I've worked in and around medical imaging for most of my career. In that time I've mostly worked around MRI applications and as a result had a number of MRI brain scans as a volunteer. You can do this with MRI quite safely as there are no known mechanisms by which the magnetic fields and radio-frequency pulses used in the machines can harm human tissue. In technical terms, MRI doesn't use ionizing radiation. Other common medical imaging techniques like X-ray, CT and PET do use ionizing radiation and there is a risk-benefit aspect to their use. Most people are very wary of "radiation" and tend to regard any exposure as far riskier than it actually is. This is probably due to associations with the nuclear arms race and accidents at places like Windscale, Three-Mile Island, Chernobyl and Fukushima - which isn't to say these weren't dangerous and highly undesirable events.
During my investigations for CLIPPERS I had CT brain and chest scans and PET head and chest scans as well as X-ray during lumbar puncture (twice). These are scans I wouldn't have had as a volunteer, because of the ionizing radiation dose and the fact that the risk from radiation wouldn't be balanced out by the benefit of diagnosing something nasty. However, I realised I had no idea what the relative risk of having a CT scan was compared with, say, flying to Spain, or how it might increase my risk of cancer later in life. So here's what I found. To start with, having one chest X-ray exposes you to the same amount of radiation the average American gets from their natural surroundings in 10 days (source: RadiologyInfo).
Yes, we are all exposed to radiation all the time.
Another version of this statistic reports the equivalent of 3 days background radiation exposure for a Brit or the equivalent of taking a return flight from London to Spain (source: Radiology for Undergraduate Finals and Foundation Years).
Yes, we get exposed to more radiation when we fly.
RadiologyInfo also points out that we have a 1 in 5 lifetime risk of dying from cancer and that a chest X-ray adds less than a 1 in 100,000 additional risk. So let's not worry too much about chest X-rays, or flying to Spain.
X-Ray Computed Tomography (CT) is a very clever imaging technique which works by taking X-rays from lots of different directions and then using a computer to figure out the internal structure of the body giving a set of slices. For the patient lying in the scanner, CT is quick, quiet, painless and not claustrophobic compared with MRI. Because it involves multiple X-rays the radiation dose is typically quoted as between 100x and 400x that of a conventional single chest X-ray (e.g. the FDA). For the kind of brain CT I had (angiography = with and without contrast injection) RadiologyInfo suggest a radiation dose equivalent to 16 months background and an additional risk of fatal cancer of around 1 in 1000 to 1 in 10000. This is similar to the exposure of commercial airline pilots. For high-resolution chest CT the dose is 2 years equivalent with similar risk.
Positron Emission Tomography (PET) is another clever imaging technique which makes use of radiation. Here a radioactive drug (FDG) is injected which behaves like glucose which human tissue uses as an energy source. The emitted radiation is detected by the scanner and again a computer generates slices through the body. So PET scans show where tissue requires the most energy (e.g. where there is active tumour growth). These days PET scans are usually combined with a special CT scan which makes the reconstructed images more accurate. It's hard to be precise but PET doses are comparable with those quoted for CT above (source: IAEA Radiation Protection of Patients). PET/CT doses can be up to three times higher than this but still comparable with the annual exposure limit for nuclear workers (source: NHS).
So what does alll this mean for risk practically? Consider 1000 unremarkable people who have never had a chest CT. On average 200 of them (1 in 5) will get a fatal cancer during their lifetime. Now consider 1000 different people who are similarly unremarkable except they have had a chest CT which gives an additional fatal cancer risk of at most 1 in 1000. On average 201 of the second group will get a fatal cancer during their lifetime but we won't know which person had their cancer triggered by the CT - just that there's a slightly higher occurrence than might be expected by chance.
So what does alll this mean for risk practically? Consider 1000 unremarkable people who have never had a chest CT. On average 200 of them (1 in 5) will get a fatal cancer during their lifetime. Now consider 1000 different people who are similarly unremarkable except they have had a chest CT which gives an additional fatal cancer risk of at most 1 in 1000. On average 201 of the second group will get a fatal cancer during their lifetime but we won't know which person had their cancer triggered by the CT - just that there's a slightly higher occurrence than might be expected by chance.
So my conclusion has to be that if I have one or two CT or PET scans I'm not going to raise my lifetime risk of cancer significantly. And if I do get cancer later on, how would I know whether it was triggered by the CT scan I had 20 years earlier? The answer is I wouldn't know and the only way these statistics can be derived is by carefully studying large numbers of people over a long time and comparing cancer rates in those people who have had CT or PET scans with those who haven't. That said, having unneccessary CT or PET scans is not sensible but minimising your lifetime exposure definitely is. For people who have several scans, it is suggested (by RadiologyInfo) that a record card be kept to avoid unncessary repeats. The bottom line is I'm not going to lose too much sleep over radiation exposure for now and feel less anxious about the CT scans and PET scan I had recently.
So I'm not Frying Tonight as much as I originally thought.
Read other articles in this series at Living With CLIPPERS.
Living With CLIPPERS by Bill Crum is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.
