Rethink Imaging

EP 56 • October 29, 2026

Prospective vs. Retrospective Gating in Coronary CT: Dose, Image Quality, and Workflow Trade-offs

Featured Guest

Dr. Ari Goldberg, MD, PhD Medicine,

Radiologist • Loyola University Medical Center

Listen Now36 min

Also on:

Let's Talk

Beyond dose

Ask us what your imaging data can actually show you.

Start a Conversation

Share This Episode

Subscribe to Rethink Imaging

Coronary CT has expanded rapidly as CT-FFR gives clinicians a non-invasive way to assess whether coronary stenosis is hemodynamically significant. Host Chris St. John sits down with returning guest Dr. Ari Goldberg to examine how acquisition choices affect image quality, radiation dose, and workflow.

Dr. Goldberg compares prospective gating, which images a narrow window of the cardiac cycle, with retrospective gating, which captures data across the cycle with dose modulation. The discussion covers motion artifacts, heart-rate management, HeartFlow requirements, scanner advances, photon-counting CT, interpretation workload, and the need to balance diagnostic reliability with radiation risk.

Host

Chris St. John

Host, Rethink Imaging • Imalogix

Featured Guest

Dr. Ari Goldberg, MD, PhD Medicine,

Radiologist • Loyola University Medical Center

Watch the Episode

Key Takeaways

CT-FFR has helped drive coronary CT adoption by providing non-invasive hemodynamic information that can reduce unnecessary diagnostic catheterization.
Prospective gating limits acquisition to a narrow window in diastole, lowering overall radiation dose but increasing the risk that motion will compromise the scan.
Retrospective gating captures data across the cardiac cycle with dose modulation, improving flexibility and image reliability while creating more data to review.
Modern scanners, motion-correction software, wider detectors, and photon-counting CT continue to narrow the dose and image-quality trade-offs between protocols.
The best protocol depends on clinical context, scanner capabilities, patient heart rate, motion risk, and the downstream requirements of tools such as HeartFlow.

Full Transcript

Rethink Imaging Podcast Transcript

Guest: Dr. Ari Goldberg

Host: Chris St. John

ARI GOLDBERG 00:00:00 It puts even more of a premium now, in my opinion, to have the best images that we can. Because you can’t create magic, you can’t send to the heart flow technology the significantly motion corrected images and expect to get back some magic answer about whether those lesions are truly hemodynamically significant or not. Garbage in, garbage out. Good data in, good data out.

CHRIS ST. JOHN 00:00:29 Welcome back to another episode of Rethink Imaging. Today I am joined by an incredible radiologist and a great guy, Dr. Ari Goldberg. He’s been on the show before. Ari, thank you so much for coming back. It’s great to talk to you again.

ARI GOLDBERG 00:00:51 Thank you for having me. It’s always great.

CHRIS ST. JOHN 00:00:53 Today we’re looking to get into this gating debate going on in coronary CT. Before I clicked the record button, I was asking for a little bit of clarity in language. I asked whether I should call it coronary CT or cardiac CT. You started to answer and I cut you off, but I would love to take that question again now that we’re live.

ARI GOLDBERG 00:01:24 Honestly, for the purposes of the debate retrospective versus prospective that would be a coronary CT. But more broadly, we can just talk about cardiac CT. Coronary CT is a subset of cardiac CT.

CHRIS ST. JOHN 00:01:45 Coronary CT has been around for a while, but more recently it seems like it’s really taken off in a big way.

ARI GOLDBERG 00:01:53 There are a few reasons for that. Cardiac CT and coronary CT have been around for more than 20 years maybe 25 but it really has taken off in the last several years. In part, that’s because of an accompanying technology that has gained a lot of traction called CT-FFR.

FFR stands for fractional flow reserve, which was originally only done in the cath lab. They would see a lesion or a stenosis, put a catheter past it, and measure the pressure before and after the lesion to determine the fractional flow reserve ratio.

