AI-Powered Teleradiology FDA Cleared for Triage in Departments Swamped by COVID-19

By Conn Hastings at Medgadget

Nines, a teleradiology company based in Palo Alto, CA, recently received FDA clearance for their NinesAI medical device, which supports the automated radiological review of CT Head images for the possible presence of two time-critical, life-threatening indications: intracranial hemorrhage and mass effect. The technology can help radiologists in triaging cases. Nines is the first company to receive FDA clearance for AI technology that triages mass effect conditions.

Teleradiology is an increasingly vital service for healthcare providers, whereby radiological images are sent to a radiologist in a remote location for analysis. Moreover, AI has an emerging role in radiological diagnostics, and can provide a supportive role for radiologists in making their assessments.

All current customers and new customers who sign up with Nines before June 30, 2020 will get NinesAI for free. This access will allow Nines’ customers to assess, identify, and triage emergent conditions of intracranial hemorrhage and mass effect while in-house radiology departments are inundated, especially as some of these departments are overwhelmed by COVID-19 cases.

Medgadget had the opportunity to speak with David Stavens, CEO and co-founder of Nines, about the technology, and how the company is supporting quality care and radiologists in a time where hospitals and healthcare providers are resource-constrained.

Stavens has an interesting background in AI, having been involved in the world’s first robotic car “Stanley” as a co-founder of Stanford’s self-driving team, which became the foundation for Waymo, which was acquired by Google, itself part of Alphabet. He later co-founded and was CEO of online learning platform Udacity.

Conn Hastings, Medgadget: Please give us an overview of your background in AI.

David Stavens, Nines: I’ve loved computers since I was very young. I remember receiving my first Apple IIGS for Christmas in the late 1980s and learning to write computer software soon thereafter. It was unusual in those days for kids to be into computers and coding. As I got older, I recognized that computers and software were not only fun hobbies, but an exciting way to make a broader impact in the world.

I am drawn to important humanitarian problems and how they could be solved using computer technology. As co-founder and CEO of Udacity, Sebastian Thrun and I worked to use the web to make world class education and great jobs available to everyone in the world. Prior to that, we worked on building one of the first self-driving cars while at Stanford, which eventually became Waymo.

Self-driving technology has the potential to save many of the 40,000 lives lost every year in the U.S. due to traffic accidents, improving mobility for those with visual or neuromotor impairment, and making longer drives more pleasant for everyone. At Nines, we believe that radiology can be made even better for radiologists and patients using technology and we are excited to help bring that change about.

Medgadget: What inspired you to start Nines?

David Stavens: Our health is one of our most precious assets — a point made even more pronounced with the onset of COVID-19. But navigating the healthcare system is very complex, especially for those without insurance or those who do not live in an industrialized country. Access to timely, affordable,high quality care should be universal. At Nines, we believe that radiology can be made even better for radiologists and patients using technology and telehealth and we are excited to help bring that change about. The adoption of telehealth has only become more essential and prevalent in the time of COVID-19.

The use of medical imaging, such as MRI, CT and ultrasounds performed in healthcare facilities, has continued to rise, according to the Journal of American Medical Association in analyzing trends from seven integrated U.S. healthcare systems. These scans are not only increasing in number, but also in complexity — scans include more images and the images are of higher resolution.

According to the American College of Radiology, there is only one radiologist for every 10,000 Americans. To keep up, a typical radiologist must review one image every three to four seconds to meet workload demands in an 8 hour workday, according to a study from the Mayo Clinic. Each one of those images has the potential to show information that is critical in making a radiologic diagnosis that is used to direct patient care.

Nines is building world class technology to help. NinesAI is FDA-cleared and supports the automated radiological review of CT Head images for the possible presence of two time-critical, life-threatening indications – intracranial hemorrhage and mass effect – to aid radiologists in triaging cases and identify them more rapidly. In an emergency room late at night, an actionable report turned around quickly could give patients a reliable and speedy diagnosis and could calm worries about what is wrong.

Medgadget: Please give us an overview of the teleradiology services provided by Nines.

David Stavens: Nines is a teleradiology practice that pairs world-class radiologists with fantastic engineers and product managers to deliver high quality medical care accelerated by advanced technology.Hospitals and medical practices can choose Nines to read some volume of their radiology studies, for example on nights or weekends when they have fewer radiologists on call. With engineers and physicians side-by-side, technology and process improvements occur rapidly for the benefit of physicians, hospitals, medical practices, and patients.

