AI in Healthcare: What the Future Holds
Healthcare is a massive industry with an upward trajectory unrivaled by most other industries worldwide. The value of the global healthcare market is expected to reach over $665 billion by 2028.
One of the biggest drivers of this growth is digital transformation of healthcare facilities and processes. Leading the charge in this digital transformation is artificial intelligence (AI). As one of the most flexible and transformative technologies available right now, artificial intelligence is being utilized in new ways across many arms of the healthcare sector.
AI in healthcare dates to the 1970s, when an AI program named MYCIN was developed by researchers to identify treatments for blood infections. Less than a decade later, the American Association for Artificial Intelligence was formed, and an avalanche of progress followed.
Over the decades, the uses of artificial intelligence in the healthcare field have expanded. Below are some ways healthcare organizations are using AI to improve processes.
- Improve and speed data processing
- Develop new drugs and treatments
- Assist during emergencies
- Improve diagnostic outcomes
- Develop better tools and medical devices
- Identify high-risk patients
- Reduce human errors
- Assist surgeons during procedures
Artificial intelligence describes the use of machine learning algorithms and similar technologies. AI allows computers to mimic human cognition, examine data, analyze probabilities, learn to adapt, and predict likely outcomes. This technology is faster and more accurate than a human being in most cases and has many benefits when applied to common challenges in the healthcare sector.
This article will look at how AI is used in healthcare to improve patient care, produce better outcomes, and lead the way into a better future.
What AI May Be Like in the Future
The future of AI in healthcare is wide open, and researchers are frequently finding new uses for this technology.
At this very moment, new drug therapies and lifesaving vaccines are being developed using AI. New and improved medical diagnostic tools are helping doctors make accurate diagnoses. People are jogging down their street while wearing an AI-smart device that monitors their physical condition and sends real-time alerts to their doctors when something goes wrong.
Researchers and developers are busy working on new applications for AI in healthcare. One example of innovation in the field is the development of a tool named MAIJU, or the “motor assessment of infants jumpsuit.” A team of researchers at the Helsinki Children’s Hospital in Finland have developed a jumpsuit, which resembles a typical infant “onesie” that includes a series of motion sensors.
These sensors can assess a child’s development by identifying latent characteristics in the infant’s movements. This technology could help track a child’s neurological development and assess the performance of interventional therapy.
AI-Powered Healthcare Data Can Saves Time on Data Analysis
Data analytics describes the way information is collected and analyzed. In the case of the healthcare industry, that analysis is used to improve overall patient care and facility operations. Artificial intelligence can be used to speed the process of data analysis and form more accurate conclusions using machine learning and specific algorithms.
Healthcare facilities have access to massive amounts of data from a wide variety of sources. There has been a push in recent years to improve the patient experience by gathering direct feedback from patients. AI allows facilities the ability to collect and categorize patient responses, which can be used to make meaningful changes that would improve the quality-of-care patients receive.
Using AI in healthcare for data analysis can help organizations gain insight that can lead to improvements in processes and procedures. On a broader level, AI can recognize patterns across the industry. These patterns can predict trends in diseases, affect research and treatment plans, and can benefit facility management.
Another application for AI involves streamlining administrative tasks. It’s estimated that between 15-30% of healthcare spending, and 25% of the healthcare labor force, are involved with administrative costs, such as: medical billing, insurance management, and overall facility administration.
AI can use data analytics to streamline many of these tasks through automation. AI systems can assist in record keeping, delinquent payment follow-up, and pre-authorization of medical insurance – all tasks that would otherwise require human intervention.
Assisting Emergency Situations and Helping Medical Staff
The emergency department is the beating heart of hospital facilities, and a majority of patients are funneled through its doors. It’s estimated that around 70% of hospital inpatients are processed through the emergency department.
Hospital staff working in emergency departments must be able to handle multiple patients at the same time and are often pulled in many directions. ER doctors must manage critical cases that require them to make life-and-death decisions very quickly.
The risk of human error rises in situations where decisions are made rapidly. This is one place where AI in healthcare can be a helpful tool for improving patient care and outcomes.
AI uses data analytics to aid in decision-making, assisting doctors in their diagnosis and treatment paths in the emergency setting. It can recommend which diagnostics to run based on data inputs, saving hospitals and patients time and money. Artificial intelligence can also help determine whether a patient should be admitted or discharged based on the gathered data.
AI can also move patients through the emergency department faster by streamlining the process involved with their visit, such as prescribing medication, treatment planning, and follow-up scheduling. In the case of rural areas, where a transfer to a larger, regional medical facility is necessary, AI can help with the decisions needed to facilitate these transfers.
Detecting Heart Attacks Over the Phone
Many medical emergencies require timely intervention to prevent serious long-term consequences or death, such as strokes and heart attacks. AI solutions can help emergency medical personnel recognize a life-threatening medical emergency from the very first call to emergency services.
