AI is being used to solve a variety of healthcare challenges, but its greatest hope remains its potential to detect and diagnose disease better than humans.
the latter is trying neural light, an Israeli startup that aims to use computer vision and deep learning to analyze videos of a person’s face and eyes to determine if the person has a neurological disorder. Just six months after its launch, the company announced it had raised $25 million last week. Koch Disruptive Technology Led the Series A financing, with Breyer Capital, Samsung Next, VSC Ventures, and individual investors. Including the funding raised last week, the company has raised $30.5 million since it launched in late 2021 six months ago.
So how does this technology work. The company’s CEO, Micha Breakstone, explained in an email response to questions that proprietary algorithms are used to extract information from video captured by standard smartphones and webcams. Specifically, the AI focused on ophthalmometric data — data about tiny eye movements — “which have been shown in more than 750 published scientific papers to correlate strongly with the progression of neurodegenerative diseases,” He says.
NeuraLight’s deep learning technology “then” extracts more than 100 of these eye parameters in one fell swoop (such as saccade and anti-saccade latency, pupil dilation speed, and eye blink rate), building a powerful proxy for a person’s neural health. Be patient,” Breakstone said.
The company’s short-term goal is to improve the design of neuroclinical trials and help develop complex nervous systems such as Parkinson’s, Alzheimer’s, ALS (Amyotrophic Lateral Sclerosis) and MS (Multiple Sclerosis). Diseases lead to better treatments. But Breakstone has a more ambitious goal, though he admits it’s still years away — NeuraLight’s technology has become the gold standard for neurological assessment.
Thus, this automatically implies a need to revisit current neural assessment modalities. According to Breakstone, the main problem appears to be that disease detection and diagnosis are very subjective and have 25%-30% inter-rater variability.
“The 25%-30% inter-rater variability in Parkinson’s assessment basically means that if two physicians assess the same patient for Parkinson’s on the same day, the physician will Significantly different assessments are given over time,” Breakstone “that affects treatment, and provides an inability to objectively measure whether and to what extent potential treatments for diseases such as Parkinson’s disease (or MS or Alzheimer’s disease) affect disease progression. drug developers present a huge hurdle.” The patient’s disease. “
According to the company’s press release, this variability and lack of objective and sensitive endpoints has led to one thing: “The approval rate for neurotherapeutics is only 6%, and the approval rate for non-neurotherapeutics is less than 50%.”
This variability and resulting scarcity of approved neuropharmaceuticals is a problem that NeuraLight’s powerful deep learning algorithms could potentially address. This is also a method that pharmaceutical companies can use to improve the design of neuroclinical trials in the future.
“The goal of pharmaceutical companies is a key stepping stone toward our ultimate goal. That is, making our objective and sensitive ocular biomarkers part of our decision-making platform to improve the accuracy of clinical trials and remove the subjectivity of neurological diagnosis sex,” Breakstone said.
Given that NeuraLight’s technology is device-agnostic, researchers can combine the company’s technology to conduct decentralized clinical trials, a move that frees clinical research from a central location, opening up the possibility of having a diverse trial population, he added.
“[Being device agnostic] It is especially valuable for clinical trials focused on neurodegenerative diseases, which make it difficult for patients to travel to central research facilities,” he said.
While it will be years away from establishing the digital biomarkers developed by NeuraLight as the gold standard for neurological assessment, it is now possible.
“In the near future, we will be contracting primarily with pharmaceutical companies and will be billed for each item based on cohort size and labeling cadence,” he said. “As our biomarkers gain further clinical validation, this model may evolve and we become part of the indication itself and/or accepted outcome measures.
Pharmaceutical companies can use the company’s platform to design clinical trials, for which NeuraLight does not require FDA approval. However, to help pharmaceutical companies develop new drugs with its technology, the company will seek FDA approval for surrogate endpoints in the future.
Now, the company will use some of the funds raised to hire more people in addition to the existing 25. The remaining funds will be used to build NeuraLight’s own infrastructure and expand trials with research institutes and pharmaceutical partners. The company hopes to get paid for its platform by partnering with drug companies based on trial size and number of cohorts.
While NeuraLight made a splash in its recent fundraising event, promising to fundamentally change how neurological disorders are diagnosed and drugs are developed, it’s not the only company working to develop alternative methods for assessing neurological disorders.
they include altoida, which leverages artificial intelligence and augmented reality to perform a 10-minute cognitive assessment using a smartphone anywhere. at the same time, beacon biosignal Neural biomarkers are being developed and validated using its EEG analysis platform to improve cohort selection during clinical trials, improve clinical assessment, and identify novel endpoints.
In the case of NeuraLight, it may hope that both its platform and its team of scientific advisors will be differentiators.In a press release announcing the fundraising last week, the company touted the fact that two Nobel laureates, neuroscientist Professor Thomas Südhof Nobel Prize 2013 work on synaptic transmission, and professor at Stanford University Alvin Ross He was awarded the 2012 Nobel Prize for his work on the kidney exchange program – a member of its Scientific Advisory Board.
“The need in neurology to establish objective and sensitive biomarkers is enormous,” Südhof said in a statement. “NeuraLight’s approach really promises to fill that need.”