For example, on average, a radiologist is interrupted about five times an hour, whether answering phone calls or coordinating with colleagues, which increases exam interpretation times, according to the Journal of Academic Radiology. At Nines, radiologists and engineers have worked together to develop the Nines Navigator™ worklist and the Nines Reading Assistant. These are administrative, non-medical device programs that aggregate clinical information and surface custom-built tools to improve radiologist focus. This includes listing relevant patient info and easier communication with hospital physicians interacting with the patients. Together with NinesAI, these tools assist Nines radiologists with providing timely, quality care.

Medgadget: How has the COVID-19 pandemic affected the services you offer and the clients you work with?

David Stavens: COVID-19 has placed unprecedented strain on all aspects of the U.S. healthcare system. We are offering in-house access to the NinesAI technology free of charge for new and existing Nines Radiology customers who sign up by June 30, 2020. Some healthcare providers are seeing early indications of a potential relationship between COVID-19 and an increase in the number of strokes in some younger patients. Stroke is one cause of intracranial hemorrhage and mass effect, the two conditions NinesAI detects. In addition, NinesAI improves radiology efficiency through triaging emerging studies, which we believe will be particularly helpful as radiologists are under strain at this time.

Medgadget: What are the strengths of AI in radiological analysis?

David Stavens: For emergent conditions of ICH and mass effect, time to intervention is critical. With intracranial hemorrhage for example, the 30-day mortality rate ranges from 35% to 52% with only 20% of survivors expected to have full functional recovery at 6 months, and approximately half of this mortality occurs within the first 24 hours.

Radiologists using NinesAI can be notified of a potential life-threatening finding in approximately 15 seconds after image acquisition is complete, meaning that potentially life-saving care can begin very quickly. The standard-of-care without AI is to read studies in the order in which they were received. We believe the role of artificial intelligence for radiology is to make radiologists’ lives better as an assistive technology that supports their service and prioritization of the patients who need care most urgently.

Medgadget: Please tell us about these most recently approved AI-powered teleradiology services.

David Stavens: At the end of April, we received unprecedented U.S. Food and Drug Administration clearance for artificial intelligence technology that triages intracranial hemorrhage and mass effect conditions on non-contrast CT scans of the head. Our NinesAI medical device supports the automated radiological review of images from these scans for the possible presence of these two time-critical, life-threatening indications to aid radiologists in triaging cases. We are proud to be the first company to receive FDA clearance for artificial intelligence technology that triages mass effect conditions and to our knowledge, the first company to receive simultaneous FDA clearance on multiple indications.

Original news can be found here.

Microfluidic Test for Viral Antibodies Takes Just 20 Minutes

By Conn Hastings
Researchers at Hokkaido University in Japan have developed a microfluidic test that can detect antibodies against a viral infection. So far, the test has been optimized to detect avian flu, but could be adapted to detect antibodies against the virus causing COVID-19. The device can provide a result in as little as 20 minutes and requires only 2 microliters of serum to run.

Antibody tests have been proposed as a way to determine how many people have been exposed to COVID-19, and may help us to understand how far the virus has spread. Given that many infected people are asymptomatic, they may never be recorded as infected, limiting our knowledge of infection levels and hampering the response to the pandemic.

However, testing random samples of people using antibody tests could help researchers to gain more knowledge in the fight against COVID-19. Such tests may also potentially be useful in determining who may have immunity against the virus, although it is not yet clear if the presence of antibodies against the virus causing COVID-19 indicates effective immunity.
Existing techniques to assess if an antibody is in the blood often rely on a visual assessment to determine whether a result is positive. This means that the accuracy is limited, as interpreting results is subject to human error. To address this, these researchers have developed a new type of microfluidic test, which is based on binding a fluorescently labelled protein to the antibody of interest in a serum sample.

The technique involves detecting the fluorescence polarization of the bound fluorescent molecules, and liquid crystal molecules are used to control the direction of this polarization. Once the sample is within the microfluidic device, it is then attached to a portable fluorescence polarization analyzer, weighing only 5.5 kg, which can measure the fluorescence signal and indicate if the antibody is present in the sample.

The researchers mix the serum sample with the fluorescent reagent and then allow it to rest for 15 minutes before loading it into the microfluidic device. They then attach the microfluidic device to the portable fluorescence polarization analyzer to obtain a reading. The entire process takes only 20 minutes, and such devices could help to speed up and streamline community testing for antibodies related to COVID-19, and allow for testing outside of medical labs.

