27
AI in healthcare: is the industry ready for it and how does it use AI?
As the Healthcare Global Forecast to 2026 suggests, artificial intelligence (AI) use in healthcare is to experience an immense rise both in terms of revenue and patient health. While only a couple of years ago the use of AI in the healthcare industry was somewhat limited, nowadays the range of use cases is mind-boggling. Drug discovery, patient care, illness diagnostics, you name it.
The use of AI in healthcare goes beyond helping with R&D and AI-powered diagnostic; the healthcare industry makes use of conversational AI to automate all sorts of processes.
In this post, we’ll take a look at the possible uses of AI in healthcare.
Originally posted on dasha.ai/en-us/blog/ai-in-healthcare
The projected AI in the healthcare market size is expected to go from over $4 billion as we saw in 2020 to over $45 billion. Such growth is attributed to the widening adoption of AI use in the healthcare industry due to the technological advancements and increase in the pool of skilled AI workforce.
Though these numbers show that the healthcare industry is predicted to be adopting AI at a fast pace, there are pitfalls. Healthcare is a complex industry. Relying on artificial intelligence brings its own risks. It’s not only high-level tasks such as having an AI-powered robot performing surgeries that carry risks when it comes to a patient's health. A benign-looking AI symptom checker can carry as much of a risk since those ill with something might resort to checking their symptoms online instead of seeing a doctor. Any healthcare-related AI must go through rigorous testing and evaluation before seeing the light of day, which slows down the process of its adoption because otherwise human lives could be at stake. Since federal law protects patient’s information, the healthcare AI providers should (and do) constantly refine security measures as any breach of data can lead to the healthcare organization’s financial losses as well as legal issues.
Ensuring that the healthcare AI security and privacy measures to be top notch remains the priority. Yet another concern still prevails among some individuals and that’s that AI will ‘steal’ jobs in the years to come. This, however, shouldn’t be seen in a negative way since more jobs will be created. Healthcare AI innovators strive to create a symbiotic relationship between physicians and technology since while artificial intelligence can do things faster and smarter, at the end of the day it’s the humans who interpret the results and draw conclusions. With AI, the efficiency and accuracy of research, diagnostics and treatment are bound to boost, which leads to physicians having more time to tend to patients in a more personalized way and, as an end result, create a healthier population.
So is the healthcare industry ready for AI? Let’s look at some numbers.
As per Medscape report, 70% of the physicians surveyed said that they believe their decisions will be more accurate with help of AI-powered software, and 66% of the respondents agreed that they’ll use it if it surpasses human abilities when it comes to diagnostic tasks. Since AI is great at automating repetitive tasks, 68% of physicians said that employing AI will help them concentrate on more meaningful and important tasks.
The Covid-19 pandemic fast-tracked the adoption of AI technology in the healthcare sector. 56% of healthcare executives said that they accelerated or expanded their AI use, as per Health IT Analytics report. Another Health IT Analytics report states that 55% of the healthcare leaders said that AI adoption improves patient’s health outcomes and another 55% said that the biggest benefit of AI is improved patient experience.
While there are some issues healthcare AI providers must work on, the outlook on AI readiness is positive. The healthcare sector is becoming increasingly up for adopting AI and recognizes that the future of AI in healthcare is bright.
As Deloitte puts it, there are three ways the healthcare industry uses AI:
- Clinician-oriented AI,
- Patient-oriented AI,
- Administrative- and operational-oriented AI.
Let’s take a look at each of these.
Clinical-oriented AI use includes medical imaging technology (MIT), which relies on computational models and algorithms based on bioinformatics, and various pathology reports. Such AI uses large sets of data to provide physicians with answers to diagnostic questions as well as potential treatment options. The benefit of using clinical-oriented AI is that it can find patterns that would otherwise be difficult for a human to find, therefore speeding up the process of diagnostics and treatment. It provides recommendations based on specific data for physicians to judge, accept or reject based on their own experience and expertise.
In the previous section we discussed the synergy between AI and humans is what brings efficiency to the healthcare system and Michael Brady’s (Professor of Oncological Imaging at the University of Oxford) words back this idea up: “When we combine AI-based imaging technologies together with radiologists, what we have found is that the combination of the AI technology and the radiologist outperforms either the AI or the radiologist by themselves.”
Eric Topol, a cardiologist and executive vice president of Scripps Research, states that in his 30+ years of experience in the medical field he found AI to be the top resource that transforms and impacts the healthcare industry the most. Various AI technologies help patients monitor their own health, arrange appointments with healthcare professionals, collect daily health data and send it to the healthcare provider. Among all existing healthcare AI tech, voice/conversational AI is the area with the highest potential, and according to Karder: “voice technology is even considered as one of the most promising sectors, with healthcare being predicted to be a dominant vertical in voice applications. By 2024, the global voice market is expected to represent up to USD 5,843.8 million”.
