Skip to main content
Loading Explore guides…
Five Core Imaging Modalities · Acedra
Home
Explore
Medicine
Five Core Imaging Modalities
Sign in to save
Chat
If YouTube asks you to sign in, open the video on YouTube.
Watch on YouTube
Reload player
Transcript
36 segments
Copy
Hide video
0:02
We're going to look at the five core imaging modalities in modern medicine. Conventional X-rays, CT, ultrasound, MR, and PET. For each modality, we'll explore why it works and how that shapes the modality strengths and weaknesses in clinical practice. Let's start with conventional X-ray images. Think of an old incandescent light bulb. Electricity heats a thin filament inside a glass bulb until it glows, giving off light. An X-ray tube works much the same way. The only differences are that the
0:39
photons it produces are X-rays, not visible light. The filament is sturdier. The glass bulb design is different. And the photons are fired at a sheet of film or digital detector panel with a patient in between. which for the sake of this diagram I'm illustrating as just a couple of atoms with pink nuclei and blue electrons. Whenever an X-ray photon enters the body, there are three potential outcomes. One, the X-ray photon may pass straight through. Atoms are mostly empty space,
1:18
so many photons make it to the detector untouched. Two, the X-ray photon may collide with an inner shell electron of an atom somewhere in the patient's body, transfer all its energy to eject it and then disappear. We call this photo electric absorption and these kind of events contribute to image contrast. Three, the X-ray photon may hit a loosely bound outer shell electron, transfer part of its energy to eject it and is deflected with reduced energy. We call this compton scatter and these kind
2:01
of events can degrade image quality or add dose elsewhere. An X-ray image records the outcomes of billions of these sort of X-ray photon patient interactions and is essentially a 2D projection of electron density. Higher atomic number tissues pack electrons more densely, absorbing more X-ray photons and appear white on an X-ray image, while lower density tissues absorb less and appear darker on an X-ray image. In practice, electron density usually matches tissue density
2:42
closely enough for us to think of X-ray images as a tissue density map, though it's not absolutely perfect. um as say in the case of calcium versus aluminum. Conventional X-ray imaging excels in settings where there are large density differences such as when we're looking for pneumonia or lung masses against a background of airrated lung. Plural eusion adjacent to lung. Pumathorax adjacent to adelctatic lung. fractures, dislocations, arthritis or prosthetic alignment relative to cortical bone.
3:27
Bow gas patterns against a background of mezenteric fattened soft tissue and metallic foreign bodies against a background of soft tissue. Conventional X-ray images tend to struggle in settings with minimal density contrast, such as when we're looking for small or deep soft tissue masses against a background of other soft tissues or cartilage, ligament, and tendon injuries.
3:58
CT imaging. A big problem with conventional X-ray images is superimposition. different structures superimposed on each other in the same 2D image. CT imaging solves this by taking many X-ray projections from different angles and using a computer to combine them into a 3D volume. Here's how it works. Let's say we have an object. You can figure out what it looks like by shining a flashlight on it from different angles. Each time you shine the light, you see a shadow on the wall. Now, if you only
4:39
have one shadow, you don't get the full picture. But if you shine the flashlight from many different angles and look at all of these shadows, you can piece together what the object looks like. Unlike a flashlight which only gives silhouettes, X-ray images or X-rays penetrate tissues to varying degrees, providing grayscale data from which computer can reconstruct not just the surface but the entire 3D volume inside a patient's body. CT imaging tends to excel in settings
5:20
where density differences exist, but overlapping structures might have obscured them on a plain X-ray, such as with lung nodules and other lung prank opacities, small plural eusions, small or complicated pneumthoroses, complex or subtle fractures, bowel obstruction with identification of transition points, and small lymph nodes. With iodinated IV contrast, tumors and metastases that were practically indistinguishable from their background organs may stand out in contrast, as may
6:02
vascular diseases such as pulpy emblei and aortic dissections. That being said, there are some settings where CT imaging struggles and where iodinated IV contrast and eliminating tissue superposition still isn't of enough help, such as in cartilage, ligament, and tendon injuries and early eskeemic stroke.
