image “flips over” on the top of the screen you will need to move the camera a bit back again. Do not touch the eye.
Third, once you have a good cSLO image and OCT signal, activate the device-specific “Eye-Tracker.” To do so, you need to either (i) press the joystick button for about 2 s (basic Spectralis device) or (ii) press the round button below the touch screen (all other Spectralis devices).
Fourth, place the volume scan with the mouse over the macula.
Fifth, once you start scanning by either (a) pressing the joystick button shortly or (b) pressing the “Acquisition” button on the touch screen, make absolutely sure to observe the OCT live image to the bottom of the screen. Always keep the scan as much horizontally aligned as you can. While the device is acquiring several images to obtain a good signal-averaged image, you may need to continue adjusting the live image with tiny up/down/right/left movements of the joystick.
Once the image is complete, continue with the ONH ring scan.
The ONH Ring Scan
First, click on “RNFL-N” (indicating the ONH ring scan) from the preset buttons. Adjust the ART according to the patient’s abilities and image post-processing needs. Higher ART means longer scanning time.
Second, ask the patient to fixate on the new target, which is now closer to her nose.
Third, once the ONH is clearly visible in the cSLO window and there is also a good OCT signal, start the device-specific “Eye-Tracker” function as described previously.
Fifth, start the OCT image acquisition by pressing the device-specific button as described earlier.