ARTICLE SUMMARY:
Phantom Neuro’s human computer interface technology, in contrast to bulky, high-cost, invasive options, potentially enables people with limb loss to move their muscles seamlessly and intuitively in real-world settings.
Several years ago, Connor Glass, MD, a young reconstructive plastic surgeon in training at the Johns Hopkins School of Medicine became engrossed in the emerging field of human-computer interfaces (HCI).
Glass was looking for opportunities to make a significant impact on healthcare. This new field was capturing scientific and investor interest, particularly for applications related to restoring functionality to people with cognitive and mobility impairment. (See “Assessing Medtech in 2025: Financings Steady in a Weak LifeSci Market,” MedTech Strategist, August 8, 2025.) The approaches being studied seemed superior to conventional treatments, but were still invasive, costly and non-intuitive.
In 2020, he identified a minimally invasive implantable device that used surface electromyography (sEMG) electrodes and machine learning to approximate implantable system behavior and started Phantom Neuro. The technology had enormous potential to compensate for damaged residual nerves and restore mobility to amputees.
Conventional sEMG systems have been used to restore some functions to impaired limbs, but are still limited by poor limb control, high cost, invasiveness, and discomfort. They do not fit seamlessly into users’ everyday lives.
The device Glass identified was developed at Johns Hopkins Medicine’s Applied Physics Lab. It consisted of a sweat-band-like strip with multiple embedded EMG sensors that could be implanted subcutaneously and wrapped around muscles. It would identify electric signals and send them wirelessly to a miniature computer module, which Phantom Neuro has branded as the “Fusion Port”. The Fusion Port is embedded into the patient’s artificial limb socket, from which it sends signals to a robotic prosthesis, initially a hand with flexible fingers. Using machine learning, the system can detect in unprecedented detail complex electrical signals from nerves remaining near the amputation and interpret them.
The entire system, which Phantom Neuro has branded The Phantom X, recently received FDA breakthrough status and has been accepted into CMS’ New Technology Add-On Program (N-TAP), opening communications doors with regulators and payors. The first in-human feasibility study is scheduled to begin in early 2026 in Australia, with a US pivotal trial aimed at FDA regulatory approval of the device for use in robotic upper limb control beginning late next year or early 2027.
The launch initially will target orthopedic, plastic, and vascular surgeons in the outpatient setting because these specialties routinely work on limbs and soft tissue procedures. The implantation is safe for a wide variety of healthcare providers, including primary care physicians, to perform, given that it will take place just under the skin, far from vital organs, nerves or blood vessels.
Glass sees a day when HCI devices like Phantom X will complement much more invasive brain-computer interface systems, which are the recipient of so much attention these days. He notes that some of the US government’s long-running DARPA programs funding advanced prosthetics research are recently completed and have delivered amazing technological advances that are laying the groundwork for industry’s work. That leaves room for Phantom Neuro and other startups to seize the moment. (See “Brain-Computer Interfaces: The Internet of Us,” MedTech Strategist, October 22, 2021.)
Subscribers, click the link below to watch this video.