In order to make the patients’ lives easier and healthier, AI companies create platforms and apps like AI-powered conversational mental health management apps, symptom checkers and voice biomarkers that can detect certain illnesses.
With machine learning techniques, a mental health management conversational AI app can analyze the tone and the speed of the patient’s voice and provide recommendations based on that. It also could be programmed to analyze patterns of a patient’s mood, activities, and overall mental state over time and, if necessary, make appointments with healthcare professionals. For instance, if it recognizes either by speech tone or words that the patient is stressed out, it could suggest a pre-programmed range of meditations or provide a range of relaxation techniques. Such conversational AI apps could be integrated with various sources of health trackers (heart rate, sleep, diet, blood pressure, sugar level) and make recommendations according to those indicators.
Such a mental health management AI app can be easily created with Dasha conversational AI API and integrated with various services and health trackers to provide further benefits for the patients.
Healthcare companies use voice AI to diagnose certain disorders. In the mental health arena, conversational AI might initiate a chat with you and based on the tone, pitch, rhyme, and energy detect whether you are suffering from depression. In fact, such technology changes the healthcare system as it has limited capacity to take every patient in during the time of need and the lives of the patients. According to Scientific American, most American citizens suffering from depression aren’t receiving the necessary treatment. However, with the help of AI, mental disorders like depression and anxiety can be discovered on time and treated by professionals. Of course this is a slippery slope that not every company will set out on. After all, if the AI is tasked to recognize early signs of, for example, suicide, and fails - a human life is at stake. We turned down the offer to use Dasha AI for such a use case two years ago due to ethical considerations.
Another way healthcare companies use vocal biomarkers is by assessing patients’ voices to diagnose Covid-19. The virus was at the outset widely spread by those unaware of being infected. The Massachusetts Institute of Technology published an article saying that its “researchers have now found that people who are asymptomatic may differ from healthy individuals in the way that they cough. These differences are not decipherable to the human ear. But it turns out that they can be picked up by artificial intelligence.” The AI algorithm uses vocal sentiments to analyze the difference in the voice of a sick and healthy individual and advises those whose voice sounded sick to stay home.
The same logic applies to detecting Alzheimer’s disorder: artificial intelligence detects the subtleties in the mood, as those with this disorder show more frustration and indifference rather than happiness and calmness, which can be picked up from their voice. According to MIT, "the researchers developed a sentiment speech classifier model by training it on a large dataset of actors intonating emotional states, such as neutral, calm, happy, and sad” and trained their AI model to distinguish between those to get accurate results when it came to assessing voice.
While vocal biomarkers are a great way of AI diagnosing disorders, it doesn’t stop there. AI symptom checkers are on the rise too, and as time goes, they become more accurate than ever.
It’s tempting to ask Dr. Google for ‘medical advice’ when we feel sick. However, in practice it’s most likely we’ll self-diagnose cancer, petrify ourselves, and schedule an appointment with an oncologist.
There are 2 main issues that AI symptom checkers are solving. On the healthcare sector side, it prevents unnecessary doctor visits. On the patient side, it provides accurate information on potential illnesses and refers the patient to a specific doctor or clinic (it can even suggest a physician in the area the patient lives).
According to Medical Device Network, “Documenting the visit takes as much as 40% of the overall time that you spend with your GP”, and employing AI to collect all the necessary data pre-visit saves time and provides a preliminary insight on a potential condition before the patient enters the clinic.
A symptom checker conversational AI app is scriptable, therefore could be easily created with Dasha conversational AI.
As mentioned, conversational AI is widely used for customer service automation. Whenever patients call the customer service number, they don’t want to be put on hold or asked to go to a health institution to get their issue solved; they want a quick response and quick results. The pandemic showed the awfulness of having to be on the line for hours. The healthcare industry should be mindful of not only improving the health of the people but also of providing them with a convenient and fast problem-solving solution, such as automating customer service with conversational AI. Any scriptable call can be automated, from patient registration to data entry to doctor appointment scheduling.
In terms of appointment reminders, a study showed that over 75% of patients prefer getting appointment reminders by receiving a call from an AI assistant. Want to get an appointment reminder digital assistant set up? You can create your own conversational AI appointment confirmation app by following this guide.
AI focused on automating repetitive administrative and operational tasks is appealing to physicians as it removes them from tedious data management, repetitive question answering and offers time to tend to patients. According to MIT Technology Review, 60% of medical staff said that when they’re equipped with AI-powered tech, they spend more time performing procedures, which is what matters most. On top of that, nearly 80%
said that AI helped to prevent burnout.
Artificial intelligence presents an opportunity for a healthier future. Not only it can process complex clinical data, assist with researching new medicine, help perform surgeries, and fight disorders, but also creates a more personalized experience for patients and lets them and their doctors have more control over their health status.
Conversational AI has huge potential to be employed for patient care as well as for making the day-to-day lives of healthcare providers easier. Now, why not create a conversational AI app that suits the needs of your healthcare institution?
27