6:28
Ultrasound imaging at its simplest, ultround images are created by sending out sound waves and detecting returning echoes. Two key physics concepts are essential. The first concept is attenuation. Sound weakens as it travels, like headlights fading in fog. Attenuation becomes a challenge when the sound path is long, as in obese patients, deep portions of organs like the liver and kidneys, and with deep blood vessels like the iliac arteries and veins, but it's rarely an issue for superficial
7:12
anatomy like the testicles, thyroid, and breast or for fluid fil structures close to the skin. The second concept is acoustic impedance. How much a material resists sound travel, like a cold stick of butter you're trying to pass a knife through. Acoustic impedance is usually measured in units of mega rails. Most tissues have similar though not identical values, but air and cortical bone are marked outliers. At a boundary between materials with different impedances, part of the
7:53
soundwave reflects back, creating an echo, and part continues onward. Small impedance differences are ideal, enough for echoes, but not so large that deeper imaging is blocked. Cordobone poses a challenge however due to its very high impedance and consequently very high impedance mismatch with surrounding soft tissues. Sound is almost completely reflected when it encounters a tissue bone interface and any transmitted sound is quickly attenuated creating a complete
8:35
shadow beyond the bone superficial margin. Air poses another challenge. Air's extremely low impedance creates a huge impedance mismatch with surrounding soft tissues too, resulting in near total reflection. Not only that, air pockets, especially small bubbles, tend to also scatter sound waves in all directions, which can produce pronounced imaging artifacts, too. As a consequence, regions of the abdomen containing bowel segments are often difficult to assess and why organs
9:12
like the pancreas can be challenging to visualize on ultrasound. Ultrasound cells in settings with gentle impedance changes. Good examples include fluid-filled structures such as gallbladder, urinary bladder, and cysts whose margins are sharply defined due to their modest contrast with surrounding tissue. Cysts being homogeneous with few internal impedance boundaries appear anacoic whereas solid lesions contain multiple subtle boundaries that contain that produce internal echoes which makes
9:48
ultrasound great for distinguishing solid from cystic lesions in organs like the liver, kidneys and ovaries. Ultrasound is also incredibly useful at characterizing fetal anatomy too. Conversely, ultrasound performs poorly in regions with large impedance mismatches such as the lungs and bowel where air scatters sound waves and near cortical bone which reflects sound almost entirely.
10:25
MRI. MRI works by exploiting the magnetic properties of hydrogen protons. Here's how it works. Let's start with hydrogen protons. Here's a grid of 45 of them. Hydrogen protons are everywhere in the body, especially in water and in fat. Each hydrogen proton acts like a tiny bar magnet. I've drawn them on this image as all neatly aligned. However, without any external force, they actually point in random directions, randomly oriented with no overall directions. This is the natural resting
11:04
state of the body. When we apply an external magnetic field like in an MRI scanner, most of the hydrogen protons stay randomly oriented. However, a small fraction of protons align with the external magnetic field. In this image, six have aligned parallel with the external magnetic field and two have aligned against or antiparallel to it. The tiny difference between these groups is what will help give us an MRI signal. Now we send in a radio frequency pulse like broadcasting a song only protons
11:48
can hear. This tips the eight aligned protons into the horizontal plane and in sync with each other. Once the pulse stops, the protons begin to relax. They fall out of sync with each other. We call this defasing and they gradually return to their original positions. As they defase, the protons meet emit weak radio frequency signals we can pick up and use to create an MR image. The six parallel protons go back to pointing up and the two antiparallel protons point down just like before the pulse.
12:31
Now, here's the trick. In MRI, we can choose when and how we listen. We can listen early
12:42
or we can listen later. Depending on when and how we listen, we highlight different tissue characteristics. These are called imaging sequences. MRI is like tuning a radio. All the protons are broadcasting signals, but each sequence chooses a different time and frequency to tune in. T1 listens early, T2 listens late, and some sequences like stir or flare use filters to block out background noise. Let's take a look at how this plays out in the brain. Even though we're imaging the same
13:18
brain, each imaging sequence reveals a different part of a story just by changing how and when we listen. On T1 weighted imaging, we tune in early. Fat is loud and bright. Fluid is quiet and dark. T1 weighted images are good for anatomy and postcontrast imaging. On T2- weighted imaging, we tuned in late. Water sings much louder relative to other substances here. So, CSF and edema show up bright. T2- weighted images are good for spotting edema, cysts, and inflammation.
