Amidst the turbulence caused by the pandemic, we witnessed impressive innovation in healthcare. Overburdened like never, medical facilities devised new ways of working with the help of artificial intelligence. Take the example of Mount Sinai, which used AI algorithms to identify patients ready for discharge during the first wave in New York. From disinfection robots to seamless PPE delivery and contact tracing, AI played its optimal role in several other cases. It is not surprising that a recent Intel survey found that the number of medical organizations already using AI or planning to use AI has doubled after the onset of the pandemic.
Towards preventive care
Experts feel AI can significantly help in delivering personalized and preventive care. In the healthcare industry, vast amounts of data are generated through research, physicians and clinics, wearables, patients, and non-physician clinical workers. Machine learning helps collect this data effectively and analyzes it to drive meaningful action. Even today, while we rely heavily on the patient for information, the truth is that preventive care would be possible with the ability to analyze patient data from multiple sources and deliver it to medical professionals. According to CDC, preventive care could save close to 1,00,000 lives in the U.S. alone.
Today’s paradigm is disease management, where most practitioners wait for patients to become ill before starting treatment. However, medical organizations will focus on disease prevention in the future, which is possible by proactively monitoring healthy individuals. Take, for instance, an overweight male who has diabetes and is also a smoker. With access to a wearable device, the patient can monitor his glucose levels, heartbeat, and exercise levels. This information can be synced with a central monitoring system and the signals used to train ML algorithms. When the algorithms detect abnormal deviations or slight pattern changes, the patient’s healthcare provider is notified, and emergency care is enabled.
A paper published by NCBI cites that with the complexity of data in healthcare, AI will be increasingly employed by payers and providers. Besides preventive care, the critical areas of applications will include:
In medical diagnosis
It has been found that 17% of complications and 10% of patient deaths 1 directly result from wrong diagnoses. Disease diagnosis can be tricky as, although large datasets are available, tools that can detect patterns accurately and with precision are limited. The use of AI, especially machine learning, reduces detection errors significantly and supports the diagnostic process. Let us take the example of neurodegenerative diseases where a heterogeneous study population and complex molecular mechanisms pose severe challenges to accurate diagnosis and treatment. Here machine learning algorithms have been applied to electronic health records and longitudinal patient data collection, which help stratify the patients and predict the prognosis. Along with NLP, ML facilitates robust interrogation of multiple datasets and detects undiscovered patterns and correlations.
In remote monitoring
Besides medical diagnosis, artificial intelligence will also play a critical role in remote monitoring. According to ScienceDirect.com, it has been predicted that telehealth and wearables which enable remote monitoring will play an increasingly significant role in future healthcare delivery models. While we would like to believe that hospitals are considerably safe, the reality is that with prolonged stays, the risk of getting infected by spreading viruses remains large, almost close to 17.6% 2. Making matters worse are the manual spot-checking methods which increase the risk of infection resulting in infrequent visits by the medical staff, unmonitored patients, and spotty data.
The deployment of remote monitoring systems and AI-powered patient analytics empowers medical teams to glean valuable insights, which are further used to direct patient care to those who need it the most. One good example of an AI feature is the early warning score system, where the algorithms are built on large datasets and help detect potential deterioration early. With this, the medical staff is better equipped to tackle critical conditions and take preventive actions.
Personalizing Treatment
While there is no substitute for the human touch in treating a person, AI can complement the efforts of healthcare professionals thereby alleviating the pressure and improving outcomes. Probably, it is only a matter of time before personalized treatment becomes the norm. Aided by machine learning that can automate complicated statistical work, medical professionals will be able to monitor the response of a particular patient to a specific treatment in real-time. The resulting outcome is a more effective treatment as doctors make informed decisions and design the right treatment plan.
Thus, AI will merge with medicine in newer ways, helping save lives which will benefit global health at large. The pandemic has already reinforced that, when combined with purposeful technology, human intelligence can drive inspirational results. Then why not take the lessons forward?
Resources:
1 . Diagnostic Safety and Quality | Agency for Healthcare Research and Quality (ahrq.gov)
2 . https://pubmed.ncbi.nlm.nih.gov/21945976/
Contributor
Sohini Deb
Lead Content Writer, Marketing