“Our analyzer could be used to conduct other bio tests if suitable reagents are developed,” said Manabu Tokeshi, a researcher involved in the study. “By reproducing fragments of spike proteins expressed in the novel coronavirus, and using them as the reagent, the analyzer should be able to detect anti-coronavirus antibodies.”

Original news can be found at Medgadget

South Korea says recovered coronavirus patients who tested positive again did not relapse: Tests picked up ‘dead virus fragments’

  • Experts in South Korea said that recovered coronavirus patients who tested positive again were not reinfected and that their virus was not reactivated, as was previously feared.
  • More than 260 people who recovered and tested negative subsequently tested positive again. The Korea Centers for Disease Control and Prevention worried that the virus had reactivated after going dormant.
  • But the country’s infectious-disease experts said on Thursday that the tests were detecting dead fragments of the virus left in patients’ bodies.


Scientists said the wave of South Koreans who tested positive for COVID-19 even after they recovered did not have the virus reactivate after going dormant and that they were not reinfected. South Korea announced in early April that some patients who had recovered from and tested negative for the virus later tested positive, suggesting that the virus could reactivate or that patients could be reinfected. The country has recorded this happening in 263 patients, The Korea Herald reported.

But the country’s infectious-disease experts said on Thursday that the positive test results were likely caused by flaws in the testing process, where the tests picked up remnants of the virus without detecting whether the person was still infected, The Herald reported.

The Herald described the experts as saying that “dead virus fragments” were lingering in patients’ bodies after they recovered and that the virus did not appear to be active in the patients.

Dr. Oh Myoung-don, a professor of medicine and the head of Seoul National University Hospital’s division of infectious diseases, said that the committee studying the cases found little reason to believe that the patients had been reinfected or that the virus had reactivated, according to The Herald.

“The tests detected the ribonucleic acid of the dead virus,” said Oh, an adviser to the Korean government and the Korea Centers for Disease Control and Prevention.

Experts were already skeptical about the theory that the virus could reactivate

Jeong Eun-kyeong, the director-general of the KCDC, said earlier in April that the positive test results in these patients could mean that the virus “reactivated” after going dormant, Business Insider’s Holly Secon reported.

Jeong said the tests were conducted within a “relatively short time” after the patients were cleared, so it was unlikely that they were reinfected.

“While we are putting more weight on reactivation as the possible cause, we are conducting a comprehensive study on this,” Jeong said. “There have been many cases when a patient during treatment will test negative one day and positive another.”

But many experts said it was unlikely that the virus would go dormant and then reactivate in people. Instead, they said, it was more likely that patients’ bodies still had some fragments of the virus. This would mean a person could get a positive test but not be ill or able to infect others.

Dr. Keiji Fukuda, the director of Hong Kong University’s School of Public Health, told the Los Angeles Times in March that this was the most likely scenario.

“The test may be positive, but the infection is not there,” Fukuda said.

KHB SARS-CoV-2 Gold Standard Reagent Helps Italy to Fight Covid-19 Outbreak

According to a report by Rai Italia, a SARS-CoV-2 IgM/IgG Antibody Detection Kit (Colloidal Gold) supplied by TGS, a subsidiary company of Kehua Bio-engineering (KHB), is extensively used in test centers in Campania, Italy.

A report by Xinhua News Agency from Rome on April 26 said that Prof. Vittorio Sambri, an Italian medical expert at Università di Bologna, expressed his positive opinion on the medical products made in China, including diagnostic test products. Prof. Sambri said, Emilia-Romagna, an administrative region where he lives, purchased SARS-CoV-2 Detection Kits from Shanghai Kehua. When tested on patients who had been diagnosed with SARS-CoV-2 infection, the detection kits demonstrated a sensitivity of 90% in patients who had contracted the virus for at least 8 days. In patients who had been infected for more than 14 days, the kit’s sensitivity is 94%. Recently, this kit has been used in Emilia-Romagna Administrative Region on medical professionals.

What is KBH SARS-CoV-2 Detection Kit?

SARS-CoV-2 IgM/IgG Antibody Detection Kit (Colloidal Gold) is an easy to use rapid test targeting to detect Covid-19 for personal use, committing high specificity/accuracy and instantly resulting in 15-20 minutes.

Content Image
Content Image