Health Care Embraces AI

Here is a look at some of the local health care companies using AI to improve diagnosis, treatment, and care for patients.

Renee

Dr. Renee Dua has served as a clinical instructor at the University of Southern California and the chief of medicine at Valley Presbyterian Hospital and Simi Valley Hospital. She founded in-home primary care practice Get Heal Inc., dba Heal with her husband, Nick Desai, in 2014. The two then moved towards utilizing artificial intelligence in clinical care with their new startup, Santa Monica-based SixD Inc.  

Renee provides a platform for patients, particularly aging and polychronic individuals, to manage doctor appointments and prescriptions. Dua and Desai founded the company in 2021, and Desai is its chief executive. The Renee platform applies AI with an eponymous assistant, which handles appointment scheduling, the arrangement of rides to and from offices, prescription delivery and tracking of patient health. “Our team is using AI to support a holistic patient experience, from ensuring the patient is adhering to daily medication and refills are being picked up to booking appointments and checking blood pressure using the camera phone,” Dua said. “Further, based on the patient data collected … we’re also using AI to share personalized recommendations for testing that may have been missed or incomplete, so that both patient and caregiver can be proactive in their quest for wellness.” 

Dua said that the company’s biggest benefit is task simplification. She said that responsibilities such as taking medication, picking up refills or remembering to schedule preventative-care appointments can be easily forgotten as people age, and that the company’s AI application fills in these gaps. 

Predict View

PredictView got its start when software developer and entrepreneur Randy Waters and sports technology entrepreneur Ross Comerford were both working for Comerford’s sports technology company. The pair began developing an application for athletic organizations to manage behavioral issues with their athletes. They started using artificial intelligence algorithms to identify signs of depression, substance use, and anxiety from patient online content such as social media posts. Early on, their program allowed their client to identify and prevent a suicide. That’s when Waters and Comerford realized the broader applications of their technology and turned it into a tool for mental health therapists to improve outcomes for their patients.

“Millions of people share their feelings and mental health issues openly on social media platforms, often with nameless faceless strangers,” Waters said. “However, they don’t always share this information with their therapists. While a growing number of therapists do follow their patients online, most therapists don’t have the means or the time to scroll through pages and pages of multiple social media platforms. Our application provides therapists new insights about their patient that they don’t already have.” 

The goal is to prevent mental health breakdowns or other problems with patients.

“We want to ultimately eliminate the sad realization of, ‘Gee, if we had only known,’ when someone’s mental state takes a dramatically bad turn or they take their own life,” Waters said.  

The company officially launched in late 2020 and has raised $6.3 million to date. A patent is pending on the artificial intelligence program that powers its platform.

SpectrumAI

Signal Hill-based SpectrumAi Inc. was founded in 2021 by Ling Shao, a parent with four children on the autism spectrum. Its mission is to increase the quality of applied behavioral analysis therapy, or ABA, the primary treatment for people with autism. SpectrumAi works by using artificial intelligence to help behavior technicians and analysts more accurately and efficiently record patients’ progress in ABA therapy and treatment outcomes. 

The company said that it’s not looking to create the “best machine-learning model” that can replace humans. SpectrumAi head of engineering Joseph Adu said many ABA providers lack extensive experience conducting ABA, and tools that increase data analysis of outcomes and patient results are crucial. 

“Much of what has been written about AI has focused on how it can eliminate the need for human intelligence,” Adu said. “We believe that this is not right for health care generally and ABA specifically, but instead espouse to create ‘human-in-the-loop’ products that leverage artificial intelligence as an aid, not a replacement.” 

Adu said that because SpectrumAi’s platform looks to supplement providers’ abilities, rather than replace them, he feels unconcerned about the potential for new regulations to restrict the company’s development.  “Practitioners and patients will materially benefit from tools that use AI to generate beneficial insights from previously untapped massive and complex data sets,” Adu said. “It’s also worth remembering that there are already some very sensible regulations in place that ensure that technology is used in a safe and responsible way within health care — for example, HIPAA and various requirements around the development of medical devices used in diagnostics.”