13:54
On flare, we have the same late timing as T2, but with the volume turned down on the CSF station. This lets us pick up subtle signals near the ventricles in some patients brains, like perventricular edema or demolination. On diffusion weighted images, we're not just listening, but we're watching how the signal changes when protons try to move. In places where they're trapped, like early stroke, the signal spikes. Here's our external magnetic field from before. But what if a metallic object gets
14:33
introduced? It turns out that metal, not surprisingly, distorts the magnetic field, producing signal voids and distortion. And even titanium can cause significant artifact and degrade image quality. So even though there are some implants that are safe to MR, they can still cause substantial image degragation. We refer to as a susceptibility effect. In addition to MRI's superb ability to depict fluid containing structures and soft tissue edema, it offers exceptional soft tissue
15:13
contrast enabling detailed characterization of organs and tumors, often distinguishing um between different tissue types and tumor components such as solid, cystic or necrotic areas. MR imaging is invaluable for brain and spinal cord imaging detecting demination, edema, infarct and hemorrhage as well as for muscularkeeletal assessment of cartilage, ligaments, menisc tendon and bone marrow. MRI is poor at detecting calcifications or fine cortical bone detail as bone
15:46
appears dark due to its low proton density. Airfield structures like lung and bowel yield little or no signal and motion from respiration can cause artifacts. Peristtoalsis introduces additional artifacts unless specialized protocols are used. Metallic implants or foreign bodies may cause pronounced susceptibility artifacts that can severely degrade image quality. Head imaging. Head imaging is based on introducing radionucleides into the body. Radionucleides are unstable atoms with
16:25
an imbalance in the ratio of protons to neutrons in their nuclei. The radionuclides we use in PET imaging can become stable, however, by allowing a proton in their nucleus to transition into a neutron by emitting a posetron. Posetrons are the antimatter counterpart of electrons. They have the same mass as an electron, but carry a positive electric charge. Now, there's always a swarm of electrons buzzing around the nucleus of an atom. And whenever a posetron happens to encounter one of
16:58
these electrons, they annihilate each other, which results in the conversion of their respective masses into energy, typically in the form of two gammaray photons traveling 180° away from each other. In positron emission tomography or PET imaging, we introduce radio tracers into the body that preferentially accumulate in particular environments of clinical interest and then we acquire a three-dimensional image map of the distribution and concentration of the
17:28
radar tracer within the body. We create this image by surrounding the patient with a ring of gammaray detectors. Whenever two detectors on opposite sides of the ring detect gamma photons at the same time, we know that a posetron electron annihilation event has occurred somewhere along a line between those two spots. Several billion of these sorts of events occur and get recorded during a typical PET scan which allow us to reconstruct a visual map of where the annihilation events mostly occurred
18:02
using a mix of integral geometry, linear algebra and numerical algorithms. So how do we get these unstable atoms or radionucleides into the body and to preferentially accumulate in environments that have specific clinical relevance. The most common method relies on the normal mechanism cells in the body use to metabolize glucose. Special proteins in the cell membrane called glucose transporters move glucose molecules from the bloodstream and into the cell. Once inside the cell, the
18:38
glucose molecule is immediately phosphorolated by an enzyme called hexocinase. This means that a phosphate group has been attached to the glucose molecule, converting it into glucose 6 phosphate. Since the glucose transporter in the cell membrane happens to work in both directions, this important step traps the glucose inside the cell. Converting glucose into glucose 6 phosphate also prepares the glucose for energy extraction via metabolic pathways such as glycolysis and the KB cycle.
19:11
Our most common strategy to get unstable atoms or radionucleides into the body relies on using a decoy that looks just like glucose to the cell. The decoy is a glucose molecule that simply had one of its hydroxal groups replaced by an unstable florine 18 atom. Florine 18 atoms have an excess of protons but can achieve stability by allowing a proton inside their nucleus to emit a posetron and become a neutron. We call this decoy molecule flu deoxylucose or FDG.
19:45
Glucose transporters will dutifully move FDG from the bloodstream and into the cell where hexocinase immediately phosphorolates it into FDG6 phosphate. However, the replacement of a hydroxal group by a florine atom prevents enzymes involved in glycolysis and other metabolic pathways from recognizing and processing the FDG6 phosphate molecule. This results in an accumulation of FDG6 phosphate within the cell which is particularly pronounced if your cell is hyper metabolic. And since all those
20:19
florine atoms are unstable florine 18 atoms, there'll be gammaray photons for you to detect from the cell coming from all the posetron electronilation events happening inside. PET imaging's versatility comes from the wide range of radioces we can use. each designed to mimic or target a specific molecule in the body. The workhorse is FDG which accumulates many tumors infections and also can be used to identify areas of brain involvement in dementia. Sodium fluoride is excellent for
20:52
detecting many bone metastases. PSMA targeted tracers labeled with either gallium 68 or florine 18 bind to a transmembrane protein overexpressed in prostate cancer. Somatastatin analoges like dotate localize neuroendocrine tumors and tracers such as acetate can target hpatoscellular carcinoma. This molecular flexibility is PET's greatest strength. But it's also why its performance depends heavily on the right tracer for the right clinical question.
21:34
X-ray images are accessible, widely available, and very low in cost. CT entails much higher ionizing radiation doses to patient um than X-rays. Uh often requires the use of IV contrast and is moderate in cost. Ultrasound is probably the most operator dependent imaging modality here, but low in cost. MRI studies can take a long time to perform and are high in cost. Finally, PET imaging, which is often combined with CT due to its poor anatomic detail, is very high in cost.
Here to help you learn
Sign in to keep a private conversation with your course tutor.
Ask me anything about the material…
Sign in to chat