Some very clever people from Stanford developed proprietary models based on fluid dynamics. They take a CT scan of the coronaries as the structural anatomy and roadmap, use the Hounsfield units the density of the contrast and accurately perform a virtual FFR without needing the patient to be in the cath lab.

CT-FFR is the generic technology, and the company that spun off from that is HeartFlow. HeartFlow is sort of like Kleenex or facial tissue in this context. What HeartFlow did was take a lot of ambiguity out of coronary CT. When I call a lesion that’s 40% stenosis, I don’t really know if that has hemodynamic significance. We use rules like 70% or more being very hemodynamically significant and 25% being less, but it depends on what’s upstream, downstream, and the specific anatomy.

CT-FFR technology reduces that ambiguity significantly. We are now able to much more accurately guide treatment and determine if someone needs a trip to the cath lab. That makes everybody happier, including insurance companies who were resistant at first because they didn’t like paying for cath lab trips where lesions turned out not to be significant. It works for all parties involved, and that’s a big reason why cardiac CT has taken off.

That technology has been around for about 13 or 14 years. I was fortunate to be one of the first implementers when I was at Loyola outside Chicago; I think we were the second site to implement it. Being a long-time user and believer in it, seeing it gain traction over recent years has changed how accurately and comfortably we can use the technology.

CHRIS ST. JOHN 00:06:11 The short version that I’m hearing is that this is a less invasive, faster way to gather the data points we need to determine which direction we need to go moving forward.

ARI GOLDBERG 00:06:25 Said perfectly.

CHRIS ST. JOHN 00:06:27 Within the center of that debate, radiation dose comes into the conversation. We’re exposing people to radiation. Can you talk about the considerations of implementing dose in prospective versus retrospective gating, the different doses associated with each, and the general tension between people taking these two approaches?

ARI GOLDBERG 00:07:08 In general, as technology has advanced, the absolute dose numbers and differences have gotten smaller.

When you do prospective gating, you are only capturing information about the heart at a certain point in the cycle. It’s not infinitesimal because we use padding. We prefer to image the heart when it’s most still, usually between 65% and 75% of the cardiac cycle during diastole. We don’t capture a single mathematical point; we capture a finite window, which is the padding. In prospective gating, we capture maybe 10% of the cardiac cycle.

CHRIS ST. JOHN 00:08:02 You’re gunning for a very brief moment in time.

ARI GOLDBERG 00:08:06 Exactly. That specific acquisition mode carries a bit more radiation per unit of time, but because you’re only capturing a small percentage of the cardiac cycle, the overall dose is less.

The downside is that you risk missing potential data. We try our best to capture the heart when it’s still, so we play games with protocol: we want the heart rate as close to 60 as possible and very regular. If a patient comes in with a heart rate of 84 and irregular, we administer Metoprolol, a beta-blocker, to slow their heart rate down.

CHRIS ST. JOHN 00:09:09 You’re not just doing sound baths and guided meditations.

ARI GOLDBERG 00:09:12 Exactly. It’s a whole process.

The opposite approach is retrospective gating, where you acquire data throughout the entire cardiac cycle and retroactively correlate it with the EKG. You can go back and see what the heart looked like at the 10% phase, 50% phase, and so on.

In theory, we would love to have full data, but having the beam on full power for the entire cardiac cycle is generally not necessary and the risk-benefit ratio doesn’t support it. Almost all retrospective gating is done with dose modulation. The dose is highest during diastole when we need the clearest images, while the rest of the cycle receives lower dose. That data is a bit noisier, but it’s still available if we need to piece together coronary anatomy that might be obscured on a prospective scan.

That tension is the origin of the debate. Half the places I’ve worked follow prospective gating, accepting that a certain percentage of scans will have motion corruption in exchange for minimizing dose exposure.

CHRIS ST. JOHN 00:11:27 I envision prospective gating like trying to hit the shutter button on your phone at the exact split-second someone jumps off a cliff, whereas retrospective gating is recording a video and grabbing the exact frame you need afterward.