Deep6 AI

One of the most frustrating things for pharmaceutical companies seeking to bring drugs to market is the length of time it takes to set up multiple rounds of clinical trials, including casting a wide enough net to find the appropriate patients from diverse backgrounds. 

“Manually reading through patient data chart by chart is tedious and time consuming,” said Tim Keliher, chief revenue officer for Deep6 AI. “Each trial has inclusion and exclusion criteria which dictate if a patient is eligible for that trial. Each of those criteria must be met, and this is only determined by reviewing all the medical data found in a patient’s records, from doctor’s notes to pathology reports to family history.” 

Deep 6 AI has developed a platform that applies AI to all this patient data. It organizes and analyzes the patient data as it seeks to identify the right patients and sites for clinical trials in real time.  

“While it can traditionally take many months to fill out a trial with all the right participants, our AI allows users to accomplish this on day one,” Keliher said. 

This in speeds up the process of clinical trials and gets drugs before the Food and Drug Administration more quickly for approval. Through this rapid analysis, Deep 6 AI’s platform also enables clinical trial operators and their pharmaceutical company clients to cast a wider net and bring in more diverse populations. This helps to address a criticism of pharmaceutical companies that their drug products are skewed toward white males. 

RadNet

RadNet was founded in 1984, and its mission for the last 30 years has been to create comprehensive radiology solutions. The Sawtelle-based company is using artificial intelligence to improve the accuracy and efficiency of radiology and diagnostic imaging services. In addition to these innovation efforts, RadNet operates more than 350 imaging centers across the U.S. Its projects include a RadNet TV platform as well as partnerships with Beverly Grove-based Cedars-Sinai Medical Center and Dignity Health, among others. 

In terms of its AI solutions, RadNet’s algorithms specifically help radiologists detect three major types of cancer: breast, lung and prostate.

“In our industry, AI applications can assist radiologists in making more accurate diagnoses earlier in a disease process, with greater efficiency,” said Mark Stolper, RadNet’s chief financial officer. “It has been proven that earlier and more accurate detection of disease leads to better patient outcomes and lower cost to the health care delivery system. We believe that AI will enable wide-scale, cost-effective health-screening programs that will yield a healthier population through non-invasive prevention and earlier detection.” 

The company launched an AI-powered suite of tools specifically for mammography last fall called DeepHealth. Stolper said that it can detect breast cancer up to two years sooner than “even the best” radiologists can. DeepHealth will be made available at its Los Angeles and Southern California locations later this summer. RadNet has received clearance from the Food and Drug Administration for its Saige-DX mammography diagnostic software and its Quantib software, which reads prostate MRIs. Stolper said he feels that AI in health technology will likely not see the same potential regulations that other types of generative AI are facing. 

Avenda Health

Culver City-based Avenda Health Inc. is focusing its artificial intelligence technology on improving treatment outcomes for people with prostate cancer. The company was founded in 2017 by Shyam Natarajan, a specialist in targeted prostate biopsies, and medical device and digital therapeutics specialist Brittany Berry-Pusey, who now serve as chief executive and chief operating officer, respectively. 

Avenda’s AI platform, called Unfold, maps a patient’s cancer in 3D to allow oncologists to better determine in advance where to intervene. It also draws on other information, including magnetic resonance imaging, PSA screening results and biopsy results, to round out the analysis of the cancer. This offers a more personalized and targeted approach to therapy.

“Its purpose is to assist physicians in making informed decisions regarding the next steps for patients,” Berry-Pusey said. “Unlike current AI tools primarily utilized in digital radiology or digital pathology for diagnosis, Unfold AI leverages multi-modal information such as MRI, targeted biopsy, pathology, and PSA results. This comprehensive approach is crucial because relying solely on MRI tends to underestimate the extent of the disease.” 

