The U.S. Census Bureau reports that with the launch of COVID-19 in 2020, 28 million Americans did not have health insurance at any time during the year. And while many Americans have health insurance, it often does not cover everything individuals need. For example, mental health services and follow-up breast cancer screening are not always covered.
This is where artificial intelligence (AI) can move forward to provide quality healthcare options at low cost. Companies like Vara and Paradromix are already working to increase accessibility, affordability and ultimately healthcare results – and investors are paying close attention.
Lu Zhang, founder of FusionFund, a venture capital firm focused on supporting early-stage startups such as Paradromics, said, “AI can really solve this accessibility problem, especially now that old age is a big trend in developing and developed countries.” ” “The key is to be able to better understand the origin of the disease and to achieve a highly personalized diagnosis and treatment plan.”
Those who do not have access or have minimal health insurance coverage are often Black, indigenous and colored people ,BIPOC) individuals and disproportionately poor. The Kaiser Family Foundation (KFF) found that “in 2019, non-aging AIAN [American Indian, Alaska Native]Hispanic, NHOPI [Native Hawaiian and other Pacific Islander] And black people are more likely to lack health insurance than their white counterparts. “And yet programs and services like Medicaid and the Children’s Health Insurance program help,” … they do not fully bridge the gap, making them less likely to be uninsured. There is more to come. ”
The 2020 epidemic worsened access to insurance, which disproportionately affected individuals in the community listed above with job losses and declining incomes, and therefore contributed to further disruptions in healthcare and medical coverage, according to the KFF.
AI-powered healthcare on the horizon
According to Harvard’s School of Public Health, “AI can improve health outcomes by up to 40% and reduce treatment costs by up to 50% by improving diagnosis, increasing access to care and enabling precision medicine.” The medical industry will cost 150 billion by 2025.
“I think we start with, for example, AI for medical imaging, AI for diagnostic or AI for medical sequencing. There is also much discussion about how we can improve workflow efficiency, ”Zhang said. “When we talk about AI, we think not only of AI algorithms, but also of other artificial intelligence products like AI robotics.”
Improve access and results in breast cancer screening
Each year in the U.S., an average of 255,000 breast cancers are diagnosed in women and 2,300 in men – according to a CDC report – and 42,000 women and 500 men die each year.
Individuals, especially women, are encouraged to have mammograms annually or every few years, depending on age, as part of active healthcare planning and treatment. However, there is an important difference, especially with regard to insurance coverage Type They should be screened.
According to multinational healthcare and insurance company United Healthcare, the annual mammogram is usually a screening covered by insurance plans because it is preventive care.
However, if a person goes for an annual mammogram, for example, and any abnormalities are found, then it is referred to a diagnostic mammogram, which is a screening that is less covered by insurance but uses breast. Used to diagnose cancer. . And since the latter is used to make a diagnosis, it usually involves higher costs, even if the insurance covers part of it, notes United Healthcare.
The high cost of diagnosis is one reason why Jonas Muff, founder and CEO of the AI-powered mammography screening platform Vara, started his company. The company offers a software screening service that can be installed on existing machines and does not require hospitals or healthcare companies to invest in significant new equipment. Once the center adopts turn-by-turn technology, the main change (except for improved functionality) is the branding partnership, which Muff noted as often simpler and with lines of “Clinic XY powered”.
Vara’s software platform operates in a radiologist’s workflow. Muff says Vara uses AI on multiple fronts. Software platforms work seamlessly to filter out common cancer-free mammograms, so radiologists can spend more time focusing on screening and analyzing suspicious aspects. In addition, Varni Technology also warns radiologists if they have missed a potential case of cancer that could otherwise be ignored. Muff said the team identified the feature as a “safety net” in turn, which could detect potential cancer more quickly through its AI and machine learning.
“The vision is really something that every woman can afford. The more clinics there are, the more women can afford these screenings, which is obviously very good for patients, but ultimately, it’s also great for everyone in the businesses and cancer treatment industry, ”Muff said.