Corti is an AI tool that uses machine learning to analyze emergency calls. It listens to vocal cues from the caller and any background noise and considers the patient’s medical history. Based on this data, and what it has learned from previous calls, Corti can alert emergency staff if it suspects a heart attack. Initial results found that Corti correctly identifies if a patient is having a heart attack 95% of the time.
Dispatching Emergency Services Automatically
Another AI tool known as MDGo, an Israeli startup, can use data to determine if an ambulance is needed after a car crash, shortening the length of time it takes for emergency responders to arrive at an accident scene. This technology means the paramedics are dispatched within seconds, instead of minutes. MDGo estimates it could reduce auto fatalities by up to 44%.
AI Can Help Discover and Develop Medicine
One of the most beneficial uses for AI in healthcare is aiding in the development of new drug therapies. This application for AI has amazing potential, evidenced by its use to help develop vaccines for the Covid-19 virus.
There are four ways that AI can aid in drug development:
- Access to new biology
- Improved or novel chemistry
- Improved success rates
- Quicker and less expensive discovery processes
An AI Treatment for Ebola
AI can expedite the processes used to pinpoint potential pathways to new drug development, and it can do so in a much more economical and efficient way. One example of this dates to 2015 when Atomwise partnered with the University of Toronto and IBM to develop a treatment for the Ebola virus, which claimed the lives of an estimated 11,310 individuals across Africa, Europe, and the United States.
Atomwise provided AI technology that was able to analyze compounds for their potential to bind with a glycoprotein that prevented the Ebola virus from entering a cell and ranked them by effectiveness. This enabled the researchers to pinpoint their testing to the candidate with the highest potential for success. The AI analysis that led to the development of the treatment for the Ebola virus only took one day. Without that technology, the process could have taken months, even years.
AI Anti-Cancer Treatments
More recently, AI was used to discover a new anticancer molecule, an A2 receptor antagonist that helps T cells fight solid tumors.
Exscientia, a pharmatech company, used a tool they called “Centaur Chemist” to sort through millions of potential small molecules to find those with the best properties for potential success. This expedited a process that likely would have taken years and completed the task in eight short months.
The potential for AI to speed drug therapy development is changing the industry in undeniable ways. This application of AI technology can save lives on a massive scale, and researchers are harnessing the incredible value of using technology to assist in drug discovery.
Cancer Research and Treatment
Cancer research and the creation of more effective immunotherapy therapies are two areas where AI is used in the healthcare industry. This use of AI also has a significant potential to save many lives and to transform the future of the sector.
Immunotherapy uses the body’s immune system to fight back against malignancies, in the hopes of shrinking tumors. The problem is that current immunotherapies are only effective in a small percentage of patients. Currently, oncologists have no way of predicting which patients will respond positively to immunotherapy, and no quick way to test theories regarding new treatments.
AI has the potential to improve the understanding of immunotherapy by harnessing vast pools of data, and potentially uncovering new pathways to developing treatments based on a patient’s genetic makeup.
Mirai – Early Cancer Detection Through AI
Another way that AI can aid in cancer research is by examining and analyzing diagnostic imaging, which is typically the first method of discovery for many cancers, including breast cancer.
AI systems can analyze images, like mammograms, and compare the results with the knowledge gained after examining other similar images. Using that data, this advanced technology can point out areas of concern at the earliest stage, identifying potentially cancerous cells before humans could.
Being able to recognize cancerous cells as early as possible greatly increases the chances of interventions being successful. An AI tool named Mirai, developed by a group of scientists from MIT in partnership with other institutions, can use mammogram images to predict a person’s risk of developing breast cancer in the next five years – using a model with greater accuracy than current methods used to evaluate risk.
AI models can even help doctors determine if specific treatments, like surgery and supplemental interventions, might be necessary by analyzing data regarding cancer cells.
Information, such as the size of the cancer, how fast it is growing, and the likelihood of it returning after treatment can be utilized by AI to decide which treatment plans will be most effective.
Reduce Manual Errors and Improve Diagnostic Processes
Statistics put the number of deaths that occur per year because of medication errors at around 9,000. In addition, hundreds of thousands are believed to experience adverse reactions or medication complications.
The resulting cost of these errors totals $40 billion.
Employing AI to analyze data related to medical treatments, medications, side effects, and more can reduce the chances of these medication errors. It can lower the likelihood of an incorrect diagnosis and remove the risk of human error in processing prescriptions and treatment plans.
For example, MedAware developed an algorithm that can alert doctors if a medication is prescribed that conflicts with a patient’s medical profile, or if a medication has a potentially negative interaction with any current medication.
Another program, called MedEye, uses a scanner to identify specific pills and capsules, employing visual recognition to compare data from millions of similar drug administrations. In doing so, MedEye can help nurses by alerting them when a pill or capsule looks different than expected.
AI can also help doctors by improving their success rate when diagnosing patients. Studies show that between 10-15% of diagnostic errors occur in the form of missed, incorrect, or delayed diagnoses.
