Surgical Robots Advance Closer to Complete Autonomy in Operations

Surgical Robot Operating on Deceased Pigs

Juo-Tung Chen/Johns Hopkins University

The AI-driven robot successfully extracted the gallbladder from a deceased pig, marking a pivotal achievement in machine-assisted surgery with minimal human involvement.

This sophisticated robot features a dual-layer AI system trained using 17 hours of surgical video, which encompasses 16,000 movements performed by human surgeons. During operation, the first layer of the AI observes the endoscopic video and generates clear verbal instructions like “clip the second duct,” while the second layer translates these directives into precise three-dimensional tool movements.

In total, the gallbladder procedure involved 17 distinct tasks, of which the robotic system executed 8 with a flawless success rate.

“While current surgical robotics technology has indeed made certain procedures less invasive, the complication rate hasn’t actually decreased compared to traditional laparoscopic surgery performed by humans,” states Axel Krieger from Johns Hopkins University in Maryland. “This research paves the way for the next generation of robotic systems beneficial to both patients and surgeons.”

“This investigation shines a light on the vast potential of AI and surgical robotics,” adds Danail Stoyanov from University College London. “Remarkable strides in computer vision for surgical footage, alongside accessible robotic platforms for research, will empower us to advance surgical automation.”

Nonetheless, Stoyanov points out that significant challenges remain before the system can be applied in clinical settings.

For instance, although the robot achieved a 100% success rate in completing its tasks, it needed to self-correct six times per procedure. This could involve a gripper that initially missed the artery during its attempt.

“There were numerous instances where self-corrections were necessary, all autonomously executed,” remarks Krieger. “It effectively identifies initial errors and rectifies them.” The robot also requested a human operator to swap one of its surgical instruments for another, indicating that some human intervention was still required.

Ferdinand Rodriguez Y. Baena from Imperial College London emphasizes the promising future of robotic surgery. “The horizon looks bright—and tantalizingly close,” he asserts. “To ensure the safety of human applications, regulatory measures must also evolve.”

The next phase involves enabling the robot to operate autonomously on living animals, where factors like respiration and bleeding could introduce complexities.

Topic:

Source: www.newscientist.com

Neuralink Question: Has Elon Musk Achieved a Revolutionary Advance in Brain Implant Technology?

Noland Arbor can play chess using Neuralink implant

Neuralink

Neuralink, the brain-computer interface company founded by Elon Musk, has revealed the identity of its first patient who says its implant “changed his life.” But experts say it’s not yet clear whether Neuralink has done more than replicate existing research efforts.

Who was Neuralink’s first patient?

Musk announced in January that the first human patient had received a Neuralink implant, but few details were released at the time. We now know from something. Live stream video by company – Who is that person and how will the test be done?

Noland Arbaugh explains in the video that an accident eight years ago dislocated his fourth and fifth vertebrae, leaving him a quadriplegic. He previously controlled the computer with a mouth interface, and is shown moving the cursor with just his thoughts, apparently using a Neuralink implant.

“Once I started imagining the cursor moving, it became intuitive,” Arbaugh says in the video. “Basically, it was like using ‘force’ on the cursor, and I was able to move the cursor anywhere I wanted. I could just look anywhere on the screen and the cursor would move where I wanted it. It was a very wild experience.”

He uses the device for reading, language learning, and computer games such as chess, and claims he uses it for up to eight hours at a time, at which point he needs to charge the device. “It’s not perfect, I’ve run into some problems. But it’s already changed my life,” he says.

What does the implant contain?

Neuralink did not respond to requests for an interview, but its website says the current generation coin-sized implant, called N1, generates neural activity through 1,024 electrodes distributed across 64 threads that extend into the user’s brain. It is said that it records. These are so fine that they must be placed by a surgical robot.

In a livestream video, Arbaugh said he was discharged from the hospital the day after his implant surgery, and that from his perspective the surgery was a relatively simple process.

The implant uses a small battery that is charged through the skin by an inductance charger and communicates wirelessly with an app on your smartphone.

Does this mean the first human trials were successful?

Reinhold Scherrer Researchers at the University of Essex in the UK will decide whether Neuralink’s first human trial was a success because the company “has not released enough information to form an informed opinion” He said it was too early.