Avenda has chosen to apply this AI technology first to prostate cancer, with urologists as its main initial customer base. 

Precision in intervention is especially important in treating prostate cancer because of the high risk of damaging nearby healthy tissue and impairing bladder function.

“We are excited about the potential to unlock precision care in prostate cancer with Unfold AI, as it is a key enabling technology for focal therapy to be a reality for urologists and patients,” Natarajan, Avenda’s chief executive, said. “In order for a doctor to treat focally, they need to know where cancer is and the healthy tissue to avoid. Last December, Avenda Health received Food and Drug Administration approval for its Unfold AI cancer analysis platform.

Pearl

Pearl was born out of the convergence of its three co-founders, Ophir Tanz, Kyle Stanley and Cambron Carter, at GumGum, a digital-media company that Tanz founded in 2008. First, Carter joined Tanz and the pair focused on applying AI technology – then in its infancy –the world of digital marketing and advertising. Stanley joined them in 2015, at which point the trio decided to start a health care company that used an AI platform. Because Stanley and Tanz’ fathers were both dentists, they decided to found a company that applied AI to the interpretation of dental X-rays.

Pearl uses AI to detect and evaluate diseases and other clinically relevant features in X-rays and 3D imagery. The technology applies generative algorithms, language processing and analytics to make sense of patient charts and treatment data. This allows the program to better detect cavities, abscesses and other conditions.

The aim of Pearl’s main program, called Second Opinion, is to help dentists read X-rays more accurately. In clinical studies of the program’s effectiveness, dentists using the program detected cavities on average 37% more accurately than dentists not using it. Another benefit: since patients can see the system’s findings alongside the dentist, the patients are more trusting of diagnoses.

In 2021, Pearl’s AI platform became the first to receive an OK from the FDA to read bitewing and whole-tooth X-rays in patients as young as 12 years old. According to Tanz, because Pearl has already obtained this authorization, company executives don’t believe any forthcoming AI regulations will be any more stringent than what the company had to go through to obtain that authorization. Last year, Pearl partnered with Westlake-based teledentistry firm Dentulu Inc. to begin offering its AI platform to Dentulu’s teledentists at 2,000 participating dental practices to use in diagnosing patients online.

Wellth

Wellth, which was founded in 2014, has developed a phone app that reminds patients to take medications and requires them to submit photos of themselves with the pills or dosages in their hand, ready to take. Patients also take photos of themselves monitoring their glucose levels and other activities.Patients achieving high levels of compliance are rewarded with payments on a stored-value credit card they receive when they sign up.

Underlying this is Wellth’s behavioral health platform. According to the company, this daily interaction turns behaviors into habits, which ultimately leads to lasting change and more positive health outcomes. And this is before the application of artificial intelligence algorithms.

“At the core, we’re removing barriers and giving patients the motivation to do what’s best for their health,” said Matthew Loper, Wellth’s chief executive and co-founder.

Wellth’s platform uses AI to identify and recognize the images and then draws upon biometric and other sources of information to understand and plot out patients’ behavior patterns. If the patient begins to deviate from the established patterns the program sends out alert messages and notifies Wellth’s human support staff, who can then work to get the patient back on track.

Health care organizations whose members use Wellth experience a 42% reduction in inpatient hospitalization on average, nearly one-third fewer emergency visits, and a 16% improvement in medication adherence, according to Alec Zopf, Wellth’s chief technology officer and co-founder. Those results helped Wellth secure $20 million in series B financing earlier this year, bringing its total raised to roughly $40 million.

Zopf said Wellth would welcome some regulation of artificial intelligence programs.

“AI is foundational to Wellth’s applications, which help health care organizations to save lives and ultimately lower health care costs for everyone involved,” Zopf said. “It’s our hope that well-designed regulations can empower organizations like ours to use AI for the greater good, supporting broader population-health goals across the country.”

READ FULL ARTICLE HERE