‘MRI’ stands for magnetic resonance imaging. It is an extraordinary tool that allows doctors to look inside the human body. The MRI produces black and white pictures of the knee, shadows if you will. But, as terrific as it is, the MRI creates nothing more than thin, flat slices of a complex, colorful, three-dimensional structure. Imagine trying to recreate in your mind the shape of a funny-looking loaf of bread just by observing the slices on a plate. Therefore, MRIs are subject to interpretation. Doctors with varying degrees of experience and knowledge will read MRIs differently.

MRIs are imperfect in four ways:

MRIs can miss certain painful conditions

Arthritis. If the arthritic area is small enough, it will not show up on an MRI. This is a common source of patient disappointment: When patients go to an orthopedist complaining of knee pain, they might be told that they have ‘torn cartilage’, they expect that an outpatient operation will fix it, and, when they have persistent pain, the doctor explains that he found some arthritis. “How could that be? It didn’t say so on the MRI!”

Caveat emptor! If you are over 50 years old and you have knee pain you may well have an area of arthritis in your knee that will not show up on an MRI. A competent doctor should explain this to you before you embark on any procedure.

Ligament Tears. Following a major injury, a completely shredded ligament will be detected on any MRI. However, an ACL (anterior cruciate ligament) that has been pulled off its attachment and has reattached itself to a neighboring structure may be read as normal.

• Kneecap problems. The kneecap (patella) can be poorly aligned in many different ways. Most of the time, the malalignment will not be reported by the radiologist. In fact, a number of MRI centers do not even bother to take all the MRI cuts (views) necessary to judge the position of the knee cap. It saves money and few people complain. The groove that the kneecap rides in is called the trochlea. Occasionally, the trochlea will exhibit an abnormal shape, but this is not likely to show up in the report either. When you are trying to convince the insurance companies that you have a real problem, it doesn’t make life easier to be holding an MRI report erroneously read as “normal”.
• Serious bruises of the skin and nerves. If you strike your knee and painfully bruise your skin and the underlying nerves, this will not be revealed on an MRI. These conditions happen to people who hit their knees against the dashboard of a car, and it happens to people who fall onto their knees.
• Cartilage tears. Large tears can be picked up on an MRI but smaller tears can go undetected. Remember: the MRI only shows you thin shadows of complex structures, and therefore small abnormalities can be missed. Small tears can still be quite painful and occasionally require surgery.
• Any condition best detected with the patient in the standing position. Because the MRI is performed on a subject who is lying down, conditions that are best observed with the patient in the standing position may be missed. These would include certain forms of arthritis and kneecap malalignment.

At the other extreme, MRIs can overread conditions:

Torn cartilage. The radiologist and orthopedist may feel that something is torn, and yet at surgery the structure is found to be intact. The reasons for this are complex and have to do with “signal artifact”. Occasionally, one finds a little blob of white in the middle of the triangle, just as there might be a little flickering on your television screen. This little white blip is called an “artifact” or “signal change”.It can be due to the MRI hardware or software, and it can also be introduced by scarring from prior surgery. For example a meniscus (cartilage) which has been operated upon can still look torn on an MRI even if it is intact. This commonly leads to unnecessary repeat surgery.

One of the most questionable practices in Radiology is the reporting of normal signal changes in meniscal cartilage as “tears”. In the early days of MR imaging, these signal changes were graded from I to III. Grade I and II tears do not represent changes visible to the naked eye.

I was most shocked a few years ago when a radiologist showed me how he could simply dial in a Grade I or II “tear” by turning a knob on his MRI console!

This made it easier for me to understand how certain radiologists might read a “tear” on every MRI. By reporting Grade I and II changes as tears, the Radiologist allows the unscrupulous orthopedic surgeon to “sell” the patient a surgical procedure: Seeing the word ‘tear’ on the MRI report, most patients readily agree to surgery. In short, the MRI report has become a license to operate. The insurers don’t know any better and readily approve the procedure. A Grade III change, on the other hand, is more likely to represent a true tear. And even then, there can be “false-positives”, the scientific term for false alarm. It is the doctor’s responsibility to correlate the test results with the patient’s symptoms and physical examination.

A good doctor treats patients not tests.

Baker’s cyst. Over 100 years ago, Dr. Baker described a pouch of fluid commonly found at the back of the knee. The liquid is normal knee joint fluid which, in some people, seeps through a small opening and collects in this pouch. The pouch varies in size from person to person. It ranges from being invisible to feeling like a large ping-pong ball.. The swelling can come and go as the fluid moves back and forth from the cyst to the knee joint. Unless the cyst is particularly large, it is usually asymptomatic and requires no treatment.

Most significantly, a Baker’s cyst is readily seen on an MRI and it is commonly listed on the MRI report. Radiologists are perfectly justified in listing it. The problem – and this is why I’ve listed the Baker’s cyst in this section - is that every body but the patient understands that it is a common, harmless finding. Patients are understandably anxious at the thought of a cyst living around their knee. It is therefore not so much question of an over-read by the Radiologist as an over-interpretation on the part of the patient.