Visions of the future of health care technology, both dark and optimistic, were put on full display this week at the University of Minnesota as more than 1,000 students, entrepreneurs and industry veterans gathered for the 18th annual Design of Medical Devices Conference.

The event kicked off Tuesday with a dive into the unsettling world of medical device cybersecurity, previewing hacks of the future like electromagnetic inference to confuse life-preserving sensors.

It ended two days later with a long look at the promise of “mixed realities,” which is the seamless blending of physical and computer-generated realities to create new perspectives, like augmented-reality goggles that can show a surgeon what’s right under the skin she’s about to cut into.

U engineering Prof. William Durfee, who is the technical program chairman for the conference, said there’s a need to balance tough topics like regulation, privacy and security — critical to a highly regulated field like medical technology — with the creative imperatives of “forward-thinkers and visionaries” who inspire others with their views of what’s possible.

“That’s kind of the struggle that you have in innovating medical technology,” Durfee said. “It is such a highly regulated space, and a high-risk space ... but you can’t let it be driven entirely by the cautionary folks.”

The conference is put on partly by the University of Minnesota’s Earl E. Bakken Medical Devices Center, a multidisciplinary research group created to catalyze collaborations between up-and-coming clinicians and engineers, whose colleges are located across Washington Avenue from one other.

Medtronic senior scientist Megan M. Schmidt, who worked the company’s booth at the conference, is working on what she called “the next next-generation” products in Medtronic’s line of tissue-ablation devices, including sensors that can work whether the ablation is accomplished with cryogenics or RF energy.

Asked what she thought about the next generation of cybersecurity risks that were described a few minutes earlier in a talk by Michigan cyber-med expert Kevin Fu, Schmidt said, “It’s one of those things where I’m like, that is incredible. That is horrifying. And I think that’s kind of the world we live in today — people can do these things, and you have to plan for them.”

Fu made his name in health care cybersecurity by leading the first team to remotely hack into a bacon-wrapped Medtronic defibrillator and then publishing the findings in 2008. (The paper withheld key details of the attack, and suggested possible fixes.) On Tuesday, Fu discussed sophisticated hacking methods that appear mainly hypothetical today, but could become serious threats down the road.

That includes figuring out the frequencies at which wires or sensor components begin to vibrate, and then remotely targeting them with interference to produce aberrant signals. That could be a serious risk if the target was, say, a pacemaker lead sensing a beating heart or the temperature sensor on a cryogenic tank for embryo preservation.

“Just remember,” Fu told the audience, “you are going to be designing for the risks of 10 years from now.”

The future looked a lot rosier on Thursday, when experts from industry and academia discussed the promise of “mixed realities,” an umbrella term that includes virtual-reality headsets, augmented-reality displays combining virtual and real images, and 3-D printing of virtual objects, among other sensory-enhancing technologies.

Health care is not just ripe for mixed-reality technology, it’s already being used to give doctors, patients and customers better views of what’s happening inside a body, a device or an assembly plant. But imagine a “flight simulator” for medicine, or a 3-D printer that can make FDA-approved titanium implants at the point of care in a hospital. Imagine virtually “shrinking” a patient and having them enter their own body to see their disease up close. All of those things are coming, panelists said.

Since such technologies will add costs, panelists emphasized that mixed-reality systems should only be used in situations where they will have a meaningful impact on outcomes.

“One thing that really hit home for me as an engineer working in a clinical space was that the ideas really don’t necessarily need to come from the engineers,” said Amy Alexander, a senior biomedical engineer at the Mayo Clinic. “Of course we have ideas, and we can build the tech and get really good at it. But we need to be sourcing information from the people who are in the OR. And that’s when the magic really happens.”