ARTIFICIAL INTELLIGENCE AND MACHINE LEARNING IN HEALTHCARE - THE RIGHT DATA AT THE RIGHT TIME

January 4, 2022

There has been a rapid accumulation of electronic data due to electronic medical records (EMRs), and the capture of other data such as imaging data, exam and procedure reports, lab values, pathology reports, prescriptions, waveforms, data from implantable electrophysiology devices, and data transferred from the imaging and diagnostics systems themselves.

A dilemma is that these systems and devices add to the overload of data that a physician has the human capacity to sift through. All this data can be overwhelming to physicians. Physicians don’t want more data; they want solutions that can alert them to a possible health issue so they can act. Better yet, physicians want an early warning system so they can provide proactive, preventive care.

AI ENABLED SOLUTIONS AND SERVICES

This is where Artificial Intelligence (AI) comes in. AI gives computers the capacity to compile and summarize enormous quantities of medical information and apply algorithms (machine learning/ML) to find connections - such as between clusters of symptoms to a particular disease - that a physician can’t always see. AI’s ability to quickly make connections in data can positively impact patient health outcomes.

The ability for AI to quickly sift through massive amounts of data and uncover complex associations that might be hard, if not impossible to see, puts it in a position to assist the physician in identifying health issues and recommending individualized treatment. Don’t worry, AI won’t be diagnosing patients or replacing doctors. Instead, AI will be used to augment a physician’s ability to have the relevant data they need to care for a patient and enhance clinical diagnosis and decision making.

AI applications are currently used to identify many clinical problems including pneumonia, stroke, lung or liver cancer, skin lesions, or retinopathy. Several companies have already started integrating AI into their medical imaging software systems.

AI IN ACTION

Consider this mind-blowing example of AI in action. An exam order is for chest pain. The radiologist calls up a chest computed tomography (CT) scan to read. Immediately, the AI capabilities built into the imaging system will review the image and identify potential findings. Next it will comb through the patient medical history, find the following relevant data, and present it to the radiologist for consideration:

• All the relevant data and exams specific to prior cardiac history such as history of smoking, high blood pressure or cancer

• Pharmacy information regarding drugs specific to COPD, heart failure, coronary disease and anticoagulants

• Prior imaging exams from any modality of the chest

• Prior reports for that imaging

• Prior thoracic or cardiac procedures

• Recent lab results

• Any pathology reports that relate to specimens collected from the thorax.

Without the assistance of AI the pathologist may not know that all this information exists, or because it would take too long to collect they would not have spent time looking for it, perhaps missing an important piece of the puzzle. This is one example of how AI makes the practice of medicine more efficient and safer.

Incorporating AI/ML into healthcare practices can offer quicker turnaround time to diagnosis, improved accuracy of diagnosis, earlier detection, and enhanced productivity - leading to enhanced patient care and affordability of care.