Chest CT Protocols & HRCT
- A chest CT "protocol" is just a recipe: when to inject contrast, how thick to slice, and which reconstruction to use. Get the recipe wrong and even a perfect scanner gives you a useless picture.
- A routine contrast-enhanced chest CT answers "what's in the chest?" — nodules, masses, nodes, effusions.
- A CT pulmonary angiogram (CTPA) times the contrast to flood the pulmonary arteries, hunting for clot.
- High-resolution CT (HRCT) trades coverage for crispness: thin slices and a sharp algorithm to show the fine architecture of the lung, mostly for diffuse lung disease.
- The lungs are always read on a wide lung window; the soft tissues on a narrower mediastinal window. Same data, two completely different pictures.
Here's a secret that took me embarrassingly long to learn: the scanner is the easy part. You can have a multi-million-dollar machine and still hand the radiologist a worthless study if you pick the wrong protocol. The protocol is the recipe — when do we inject the dye, how thin do we slice the bread, and how do we sharpen the photo afterward. Same oven, wildly different cake.
Let me walk you through the recipes you'll actually see ordered, and why each one exists.
The everyday workhorse: routine contrast-enhanced chest CT
This is the bread-and-butter "look at the chest" exam. You inject iodinated contrast into an arm vein, wait until it has filled the blood pool, and then scan. The contrast makes vessels, lymph nodes, and the walls of things light up bright white, so a lymph node stops hiding next to a vessel and a mass stops blending into the heart.
We use this for staging cancer, chasing infection, characterizing a mediastinal mass — basically anything where you need to tell soft tissue from soft tissue. The whole chest is scanned in a single breath-hold, usually a few seconds. Hold your breath, the table glides through the donut, done.
"With contrast" and "without contrast" are not interchangeable, and they're not the same as the timing recipes below. A radiologist can usually spot an effusion or a big mass on either. But subtle node enhancement, vessel filling, and abscess walls live or die on the contrast — and on getting it to the right place at the right second.
When you're hunting clot: the CT pulmonary angiogram
A pulmonary embolism is a clot lodged in a pulmonary artery, and to see it you need the artery glowing brilliant white with clot showing up as a dark gap inside that glow — like a pebble silhouetted in a lit-up garden hose. The trick is timing. Inject too early or too late and you've photographed the hose a second before or after the water arrived: gray, unhelpful, non-diagnostic.
So a CTPA uses fast injection and clever timing (the scanner watches the contrast arrive and triggers the moment the pulmonary arteries light up) to catch the arteries at peak brightness. It's the same body part as a routine chest CT, but a different stopwatch.
A "routine" chest CT is NOT a rule-out-PE study, and a CTPA is timed for the arteries, not the rest of the chest. Ordering the wrong one means either a non-diagnostic clot hunt or repeating the scan and the radiation. If the clinical question is "PE?", the order needs to say so.
The crisp specialist: high-resolution CT (HRCT)
Now for the one that confuses everyone. HRCT is not a different scanner and usually not a different dose of contrast — in fact, for diffuse lung disease it's often done without contrast. HRCT is a way of reconstructing the data to see the lung in extreme detail.
Two things make it "high-resolution":
- Thin slices. Instead of viewing the lung in fat slabs, we view it in very thin sections. Thin slices resolve fine structures that thick slices blur together — like the difference between a single sheet of paper and a stack of fifty.
- A sharp reconstruction algorithm. The same raw data can be rebuilt to look smooth or to look crunchy-sharp. HRCT uses the sharp ("high-frequency") algorithm that emphasizes fine edges, so the delicate lung architecture — the secondary pulmonary lobules, fine lines, tiny cysts — snaps into focus.
The payoff is pattern recognition. HRCT is the tool for interstitial lung disease: the honeycombing, reticulation, ground-glass, and traction bronchiectasis that let you name the pattern live in those thin, sharp images.
Two extra HRCT tricks worth knowing: prone imaging (scanning the patient face-down) separates real early posterior fibrosis from harmless gravity-dependent atelectasis that vanishes when the dependent lung is on top; and expiratory imaging (scanning after breathing out) reveals air trapping — areas that stay stubbornly dark and lucent because air can't escape through small diseased airways.
The window matters as much as the scan
Here's the part that mystifies trainees: the lung and the mediastinum are read from the same scan data, just displayed differently. CT measures density in Hounsfield units, and a "window" is simply which slice of that gray-scale ruler you stretch across black-to-white.
| Window | What it's tuned for | What you read on it |
|---|---|---|
| Lung window | Very dark, air-filled lung | Nodules, emphysema, HRCT patterns, pneumothorax |
| Mediastinal (soft-tissue) window | Narrow range around soft tissue | Heart, vessels, nodes, masses, effusions |
| Bone window | Wide, tuned for dense bone | Rib, spine, sternal lesions and fractures |
Always look at the lungs on lung window and the soft tissues on mediastinal window. A nodule can be glaringly obvious on one and nearly invisible on the other. Flipping windows isn't optional polish — it's reading two studies for the price of one acquisition.
The one thing to remember
The scanner doesn't decide whether you find the answer — the recipe does. Match the protocol to the question: routine contrast for "what's in the chest," timed CTPA for "is there a clot," and thin-slice sharp HRCT for "what's wrong with the lung tissue itself." Pick the wrong recipe and you'll bake a beautiful, perfectly useless cake.