An insider's perspective on why ultrasound has limits in confirming ICD device specifications and what that means for medical device procurement and quality assurance.

I've reviewed over 200 product specifications this year alone. And every time someone asks about using ultrasound to verify ICD device quality, I have to pause. Because the question itself assumes something that isn't true: that ultrasound shows everything you need to know.

It doesn't. And assuming it does can cost you.

The Surface Problem: What Everyone Asks About

When procurement teams or new quality engineers come to me with ICD device verification questions, ultrasound is almost always their starting point. It makes sense—ultrasound is non-invasive, widely available, and relatively inexpensive compared to other imaging modalities. They want to know:

• Can we use ultrasound to confirm internal component placement?
• What resolution can we expect for checking lead connections?
• Is ultrasound sufficient for pre-implant verification?

These are reasonable questions. The kind someone asks when they're trying to be thorough without blowing the budget. I get it. I've asked similar ones myself. But the real answer is more complicated than a yes or no.

The Deeper Issue: What Ultrasound Actually Resolves

Here's where things get interesting—and where most people miss the point entirely.

Ultrasound operates at frequencies typically between 2-18 MHz for medical imaging. Higher frequencies give better resolution but less penetration. Lower frequencies penetrate deeper but resolve less detail. For ICD devices, the relevant structures are often small—lead tips, connector pins, internal circuitry—and they're encased in materials designed to be acoustically challenging (titanium, silicone, epoxy).

Most buyers focus on whether ultrasound can see the device and completely miss whether it can measure it to spec.

The question everyone asks is: 'Does ultrasound show the device?' The question they should ask is: 'Does ultrasound show the device within our required tolerances?'

Industry standard for dimensional verification in medical device manufacturing often requires measurement accuracy within ±0.1 mm or better for critical features. Ultrasound, in practice, gives you spatial resolution in the range of 0.5-1.0 mm at typical imaging depths for ICDs. That gap matters.

The Uncomfortable Truth

People think higher frequency ultrasound gives better resolution, which is true—up to a point. But higher frequencies also attenuate faster in tissue and device materials. By the time you're imaging through the chest wall and the device housing, the effective resolution for internal components drops significantly. (Honestly, I'm not sure why this isn't more widely taught in procurement training. My best guess is that most people learn ultrasound basics from clinical applications, not manufacturing QA.)

The assumption is that ultrasound is 'good enough' for verification. The reality is that it depends entirely on what you're trying to verify—and whether you can accept the uncertainty.

The Cost of Not Knowing

This isn't an academic point. I've seen the consequences play out.

In Q1 2024, we received a batch of 500 ICD devices where the lead connector dimensions were visibly off—0.4 mm deviation from our specified tolerance. Normal tolerance is ±0.2 mm. The vendor claimed it was 'within industry standard.' We rejected the batch, and they redid it at their cost. Now every contract includes specific dimensional verification requirements with minimum resolution standards.

That quality issue cost us a $22,000 redo and delayed our launch by six weeks. (I still kick myself for not specifying the verification method upfront. If I'd included resolution requirements in the original contract, we'd have caught it earlier.)

The hidden cost of relying on ultrasound alone isn't the equipment—it's the false confidence. When you think you can verify something and actually can't, you don't look for other methods. You accept risk you shouldn't.

For our 50,000-unit annual order, even a 0.5% defect rate that goes undetected means 250 devices with potential issues reaching inventory. At an average device cost of $18,000, that's $4.5 million in at-risk product.

The Hard Truth: Verification Requires Specificity

So what actually works? The answer isn't sexy, and it isn't a single technology.

For dimensional verification of ICD devices, the appropriate method depends on what you're measuring:

• External dimensions and connector geometry: Vision systems with telecentric lenses can achieve ±0.01 mm accuracy. Expensive? Yes. Necessary for critical features? Also yes.
• Internal component placement: X-ray or CT remains the standard for verifying internal structures. Ultrasound has a role for gross positioning, but not for detailed verification of solder joints or component alignment.
• Seal integrity and housing: Pressure decay testing and helium leak detection are the standards. Ultrasound can detect gross leaks in some cases, but not the micro-leaks that matter for long-term implant reliability.

The question isn't whether ultrasound is useful—it is, for specific applications. The question is whether it's sufficient for your verification requirements. In my experience, it rarely is for critical specifications.

I learned this in 2020. Things may have evolved since then—newer ultrasound technologies like high-frequency micro-ultrasound (40-70 MHz) used in small animal imaging can achieve much higher resolution, but they're not practical for through-tissue imaging of implanted devices in humans. (Ugh. The physics is the physics.)

The Bottom Line

This was accurate as of early 2025. Medical device verification standards change, so verify current requirements before finalizing your QA protocol.

Ultrasound shows you something. It doesn't show you everything. And in medical device quality, 'something' isn't always enough. The fundamentals haven't changed: specify your tolerances, then verify with methods that can actually measure to those tolerances. Everything else is just hoping.