Doctors often use radiology tests to help guide a diagnosis, a process that takes time and often leads to unintentional errors in making a diagnosis.
AI models can help radiologists interpret diagnostic images, find anomalies and patterns that might otherwise be missed, and result in a more accurate and timely diagnosis. A correct diagnosis leads to the development of an effective treatment plan and a better chance at a successful outcome for the patient.
Patient Risk Identification Can Be Improved with AI
Identifying patients who are at risk of developing certain diseases can help doctors begin interventions sooner. However, before the intervention of technology, doctors didn’t have many tools at their disposal to do so, which forced them to rely on a patient’s verbal medical history as their main source of information.
AI in healthcare is changing that by providing doctors access to a vast amount of historical patient data. Using AI models and machine learning, this data can be used to help doctors and clinicians identify risk factors in patients that would otherwise go unnoticed.
AI systems can use data to determine a patient’s risk for readmission within 30 days after a hospital visit. Identifying these patients can lead doctors to develop a more thorough follow-up care plan, hopefully lowering the likelihood the patient will need a secondary visit.
This can help cut costs for facilities and the patients themselves.
Artificial Intelligence is Reducing the Number of Blood Infections
Researchers are also using AI to identify patients with a high risk of developing blood infections, one of the leading causes of morbidity and mortality worldwide. Early detection of blood infection is crucial to successful treatment, and delaying diagnosis increases the chance that the infection will spread to organ systems, or that the body will develop sepsis or shock.
Based on patient data such as demographic, medical history, diagnosis, and blood test results, the program was able to identify those patients with risk factors – with an accuracy of 82%.
AI to Calculate the Risk of UGI Lesions
AI helps doctors identify patients who are at high risk for upper gastrointestinal lesions, thereby identifying individuals who would benefit from an endoscopy.
Early detection of these lesions is important in their treatment, but endoscopies are not a typical diagnostic test offered to the public. By using AI to identify patients with a high risk of developing lesions, doctors can direct use of this diagnostic testing to those who stand to benefit the most.
AI to Save Graves’ Disease Patients from Heart Risks
Graves’ disease is a condition that is often associated with atrial fibrillation and heart failure. Graves’ disease is an autoimmune disorder that typically causes an overactive thyroid, a condition that can lead to increased heart rate and palpitations.
Treatments can vary depending on the severity of the patient’s symptoms and may involve the use of drugs to slow thyroid activity. Alternative treatments are more aggressive and include radiation therapy or surgery.
Using an AI model, researchers were able to identify “silent” atrial fibrillation from a patient’s initial ECG testing and found a correlation between that early discovery and a later Atrial fibrillation (AF) or heart failure incident.
Developing Efficient Radiology Tools and Other Medical Devices
Artificial intelligence can aid in the development of more effective tools and useful medical devices. The use of technology to create new tools that can help people function in their everyday lives is exciting and has amazing potential to improve people’s quality of life.
AI Robots Assisting in Surgery
Robots can help surgeons during surgery, aiding them in procedures that require extreme precision. They can lessen recovery time, reduce pain, and lower the chance of error. This use of AI in healthcare, especially in surgery settings, can greatly improve outcomes and save organizations time and money.
Improved Radiology Tools
Artificial intelligence can be used to develop more accurate radiology tools, which can then help doctors make a more informed diagnosis. AI models can recognize potential concerns in images like x-rays that can lead to early detection of conditions and improve patient care.
AI-Powered Medical Devices
Smart medical devices and machines can continuously monitor a patient’s physical condition and alert patients and doctors when issues are detected. This technology has the potential to benefit doctors and patients in a facility setting. These machines are especially useful in critical care settings, like intensive care units, where patients need 24/7 monitoring for life-threatening conditions.
AI-Powered Wearables
Wearable health devices can monitor a patient’s physical condition no matter where they are. This gives doctors a better picture of a patient’s overall health and can help identify concerns that may be overlooked in a hospital or office setting.
Final Thoughts on AI in Healthcare
The use of artificial intelligence in healthcare is already saving lives, improving people’s health and well-being, and helping healthcare organizations provide better care.
However, there is a real concern among some people that artificial intelligence could take jobs from actual human beings. The truth is that even as advanced as AI has become, there is little chance that AI could replace human beings in most places, especially in the healthcare sector.
When AI is employed in healthcare settings, or any industry, more jobs will be created because of that technology. There is already a shortage of individuals with the skills and technical expertise required to work with AI. Developers, engineers, researchers, and data scientists are just a few examples of the jobs that need to be filled right now.
Organizations will need experts to help them adopt and deploy AI technology in the field, and software engineers to help them integrate the technology into existing systems.
There are vast possibilities on the horizon for this technology. It will take time and effort, and substantial changes won’t come overnight; but we’re already seeing how many ways AI is improving the healthcare industry.
