We’re gobsmacked by the sheer number of special-purpose acquisition companies (SPACs) launching, given that most of these SPACs offer little value to the retail investor. One of the exceptions (if not for the premium stock price) is the recent IPO by Butterfly Network (BFLY), which is disrupting the ultrasound industry with a portable medical device powered by artificial intelligence. We recently came across a life sciences startup called NovaSignal with a different angle on the technology that uses robotics and AI for ultrasound brain imaging to detect disease.
An AI Healthcare Startup
Founded in 2013, Los Angeles-based NovaSignal changed its name from Neural Analytics just last year to reflect the company’s intentions to expand the use of its autonomous tech for measuring blood flow in the brain to applications outside of stroke detection. NovaSignal has raised about $111 million in disclosed funding from a dozen investors, including a couple of local LA area VC firms. The company has also enjoyed some government love, with about $26 million of that total coming in research grants from an alphabet soup of agencies, including DoD, NIH, and NSF.
Using Robotics and AI for Ultrasound Brain Imaging
At its most basic level, ultrasound technology uses acoustic energy to create an image. A key component of an ultrasound machine is the transducer, which both emits these high-frequency sound waves and detects the echoes reflected back. The handheld ultrasound from Butterfly Network, for instance, only needs a transducer and smartphone to work. In the case of NovaSignal, the company has developed a tricked-out version of a transcranial doppler (TCD) ultrasound, a noninvasive medical device designed to evaluate blood flow in and around the major arteries of the brain.
Credit: NovaSignal
The company’s Lucid Robotic System consists of a portable ultrasound machine (Lucid TCD 2.0, above) and a robotic headset (NovaGuide) that looks like it could have been used to listen to your 8-track tape of Iron Butterfly’s “In-A-Gadda-Da-Vida” back in the 1960s:
Credit: NovaSignal
After the robotic headset is placed on the stoner’s patient’s head, the NovaGuide uses artificial intelligence to autonomously search for the main cerebral arteries, automatically adjusting the depth of the ultrasound beam to monitor changes in blood flow velocity in real time along the length of the blood vessel. It can even count the emboli, the free-floating plaque that blocks blood vessels and reduces blood flow to the brain. If you don’t believe us, you can read some of the 100 peer-reviewed research papers for yourself. The system doesn’t require a technician with specialized skills and has shown to be 99.7% accurate to an expert sonologist, according to NovaSignal. The tech is reliable enough that it has been cleared for use in both the United States and Europe, and is covered by at least 15 patents.
More recently, the company launched an app that directly alerts physicians when patients are experiencing blood flow changes that could indicate a stroke or other emergency, according to a report in BioWorld. The medical news site also reported that NovaSignal has tinkered with the AI component, which now collects data from every scan to provide clinicians with predictive insights about how a patient is likely to progress over time.
Use Cases for a Smart TCD Ultrasound
The most obvious use case is to predict or diagnose stroke for better treatment outcomes, but that’s really only the beginning. The autonomous system can be used for long periods of time, such as during cardiac surgery, which may require patient monitoring through the entire procedure. Measuring blood flow in the brain can also detect a hole in the heart, a condition that you might be surprised to learn affects about 25% of us (and 100% of Christmas grinches). The condition is benign – until it’s not. Troublesome blood flow may also indicate poor gas exchange in the lungs, which looks like tiny microbubbles in the blood.
Credit: NovaSignal
That last insight is particularly relevant in the Age of Rona. In fact, NovaSignal published research last year that showed that the Lucid Robotic System can provide additional insights about the pathology of COVID-19 and may be another remote patient monitoring tool in the fight against the disease. Specifically, the researchers found that 83% of 18 mechanically ventilated patients with severe COVID-19 pneumonia had detectable microbubbles passing through their cerebral arteries. That’s a sign of pulmonary shunting, where deoxygenated blood bypasses the normal circulatory system, resulting in lower levels of oxygen in the blood and contributing to respiratory distress.
Other potential future applications include Alzheimer’s disease and traumatic brain injury, as NovaGuide continues to build algorithms for new use cases.
Conclusion
AI is helping transform ultrasound technology into a multifaceted diagnostic imaging tool. Companies like Caption Health, for example, are just churning out algorithms to turn any Joe MBA into an expert sonographer. Beyond diagnostics, ultrasound tech can also be leveraged as a therapeutic tool, such as breaking up kidney stones or even burning off cancer cells. NovaSignal is certainly hoping physicians can leverage its AI-powered TCD technology for monitoring cerebral blood flow as an everyday diagnostic tool in clinics and hospitals.
That kind of brings us back to Butterfly Network and its founder, serial entrepreneur Dr. Jonathan Rothberg, who is apparently going back on the SPAC attack by taking another one of his medical device startups public through a reverse merger with a blank-check company. In this case, Quantum-Si is developing protein sequencing technology using a unique semiconductor chip – yet another riff on the lab-on-a-chip paradigm. Basically, as the company is selling the concept, you can buy a desktop machine that can “decode the molecules of life.”
This trend – the shrinking of sophisticated medical equipment into handheld devices or desktop machines – is an important one to watch for retail investors. This is where real disruption in the medical field is taking place, assuming these machines can deliver results equal to or better than current legacy technology. It’s not just about better patient outcomes, but driving down medical costs and putting cutting-edge tools into more hands without the need to employ a bunch of specialized experts in different disciplines.
There are many AI healthcare stocks out there, but we're holding just one. Their software is used by all of the top 20 pharmaceutical companies, and they're also developing their own drugs. Become a Nanalyze Premium annual subscriber today and find out all five AI stocks we're holding.