[00:19] Hello. My name is James Whittaker and I’d like to ask you to remove all the ferrous objects from about your person before I welcome you to another episode of ‘Conditional 1’ my occasional podcast focussing on all aspects of MRI Safety.
[00:33] This time I’m going to be talking about an aspect of MRI Safety that has always been one of my favourites, throughout the 20+ years of my practice. I think it’s always been a topic of special interest to me because it was the first part of MRI Safety that I encountered, before I even began training in MRI itself and was working as a Radiographer in the UK.
[00:56] I am of course, talking about Intra-orbital foreign bodies or IOFBs. I touched on this briefly last episode but today I’m going to delve deeper into the subject. In this episode I hope to achieve three things. First, I’ll give a brief overview of the relevant anatomy of the eye, and it’s associated structures. Then I’ll describe some of the more common types of calcifications found within and around the eye and how they appear in imaging. And finally, I’ll tell you about the basic method I use to simplify differentiating between them.
[01:32] But before I do any of that I want to explain why all of this is important in our practice. To do that I need to go back all the way to 1986, when two things happened. Firstly, in Somerset England, a man named Peter Gabriel released what is widely considered to be his most successful album, entitled ‘SO’. It brought us the hugely influential song and video Sledgehammer but more importantly it contained the song that shares it’s title with this Episode of Conditional 1 – ‘In your Eyes’. Is this important in the world of MRI safety? No, but I always think of it when discussing this subject and it’s a great song on a superb album and I wholeheartedly recommend giving it a listen.
[02:15] Of greater importance in 1986, Kelly et al. described in the American Journal of Neuroradiology an incident where a 63-year-old man attended for an MRI scan to further characterise suspected cerebral metastases seen on CT scanning. The scan was carried out without incident, but as the patient was being pulled from the scanner he experienced a ‘tugging’ at his left eye, a flash of light & a rapid decrease in vision. Later examination revealed a gross vitreous haemorrhage caused by a retained 2x3mm metal fragment and the patient had to undergo enucleation, meaning his entire eyeball had to be surgically removed. This man had previously been employed as a sheet metal worker and the fragment had been seen on his prior CT scan. However, the Radiologist felt that it was most likely a calcified granuloma from a previous parasitic infection.
[03:13] Following this original shocking incident there was a near universal improvement in the screening of MRI patients and I’m sure that all of you listening to this that work in MRI will be more than familiar with the screening practices in place at your hospital, clinic or workplace. There have been a handful of other described orbital injuries related to MRI, but these are thankfully exceedingly uncommon due to the constant vigilance of Radiographers, MRI Techs and Radiologists worldwide.
[03:42] So, just how common are IOFBs? In 2019 a team of Ophthalmologists published a ten-year national review of ocular injuries here in New Zealand. They found that the annual incidence of eye injuries was 1007 cases per 100,000 population, or just a fraction over 1%. That’s right, one in a hundred of everyone living in New Zealand has an eye injury requiring treatment or at least medical consultation, every year. The most common mechanism of injury was ‘struck by object’ in over 55% of cases and the use of protective eyewear was recorded in only 6% of cases. I would be interested to find out if similar studies have been carried out in other countries, so if anyone listening has any insight on that, please let me know.
[04:35] Now, to make sense of IOFBs and orbital calcification you must have at least a basic understanding of the anatomy of the eye and the surrounding muscles, so let me quickly run through a few anatomical terms that will be helpful moving forward. The cornea is the clear front portion of the eyeball that covers the anterior chamber of the eye containing the iris, pupil, and lens. The conjunctiva is the membrane that covers the sclera or white of the eye and lines the inside of the eyelid.
[05:05] Within the eye itself there are three main structures of interest to us. These are the retina, which is basically the inner wall of the eyeball. This is made up of light-sensitive cells called rods and cones which allow us to detect colours and shapes. The retina is separated from the sclera (the outer layer of the eye by a middle layer of vascular and connective tissue known as the choroid. Next is an area of the retina called the macula. This tiny area is responsible for our detailed central vision and sits at the back of the eyeball near the final structure of interest to us – the optic nerve. The optic nerve transmits all the information from the retina to the brain and passes through the retina via a tiny circular spot called the optic disc.