ARI GOLDBERG 00:11:46 That’s exactly right.

CHRIS ST. JOHN 00:11:50 I’m curious about the timing and automation involved in prospective gating. Hitting the button at the right split-second with electronic delays seems complicated. How does that typically go down when the scan is happening?

ARI GOLDBERG 00:12:20 Because it’s a dynamic system, it involves preparation. The technologist notes the heart rate and trains the patient beforehand to see how their heart rate responds when they take a deep breath. Prospective gating involves some art because you’re trying to capture a single moment amidst patient motion and natural cardiac variability.

It comes down to two camps: do we live with some resulting motion artifacts, or do we deliver more dose to get full data and live with less variability?

Personally, while it’s important to be cognizant of dose exposure and adhere to guidelines, this debate has shrunk somewhat because modern CT technology has significantly reduced absolute dose numbers, even for retrospective gating.

CHRIS ST. JOHN 00:14:00 I know dose-reduction technologies like photon-counting CT or dual-energy CT involve larger gantries. Are all of those technologies at play for these procedures as well?

ARI GOLDBERG 00:14:27 That’s right. Gantry rotation speed and detector width allow us to capture more data in a single spin. All these lower-dose and motion-tolerant techniques are in play, though not every scanner has them. There is a huge curve of legacy equipment out there.

If someone is using a legacy scanner without modern dose-saving features, the debate is very relevant. But with up-to-date scanners, it becomes less so.

Furthermore, consider the patient population. We are evaluating serious disease that often presents with sudden death, typically in patients at an age where a minor radiation disparity won’t alter their overall life trajectory compared to the clinical benefit.

Technologies like HeartFlow CT-FFR put a premium on having the highest quality images possible. You can’t send motion-corrupted images to HeartFlow and expect an accurate answer about hemodynamic significance garbage in, garbage out. So while it’s not an irrelevant debate, we have to be specific about the clinical context.

CHRIS ST. JOHN 00:16:48 I spoke with a physicist who advocated strongly for prospective gating, feeling that retrospective gating wastes dose and that prospective scans can be executed cleanly with proper protocol. Where do you see the line in image quality when comparing prospective to retrospective gating?

ARI GOLDBERG 00:17:53 Theoretically, if you check every box, prospective gating can give you the coronary data you need. But it comes down to tolerance for imperfection.

Even at high-volume centers with experienced technologists and refined protocols, I would estimate that a solid 20% of prospective scans have enough motion artifact to be frustrating.

Several years ago, I worked with a major vendor to help develop their next-generation motion correction software, which arose directly out of the need to recover prospective scans affected by motion. But limited data can only yield so much; those algorithms might patch up another 5% of cases, leaving roughly 15% where you wish you had clearer data.

I lean toward the retrospective approach. However, that puts the responsibility on us to rigorously optimize our dose reduction techniques when conducting retrospective scans.

CHRIS ST. JOHN 00:20:40 Does prospective gating require more staff hands on deck due to the precise timing and heart rate management?

ARI GOLDBERG 00:20:58 No, staffing is the same. Retrospective scans often still require administering heart rate medications.

CHRIS ST. JOHN 00:20:00 In emergency room situations where a patient is stressed out and has an elevated heart rate, controlling timing for prospective gating seems particularly complicated.

ARI GOLDBERG 00:22:40 Retrospective gating faces similar challenges if an ER patient has a heart rate of 110. You’re still bound by the same operational infrastructure. It comes down to whether the ER team needs to give one dose or two doses of heart rate medication, taking 30 minutes versus 60 minutes. Those considerations are weighed when designing retrospective versus prospective protocols.

CHRIS ST. JOHN 00:23:40 Beyond basic coronary visualization, retrospective gating provides multiphase reconstruction and functional data. How clinically significant is that added flexibility?