“While the video is impressive and there is no doubt that it took a lot of research and development work to get to this stage, it is unclear whether what is being shown is new or groundbreaking,” he said. Masu. “Although control appears to be stable, most of the studies and experiments presented so far are primarily replications of past studies. Replication is good, but major challenges still remain. ”

Who else is working on brain implants?

Neuralink isn’t the only group exploring this idea. A number of academic organizations and commercial startups have already conducted human experiments that have successfully interpreted brain signals and produced some sort of output.

A team at Stanford University in California placed two small sensors just below the surface of the brain of a man who was paralyzed from the neck down. Researchers may be able to interpret the brain signals when a man decides to put pen to paper and translate them into text that can be read on a computer.

When will Neuralink be available and how much will it cost?

It’s too early to tell, as this has a long way to go before it becomes a commercial product, with much testing and certification to come. But Musk has made it clear that he intends to commercialize the technology.of The first product planned was named Telepathy.allows users to take control of their mobile phones and computers.

topic:

Source: www.newscientist.com

Shield AI secures $200 million funding with $2.7 billion valuation to advance military autonomous flight technology expansion

Shield AI The company has secured $200 million in fresh funding to expand its autonomous flight systems for the U.S. military and its allies.

Established in 2015, the startup currently holds a valuation of $2.7 billion. The latest funding round was led by US Innovation Technology Fund (USIT) with significant participation from Riot Ventures, a previous investor in Shield AI. Other contributors include existing investors Disruptive and Snowpoint, as well as new investor ARK Invest, an investment management company founded by Cathie Wood.

USIT, guided by billionaire Thomas Tull, served as the sole investor in Shield AI’s initial $60 million Series E funding. This substantial Series F round is a testament to the company’s successful track record of fundraising – the Series E raised a total of $225 million, and its Series D ranged between $210 million and $300 million.

It also highlights the capital-intensive nature of defense-focused startups, even for companies like Shield AI that offer more affordable systems compared to their traditional counterparts.

The startup specializes in developing hardware and software to transform drones and aircraft into autonomous systems capable of carrying out missions in conflict zones. The company’s flagship product is Hivemind, an AI pilot software that enables drones and aircraft to operate autonomously without relying on GPS assistance. Shield AI has also introduced a drone swarm feature called V-Bat Teams, which allows a single human operator to command at least four V-Bat drones (developed by Martin UAV, acquired by Shield AI in 2021).

“Our nation faces the challenging reality of having insufficient pilots, and rule-based autonomous solutions are insufficient for the existence of such swarms,” said Ryan Tseng, CEO and co-founder, when announcing V-Bat Teams earlier this month. “Shield AI changes this. For nearly nine years, Shield AI has been building the world’s most advanced AI pilots using a unified AI foundation that is applicable and deployable across all aircraft types, from quadcopters to F-16s. We’ve been steadfastly focused on that.”

The San Diego-based company is also working on integrating Hivemind into unmanned fighter jets and other aircraft. The Shield AI technology stack has garnered significant interest from the Department of Defense due to increasingly sophisticated counter-drone technologies that focus on disrupting drone communications and navigation.

“The battlefield is increasingly dominated by drone warfare, and adversaries are turning the battlefield into a hostile environment by disrupting communications and GPS,” stated Stephen Marcus, co-founder and general partner at Riot Ventures. “We are doing everything we can to address this. Modern Air Forces are operating blindly. Shield’s AI pilots are intelligent and adaptable to their environment, requiring no GPS or communications. Their AI is trainable and adaptable for diverse missions, and they have successfully flown teams of copters, V-BATs, and modern fighter jets. The most comparable technology we have seen thus far is what Tesla is doing with their self-driving stack.”

The new funding arrives amidst a surge in investor support for defense technology startups, driven in part by escalating geopolitical tensions and the U.S. lagging behind its adversaries. Engineers and the Pentagon are keenly aware of the risks at hand. In fact, some Shield AI executives have made noteworthy comparisons: Back in 2021, co-founder Ryan Tseng drew parallels between the Chinese military and Netflix, and the U.S. military and Blockbuster.

Source: techcrunch.com