[05:52] There are two types of eye muscles: intrinsic muscles that control how much light enters the eye and control fine focus; and extrinsic muscles that control the position and movement of the eyeball. There are seven extrinsic muscles, though one of them, the levator palpebrae superioris doesn’t directly attach to the eyeball, instead raising the upper eye lid and maintaining it’s position. There are four rectus muscles that move the eye up and down and from side to side. These are the superior and inferior at top and bottom, and the medial and lateral on the inside and outside respectively. Finally, there are two oblique muscles that control rotation of the eye, the superior and inferior oblique. The superior oblique attaches to the sphenoid bone via a fibro-cartilaginous pulley-like structure called the trochlea at the top of the orbit. Remember that name, we’ll be coming back to it later.
[06:49] There are three types of intra-orbital calcifications commonly seen on CT scanning that can masquerade as IOFBs, with examples in the literature where each led to delay or cancellation of MRI scanning due to misdiagnosis.
[07:04] The first are drusen, deposits of protein or fat that can form within the retina. The word ‘Drusen’ comes from the German word for node or geode due to their sparkling mineral appearance on retinal photography. Small drusen are normal findings in the aging eye, with just under half of us (48%) having one or more by age 50. Drusen are normally not visible radiographically, but they can calcify and multiple drusen can coalesce to form larger deposits which canbe seen on CT scanning.
[07:39] A less common subset of drusen are optic disc drusen, which occur in 1 – 2% of the population. Usually these are asymptomatic and are found incidentally when the patient attends with another issue. However, they can lead to progressive peripheral visual field defects. Optic disc drusen can form uni- or bilaterally, with bilateral being more common. In fact, up to 70% of optic disc drusen appear bilaterally and can be seen on bone window CT, though they are more conspicuous on soft tissue window images. Optic nerve drusen are often on the surface of the optic nerve and are clearly visible on retinal photography but they can be ‘buried’, giving the optic nerve a swollen appearance.
[08:29] This is known as pseudopapilloedema as it mimics swelling of the optic nerve due to raised intracranial pressure or other causes.
[08:38] Next, we come to senile scleral plaques or SSPs, which are degenerative areas of calcification within the sclera of the eye. They appear as grey discoloured oblongs or elliptical areas that commonly sit just anterior to the insertion points of the horizontal rectus muscles, a few millimetres away from the cornea. As the name suggests, SSPs grow more common with age and are found in 8 – 22% of the population. Being degenerative, SSPs are often thinner than surrounding sclera. However, as they often calcify, they paradoxically appear thicker on CT. SSPs may be unilateral (one in one eye), unilaterally bilateral (two in one eye), bilaterally unilateral (one in both eyes) or bilaterally bilateral (two in each eye). They are most likely to be confused with IOFBs when unilateral.
[09:38] Now, finally, do you remember the trochlea, the fibro-cartilaginous pulley that the superior oblique muscle hooks around? It can calcify, though nobody really seems to know exactly why. It grows more common as we age, being an incidental finding in 7 – 13% of CT examinations and in younger patients it appears to have a possible link to diabetes. In 50 – 75% of case, trochlear calcification appears bilaterally and can have a range of appearances. Most usually it calcifies in a ‘comma’ shape, making up 2/3 of cases, though it can also calcify in a dot or inverted ‘U’ shape.
[10:20] There are also a range of other tumours, diseases and inflammatory processes that can cause intra-orbital calcification though these are significantly less common.
[10:32] Retinal astrocytic hamartomas are benign glial cell tumours that can form anywhere within the retina. They are thought to be congenital but are often not found until later in life, though they are more commonly found in patients with tuberous sclerosis. They appear very similar to drusen on CT.
[10:59] Choroidal osteoma is a benign tumour where the middle layer of the eye between retina and sclera is replaced by bony tissue. Choroidal osteomas are most commonly found in young Caucasian females and often result in gradual, painless visual loss. In 75% of cases, these are unilateral and appear on CT as curvilinear or crescent shaped densities on the posterior aspect of the eyeball.