ARI GOLDBERG 00:24:02 Having functional information can be valuable, but the vast majority of patients undergoing cardiac CT also receive an echocardiogram. Echo remains superior for evaluating valves and cardiac function due to its temporal resolution.

In my practice, we don’t place heavy emphasis on the functional data provided by retrospective CT because it’s slightly less accurate than echo and rarely necessary. That aspect is a minor part of the overall decision.

CHRIS ST. JOHN 00:25:30 Given the current radiologist shortage and burnout levels, shouldn’t facilities default to providing the cleanest, easiest exam to read to save interpretation time?

ARI GOLDBERG 00:25:45 It’s a good question. However, if you are provided retrospective data, there is a responsibility to review or analyze the functional data presented. More data requires more time to evaluate unless you have a dedicated 3D lab to process it for you.

While struggling with motion artifacts on prospective scans takes time, evaluating the extra datasets in retrospective scans also adds volume to the workload.

This touches on a fundamental question for radiology: as study volumes grow and the workforce ages, what is our medical-legal responsibility for all data acquired in a scan? Traditionally, if it was acquired, the radiologist had to interpret it. We may reach a point where we acquire comprehensive data but explicitly protocol and interpret the study for a single specific indication.

While HeartFlow was the pioneer with proprietary algorithms, other industry players like Siemens are developing AI-based methods to derive CT-FFR data. CT-FFR is not tied strictly to HeartFlow, but the concept itself is the single biggest factor driving the growth of coronary CT in recent years because it provides a non-invasive alternative to diagnostic cardiac catheterization.

CHRIS ST. JOHN 00:28:53 It redistributes patients by providing an evaluation step before sending them to the cath lab.

ARI GOLDBERG 00:29:09 Exactly. It allows us to be much more selective about who needs an invasive catheterization procedure.

Photon-counting CT represents the next frontier. It offers higher spatial resolution and cleaner images. One perspective is that higher resolution makes prospective gating even more viable; the counterargument is that higher-resolution scanners should be paired with retrospective gating to maximize overall data quality.

Five years from now, this debate will shift toward how best to utilize photon-counting capabilities. Until technology renders the differences negligible, this discussion will continue.

CHRIS ST. JOHN 00:31:41 It comes down to clinical risk versus radiation risk ensuring we gather diagnostic data while managing dose. Listeners who advocate strongly for prospective gating are welcome to reach out to the show.

ARI GOLDBERG 00:32:36 There is no single right answer; it depends on how one weighs risk-benefit ratios.

An impactful study could be conducted measuring interpretation time between prospective and retrospective cohorts balancing time spent working through motion ambiguity in prospective scans against time spent reviewing larger datasets in retrospective scans. As study volumes increase, optimizing that workflow is essential.

CHRIS ST. JOHN 00:35:07 That’s why we’re here. Dr. Ari Goldberg, thank you so much for joining us on Rethink Imaging. If you have thoughts on prospective versus retrospective gating, please write in to the show.

ARI GOLDBERG 00:36:02 Thank you for having me.

Questions from This Episode

What is the difference between prospective and retrospective gating in coronary CT?
Prospective gating activates the X-ray beam during a narrow part of the cardiac cycle, usually in diastole, to reduce overall radiation dose. Retrospective gating acquires data across the cycle with EKG tracking and dose modulation, providing more reconstruction options and reducing the risk that motion will leave the scan unusable.
How does CT-FFR influence coronary CT protocols?
CT-FFR estimates fractional flow reserve from coronary CT images to help assess whether a stenosis is hemodynamically significant. Since downstream analysis depends on clear structural data, teams must prioritize image quality and motion control when selecting and executing a gating protocol.
How do modern CT scanners affect the gating debate?
Wider detectors, faster gantry rotation, dose modulation, motion-correction software, and photon-counting technology can improve image quality and reduce dose. These advances narrow the gap between prospective and retrospective techniques, but the best choice still depends on the patient, protocol, and scanner.