In clinical trials in Germany, where the company was founded, Muff claims that the pig found about 40% of the cancer that was missed by radiologists. To get an idea of how much savings AI can offer in this way, turn screening services are offered in Mexico for about $ 15, which Muff noted is usually self-paying. He said women pay for the service with their credit cards, although they do not have insurance to get the screening. If they choose to be screened elsewhere in private clinics without taking turns, Muff claims they can expect to pay between $ 50 and $ 150 per screening in Mexico.
Personalizing diagnosis and treatment in mental health
Like breast cancer screening, mental health care and treatment are often excluded from insurance coverage in the US In fact, the National Institutes of Mental Health (NIMH) reports that one in five US adults lives with a mental illness. However, there are many barriers to insurance plans that can often delay treatment for these conditions, cause individuals to travel too far for network providers, or do not cover mental health care at all, making individuals pay more. Out of pocket expenses.
The National Alliance on Mental Illness (NAMI) quoted above in a 2020 blog post and said that although steps have been taken to make mental health care more accessible, it is not enough.
“The 2008 Mental Health Parity and Addiction Equity Act, the Affordable Care Act and the State Mental Health Parity Act require specific healthcare plans to provide mental and physical health benefits equally. And yet, insurance companies do not cover mental health care the way they want, “the post reads.
NAMI reports that, in addition, the organization has identified individuals in need of this type of treatment report. “Difficulty finding in-network providers and facilities for mental health care compared to general or specialized medical care. Often, going out of network was the only option for treatment. And individuals reported difficulty finding accurate information about in-network providers for their health plans. .
This can leave people who need treatment with few options or options that are too expensive. This is where the AI-powered company, Paradromix, hopes to fill this gap.
The purpose of paradromics is to develop a data interface that interacts directly with the brain’s neural signals using AI and machine learning. A technology the company is developing, called the “Connex Direct Data Interface”, collects huge amounts of individual neural signals with a fully implantable device designed for long-term daily service. Paradromics reports that its first clinical application is a support-communication device for patients who have lost the ability to speak or type, but the technology will expand into mental health diagnoses in the future.
“We can imagine a future where certain mental health diagnoses are better understood through neurological – rather than psychiatric – frameworks. This kind of understanding can contribute to the condemnation of these distortions, “said Matt Angle, CEO of Paradromics. “It is well known that pharmaceutical treatments, which are broad-spectrum and have non-specific action, are not universally effective and pose challenges to personalize mental health care. In a wide range of mental illness and mood disorders, more than 5 million patients in the U.S. suffer from severe, drug-resistant mental illness and can benefit immediately from new methods of treatment.
Although the technology is not yet commercially available, the goals of Paradromics include applications that focus on detecting and treating debilitating mental illness. Paradromics devices will be surgically implanted to function and will be used therapeutically once the condition is diagnosed.
“Researchers have shown that depression and mood disorders, for example, occur at the brain-network level. Hopefully, mood states can be both decoded and modulated using implanted electrodes,” Angle said. “Already we are seeing clinical trials for depression using older generation brain transplants (deep brain stimulators) and the ability to decode and modulate mood and other neuropsychiatric conditions will only improve when DDIs. [Direct Data Interfaces] Be medically available. “
Maintain privacy and eliminate prejudice
While AI can help improve equity and access when insurance coverage is low, privacy can still be a concern.
“We really need better technical solutions to show that we can protect data privacy. We should not just say that anyone who uses technology should have privacy, but should enhance the technology,” Zhang said. “For example, you can look into encryption. That technology solution could enable us to show people that data is already secure. This will help them alleviate their privacy concerns.”
Similarly, bias can cause problems throughout healthcare, so it is equally important to properly train algorithms while maintaining confidentiality.
“It’s important that we find the right model where we fully consider the human with the training data loop and we find the right workflow for medical specialists,” Muff said. Not sure if the algorithm will work on every other population, for example. It’s important that you evaluate your algorithms on medically relevant subtypes. If you don’t, it can do more harm than good. “