[11:16] Retinoblastoma, which is the most common form of intra-ocular tumour in children, can present as a heterogenous mass of calcified and necrotic tissue with floating debris and haemorrhage. Retinoblastomas can appear bilaterally though in two thirds of cases they appear unilaterally. Most cases are diagnosed by age four, which should limit exposure to welding, angle-grinding, and metalworking. That said, I did an online image search for children welding while researching this piece, and the results were frankly terrifying.
[11:51] Ok, so our patient presents with either a strong history of occupational exposure to high-speed metallic fragments or a history of previous injury, and IOFB needs to be excluded before MRI can proceed. How this is done will very much depend on the policies of your workplace and in my experience, digital radiography is usually the most accessible modality and is generally more than adequate for this purpose. When I was training back in the days before DR, and even CR I remember working at one site that used non-screen film for orbital radiography. I shudder to think of the dose those poor patients sustained.
[12:30] However, before x-raying your patients’ eyes, always ask if they have had previous imaging of their orbits. I can’t tell you how many times in my practice a patient has had orbital radiographs because their referrer thought it was necessary even though they have recent CT head imaging that included the orbits. Always check their imaging history in case because every little helps.
[12:55] When it comes to reviewing previous imaging to determine whether what you are seeing is an intra-orbital foreign body or a calcification this almost entirely relates to CT, as the minute calcifications we have discussed previously are entirely invisible on radiographs as they share the same density as the surrounding bone. If you see a density on your radiograph, it is overwhelmingly likely to be a foreign body.
[13:20] Obviously, MRI can be used to differentiate between calcium and metal very effectively, but I would definitely not recommend it. One high point of my career was receiving a referral for an MRI of the orbits to determine if a previously visualised foreign body was ferrous or not. Explaining to the Emergency physician why that was not a good idea took entirely too long.
[13:44] When trying to differentiate between IOFBs and benign intra-orbital calcification I use a checklist method to determine whether what I am looking at is likely to be contra-indicated for MRI or not. I call it the LASSH method and it stands for Location, Attenuation, Shape, Symmetry and History. I find this mnemonic works well for making sure nothing gets missed out. You could swap the order you approach things and think about History, Attenuation, Shape, Symmetry and Location or HASSL but I find that to be more trouble than it’s worth.
[14:21] Just like in real estate, when it comes to IOFBs it’s all about Location, location, location. The 2019 study of New Zealand eye injuries by Wallace et al. found that 98% of IOFBs were corneal in location, with almost all the remaining 2% being conjunctival. Only a very tiny percentage of IOFBs actually penetrate the globe itself. When reviewing the imaging with your radiologist remember where the most commonly occurring calcifications are likely to be located – posteriorly for drusen, antero-laterally for senile scleral plaques and antero-superiorly for trochlear calcification.
[15:01] In X-ray terminology, attenuation is the term used to describe how much something blocks x-rays passing through it. It’s how image contrast works and is necessary for us to get a useable radiograph. In CT, density is measured in Hounsfield Units (HU) and the denser the object, the greater the attenuation and the greater the Hounsfield Units. Water is considered to have a density of 0 and forms the reference value for our scale. Air is considered to be the least dense substance on the Hounsfield scale and has a value of -1000. This means that 1 Hounsfield Unit equals 0.1% of the difference in density between water and air. Metal objects often exhibit a Hounsfield Unit value in excess of 3000. Particularly dense tissue or objects can cause a streaking appearance on CT images. This is because they block all the x-rays from passing through and can be thought of as the shadow that the object is casting on that area of the image and spreads out in rays as the x-ray tube and receiver circle around.
