Technological advancements — in conjunction with innovative thinking — have led to changes in the radiology industry over the last few years, which has delivered major advantages to both radiologists and patients. These solutions allow radiologists to deliver higher quality services faster than ever before while also helping prevent workplace burnout within the industry.
However, determining which technologies to add to your diagnostic toolbox can be a challenge. To take full advantage of these benefits, radiologists should educate themselves on what is available and then equip themselves with the solutions that will provide them with the most benefit. As a result, these radiologists will have the ability to better compete on price and value, while offering patients a higher quality of service than those who do not adopt these innovative practices.
As we look toward a new decade, below are industry technology trends to watch out for in 2020.
Hyperautomation refers to the use of technology to automate tasks. By applying advanced technologies, such as machine learning (ML) and artificial intelligence (AI), processes that previously required humans can now be completed via automatization. This technology can be used to automate numerous tools, which is essential because a single tool cannot successfully replace a human being.
One way that hyperautomation can be leveraged effectively, for instance, is by better streamlining and automating patient communication, such as through automated text messaging. By leveraging this process, patients can receive real-time secure messaging on updates related to their appointments, updates on when scan results are ready and more.
It is important to know that successful hyperautomation requires the implementation of a combination of tools, including robotic process automation (RPA), intelligent business management software (iBPMS) and AI. Ultimately, the end goal of hyperautomation is to increase AI-driven decisions.
Artificial Intelligence and Machine Learning
The uses of AI and ML are breaking barriers in radiology and offering major improvements in the efficiency, quantity and quality of work. This includes improving diagnostics, allowing for continuous training and ensuring proper security measures are in place.
AI assistance to improve diagnostics. Thanks to the combination of AI and ML, diagnostic tools can now be trained to read radiologic scans and tissue samples. Radiologists who decide to take advantage of this cutting-edge technology can expect to expedite patient diagnoses. Furthermore, with ability to automate monotonous tasks and assist with prioritizing cases or secondary reads according to tumor detection, radiologists will be able to provide patients with a more precise diagnosis.
AI will allow for continuous training. The use of AI and ML will allow health professionals to continuously train through the creation of feedback loops, which is the process where the inputs of a system are plugged back in and are used as iterative inputs. While the algorithm gains new information from the perpetual loop, it can easily allow for continuous training for the radiologist. For example, if a radiologist is notified of a pathology and/or surgery outcome, he or she can easily access previous cases similar to the outcome or diagnosis as part learn more and better improve future processes.
AI security. The use of this technology is also incredibly beneficial in cybersecurity. In 2019, nearly four out of the five industry cybersecurity breaches that occurred involved a healthcare organization. By the end of 2019, healthcare data breaches will total $4 billion and it is anticipated that the cost of 2020’s data breached may be even higher.
What’s more, according to a Black Book Market Research survey, 96 percent of the IT professionals who took the survey agreed that the data attackers are outpacing the medical enterprises. The survey also showed that more than 50 percent of the attacks in 2019 occurred due to external hacking.
Although hyperautomation has many benefits, it can also lead to vulnerabilities; therefore, radiologists must implement an effective AI security solution that can understand patterns and protect data, which must include the following:
• Protecting all AI-powered systems — training data, ML models and training pipelines;
• Leveraging AI to improve security defense — use ML as a means to understand patterns, uncover attacks and use automation in parts of the cybersecurity process; and
• Anticipating nefarious use of artificial intelligence — identify attacks and protect the data.
While there are a few AI-assisted workflow management systems that are currently going through the testing phase, there are several more that are still being developed. However, for these systems to be beneficial, they need to be based on the radiologist’s workload and be capable of the following:
• Planning and organizing exams automatically by using various factors such as time of day, department and specialty.
• Balancing workloads automatically by matching the radiology exam workload with reading capacity.
• Automatically escalate and assign studies based on the availability of the radiologist, location, subspecialties and time of day.
• Sending specific exams to be analyzed by artificial intelligence first. Specific exams may be based on the type of examand the request; for example, AI-assisted pre-analysis may be requested for an exam that detects a lung embolism.
• Prioritizing radiologists’ reading queues and setting configurable deadlines for escalating and monitoring service level agreement (SLA) exams.
In the future, AI-based workflow systems will have the ability to use and apply data, as well as assist with operations management. For example, these systems will eventually be able to recognize bottlenecks and use that data to route exams to the radiologists that are available, reducing waiting times and streamlining the delivery of care. Implementing an AI-based workflow orchestration system early on allows radiologists to get a head start by increasing their productivity and overall reading workflow.
It is anticipated that forecasting capabilities and predictive analytics will soon become more readily available. With the ability to forecast capabilities based on patient data generated from electronic medical records (EMR) and other systems, this feature will be an essential component of tomorrow’s workflow system.
Cross-discipline Collaboration Within Departments
To provide optimal care, the industry has become increasingly dependent on collaborating with other specialists, giving radiologists the ability to take the lead in the management of cross-disciplinary workflows. Since radiology is one of the most IT-knowledgeable disciplines within the healthcare industry, radiologists will have no problem applying integrated diagnostics using digital technology. Now that pathologists are moving toward reviewing images digitally, a more disciplined process for cross-collaboration is on the horizon.
Dhruv Chopra’s experience in radiology began well before his tenure as CEO of Collaborative Imaging. He previously spent 12 years as an executive for a leading billing company in the radiology industry where he gained an appreciation for how much physician money is lost due to the inefficiencies with data gathering, claim submission and adjudication, documentation, charge capture and coding, and the ever-changing requirements of insurance carriers. Unable to tolerate the revenue losses that physicians were facing as well as the workflow inefficiencies that plagued radiology practices, Chopra created his own path to combat these obstacles with the formation of Collaborative Imaging.