Source: Businesswire

Artificial intelligence (AI) has caused a seismic shift in all the industries it has been applied to. One such industry is medicine where K Health is using an AI chatbot to diagnose patients before connecting them to doctors or nurses for treatment.

AI in Medicine: A Bold Feat

The integration of AI into medicine has become a massive undertaking in the industry as doctors and stakeholders work to better the services they offer using the technology. According to a report by Allied Analytics, in 2018, the global healthcare chatbots market was valued at $116.9 million and it is expected to reach $345.3 million by 2026.

This growth makes it a lucrative space to venture into. However, from past experiences such as with the internet, integrating technology to play the role of a doctor even in the slightest capacity has been difficult.

The internet, for instance, has proven to be unreliable in giving diagnoses by either underdiagnosing a patient or giving an extreme condition that is more often than not a false diagnosis.

Regardless, K Health, an Israeli startup, has ventured into the area and has succeeded in offering over 3 million patients a diagnosis and access to further medical assistance. The company’s AI works such that it asks a series of questions that a doctor would typically ask in order to derive the precise signs and symptoms.

After that, the bot goes through its database, scraping other patients’ records to come up with a list of potential diagnoses. Thereafter, it provides patients with the alternative of speaking to a doctor or nurse for further medical attention.

K Health’s AI chat bot does not replace a doctor, instead, it works as a digital medical assistant, offering the doctor a starting point. “Doctors spend a lot of time collecting information from forms asking basic questions that machines can do,” says the company’s CEO and co-founder, Allon Bloch.

Not only does the bot serve the patient, but it also helps the doctor by providing them with an overview of the patient’s status as well as a list of potential diagnoses. This saves the doctors time and allows them to focus on the patient’s treatment more.

So far, the chatbot assists with primary care, urgent care, and some pediatric services, as well as chronic disease treatment, such as weight loss management.

K Health Maneuvers the Challenges of AI in Healthcare

Source:MobiHealthNews

Developing a medical AI chatbot and putting it in production is a bold feat considering how complex and high-risk such systems are.

One major concern is normally the accuracy of such systems since a small error could easily result in the loss of life. As a result, many systems have been built using AI but very few are in production despite having what would qualify as high accuracy in other industries.

For K Health, a retrospective analysis of over 100,000 patient visits at K Health between October 2022 and January 2023, showed that 84.2% of the time, human doctors and nurses agreed with an AI-recommended diagnosis. Moreover, 60.9% of the time, medical practitioners agreed with the top-ranked diagnosis provided by the AI chatbot. Notably, the study has not yet been peer-reviewed.

The chatbot’s accuracy is varied in that it performs better at diagnosing some conditions than others. According to the study, high levels of accuracy were obtained for bladder infections and pink eye, while the model proved to be less accurate for issues surrounding symptoms of skin rashes and abdominal pain.

This variability however is possibly due to the availability of data with such issues, says Dan Zeltzer, an economics professor at Tel Aviv University who studies healthcare decision-making and was the lead author of the study.

Another difficulty in developing such models is data, its availability, biases, and privacy concerns. Finding data to train the model was difficult in the US since the licensable data from medical insurance lacked medical notes.

Luckily, when Bloch and his team looked for data in Israel, they were able to license a dataset from Maccabi, an Israeli health insurance company, that included anonymous records of 2.1 million patients over 20 years. Maccabi provided high-quality data without breaching its users’ privacy, enabling the development of K Health’s AI chatbot.

Like ChatGPT, the AI chatbot has the ability to learn from its users’ information as well as feedback from the doctors and nurses that use it. This has enabled it to become very good at predicting the next best question to ask a patient when gathering symptoms.

The model is still a work in progress and in some cases it may fail to provide a diagnosis. Unlike ChatGPT, however, Bloch says, the AI cannot make up a fake disease that doesn’t exist. K Health is working on finetuning its model while experimenting with others with the goal of being good enough to IPO the company.

Despite being in existence for 7 years, Bloch acknowledges that being in the medical industry, the company still has more to do before it is able to go public. “Now in the future, is there a possibility to fully automate certain diagnoses and treatments? I think so down the road, but we’re not ready for it yet,” he said.

The Medical Unicorn

As a business venture, the company has yet to make profits despite generating $53 million in revenue last year. According to Bloch, the company’s business is through direct contact with consumers(40%) while the remaining is through enterprise contracts such as with Elevance Health, an insurance company.

K Health has grown since its launch, seven years ago, from being a symptom checker in 2018 to working with three doctors in two states in 2019 to now working with doctors across 48 states.

Through its existence, K Health AI has worked with over 10 million patients, with 7 out of 10 of its patients nowadays choosing to see a doctor after a chat.

The AI-based platform has managed to attract the interest of investors and in 2021, K Health raised $59 million from the biggest hospital in California, Cedars Sinai, and other investors. However, for this funding round, the medical AI company did not release its valuation.

Prior to this, the company had been valued at $1.5 billion when it raised $132 million. Unfortunately, due to the harsh market conditions that have caused a decline in unicorn valuations, K Health is estimated to be worth $800 million by Forbes. In total, the company has raised $330.3 million.

While these numbers are impressive, the service has also been on the receiving end of a lot of criticism and complaints. Some of its patients complained about having to see a doctor after paying to speak with the chatbot. This shows just how much patients expect AI to do and how much AI still has to be advanced in the medical space to be able to replace doctors or move patience away from the problematic ‘Google Doctor’.

What's the Best Crypto to Buy Now?

  • B2C Listed the Top Rated Cryptocurrencies for 2023
  • Get Early Access to Presales & Private Sales
  • KYC Verified & Audited, Public Teams
  • Most Voted for Tokens on CoinSniper
  • Upcoming Listings on Exchanges, NFT Drops
See the 15+ Coins