[16:09] One issue with CT is that images are often manipulated to show only narrow ranges of Hounsfield Units. A Head CT for instance utilises a narrow window width to maximise the image contrast between grey and white matter while still maintaining sensitivity for haemorrhage. White matter has a density of 25 Hounsfield Units and Grey matter is only 10 Hounsfield units denser. Any visual range that could include 3000+ Hounsfield Units to exclude metallic objects would have such a minute difference in appearance between grey and white matter that no Radiologist could diagnose from them. That’s why a wider bone window is our friend in this instance, as calcification will be visible with their density of 500 – 700 Hounsfield Units while metal will still be apparent. Use of a Region of Interest or ROI over the object will give the final answer and I’d be very surprised if any PACS in use today doesn’t have that functionality.
[17:08] Shape, or morphology can also be very useful in differentiating between a foreign body and calcification. Usually, calcification within the orbit has a soft and rounded appearance while foreign bodies can have any shape but are often irregular with sharp margins.
[17:26] With regards to symmetry, foreign bodies in the orbit are almost always singular or closely grouped within a region while calcification may develop symmetrically, and this can be a hugely useful diagnostic aid.
[17:40] When it comes to history, as I told you in the last episode, I don’t ask “Were you welding yesterday?” or “Have you got metal in your eyes in the last week?” No. “Have you EVER had any injuries or accidents where metal objects may have entered your eyes?”
[17:55] People are often as unreliable when completing health screening questions as they are in the witness box and the reason for this, to my mind at least, is quite simple. For us as MRI Techs, this is a vitally important question that we ask with full understanding of the consequences of making a mistake. To the patient we are asking them to recall a relatively inconsequential incident that may have happened many years prior, so it is unsurprising if nothing rises to the surface of their memory.
[18:25] I’ll finish with two cases that highlight how patient history can be an imperfect tool in the context of orbital foreign bodies. One come from the literature and the other comes from my own practice.
[18:38] In 2017, Platt et al. wrote a very interesting historical case report in the American Journal of Ophthalmology. They reported that in 2009 a then 10-year-old boy attended for a brain MRI in hopes of determining a cause for his neurodevelopmental delay. He had previously had MRI aged 5 without issue, and as was standard practice at the Canadian hospital he attended, his parents completed the pre-scan safety checklist. Of note, there was no history of orbital trauma or surgery.
[19:11] On the first localising scan a huge susceptibility artefact was demonstrated over the left eye and the scan was understandably abandoned for safety reasons. The 10-year-old reported no pain or discomfort during or after the scan. X-rays and a CT were performed which confirmed a foreign body within the globe of the eye adjacent to the optic nerve. The mechanism of injury and exact make-up of this object were never determined as it was decided that removing the object would pose a significant risk of damage to the optic nerve. A tiny traumatic cataract was discovered that the authors believe was the entry wound and was not due to the MRI scan.
[19:52] The other case is taken from my personal historical case file. A 59-year-old male attended for MRI brain after collapsing while intoxicated, with a multi-year history of previous collapse. Upon questioning he denied any history of eye injury or working with metal in any capacity. Despite this, a large area of magnetic susceptibility was immediately apparent on localiser images and the examination was terminated.
[20:20] As mentioned previously, this gentleman had an extensive previous history including 7 CT head scans over an 8-year period. Upon review with the Radiologist, it was determined that the small senile scleral plaque that had been visualised, but not commented upon, in all of his previous imaging was in fact the foreign body in question and was found to be 1800 Hounsfield units in density. Upon further questioning the patient ‘remembered’ being a panel beater at some time in the past. Unfortunately, he was lost to follow-up at that time, though I’m sure that was not the end of this gentleman’s story.
[20:59] So, there you have it. I hope that all this has been helpful, or at the very least interesting. I appreciate it is not up to us as MRI Techs to make the final call whether that density on CT is a foreign body or a calcification, but another pair of eyes is always useful and sometimes all it takes to prevent an accident is a willingness to speak up.
[21:25] I’d like to say thank you to Purpleplanet.com for the use of their music, and thank Jaenelle Whittaker for graphic design. If you have any questions about the content of this podcast, or ideas for future episodes of Conditional 1, please email me. My email address is podcast @ conditional1.com. You can also message me on the website www.conditional1.com.
[21:51] And remember, if anyone ever tells you that being an MRI Technologist isn’t Rocket Science, tell them No, but it is Nuclear Physics.