To err is human, to AI is to define
Radiology is becoming an increasingly critical aspect of medicine. There are more studies, more images, and more data. Radiologists are required to flex their expertise to make more decisions and comparisons. Radiology departments are facing shrinkage in human resources. Technology is introducing more dynamic and scalable solutions that allow for deeper analysis and richer integration, but this, in turn, pushes out more data, more studies and more images. However, technology also forms the backbone of the solution to these challenges. Artificial intelligence (AI) is increasingly embedding itself into the fabric of radiology solutions, providing the medical professional with additional support, data and information with which to make deeper and more relevant medical diagnoses and decisions.
Artificial Intelligence in Radiology
According to the study entitled ‘Artificial intelligence in radiology’ released by Ahmed Hosy, Chintan Parmer et al, in Nature Reviews Cancer, AI methods excel at ‘automatically recognizing complex patterns in imaging data and providing quantitative, rather than qualitative, assessments of radiographic characteristics’. Another study – Stand-Alone Artificial Intelligence for Breast Cancer Detection in Mammography: Comparison with 101 Radiologists’ by Rodriguez-Ruiz et al, found that ‘the performance of the AI system was statistically noninferior to that of the average of the 101 radiologists’. The AI system had a 0.840 (95% confidence interval [CI] = 0.820 to 0.860) area under the ROC curve and the average of the radiologists was 0.814 (95% CI = 0.787 to 0.841).
The results showcase that AI has inordinate potential when integrated into the PACS/RIS radiology workflow, however, it does not indicate that the role of the radiologist is defunct. Conversely, it emphasizes the value that medical professionals can derive from a symbiotic relationship with AI and its potential as an emergent technology within the radiology arena.
Integration of AI into your PACS and RIS systems
Picture Archive and Communication Systems (PACS) and Radiology Information Systems (RIS) are designed to enhance the role of radiography through the use of technology. As influential in their introduction as AI is today, these systems allow for improved workflows, collaboration, and visualization while significantly reducing waiting times and improving patient care. The integration of AI into PACS and RIS systems has the potential to maximize radiologist capability and supports their role in the discovery of lesions or areas of concern.
“One way to solve the myriad challenges impacting on the time of the radiologist is to embed AI,” says Dr. Paul Parizel, Chair Department of Imaging, University of Antwerp. “It doesn’t replace radiologist, but it does have the ability to take over simple and repetitive tasks that radiologists traditionally have to do. This frees up more time for them to undertake the tasks that really matter such as consulting with colleagues. We have been using Aidoc for the detection of intracranial hemorrhages – the software can detect high-density lesions and annotates them so the radiologist can see the abnormalities. The technology draws the eye, showing the radiologist all points of concern.”
AI is the tool that defines the decisions made by the radiologist, offering up additional data and points. It also minimizes the amount of work that the radiologist is expected to manage on a daily basis. Curious how integrating AI with your PACS and RIS really works? The answer lies in the solution, in ensuring it is platform agnostic and capable of seamless integration within any system, without heavy IT support or downtime.
Aidoc has invested into ensuring that the integration of AI into medical facilities PACS and RIS is seamless – setup and maintenance are primarily handled remotely and there is no need to invest in additional hardware in order to gain the benefits of always-on AI. There is also poetry in the fact that the weight of data currently impacting on radiologist delivery and workload is the exact fuel that drives improved AI delivery and capability. The richer and deeper the integration, the more value the AI can deliver.
What an AI-integrated radiology workflow looks like
Radiology departments using AI, PACS and their RIS together can expect a significantly more powerful and seamless experience than those without the boost of AI. The software is always-on, running in the background constantly to ensure that the data endlessly analyzed and the results relevant. It is consistent and constant, thereby reducing radiologist workload while simultaneously ensuring that high standards are maintained.
As Dr. Chen Hoffman, Head of Neuroradiology Department at Sheba Medical Center points out: “The workload in one day in 2018 is equal to a week in 2008 and a month in 1998, so we need help.”
An AI-integrated radiology workflow ensures that the most urgent cases are flagged and diagnosed first. With Aidoc’s AI-based triage system, cases are prioritized for the radiologist as they enter the system and the data is analyzed. The workflow is then seamlessly managed from scan to diagnosis to ongoing patient care.
The benefits and successes of PACS and RIS integrated AI
There are significant benefits in combining AI with PACS and RIS, beyond what has already been outlined. Of course, seamless workflow and PACS and RIS integration are a start, alongside improved patient care and consistent radiologist support. And, software that has been effortlessly integrated into PACS RIS and workflow delivers a tailored ecosystem experience to the radiologist and department. Diagnosis is faster, more accurate, more consistent, and it is done within tightly defined parameters that prioritize patients according to levels of urgency and need.
Already, global medical institutions are experiencing the benefits of PACS and RIS workflow integrated AI. At the University of Antwerp, Dr Paul Parizel says that AI is ‘important in all elements of radiology and will pave the way forward for radiologists to deal with the amount of work they face. It is the first big thing to happen in the past 20 or 30 years in radiology and our best hope for the future.’ In this video case study, Dr Parizel outlines the impact and application of AI at the university.
At the Sheba Medical Centre, Dr Chen Hoffman, Head of Neuroradiology Department, says that ‘Aidoc is helping us to know who is urgent and who is not while we are doing our teamwork. It helps us to become better doctors. Additional knowledge about the case provides faster treatment and it isn’t about who came first gets treated first, but rather who is more important is treated first.’ The video case study examines how Aidoc’s always on AI has helped the radiology department manage patient care more effectively.
Perfecting the PACS RIS & AI Integration: The full package
The benefit of AI PACS RIS integration is long and varied. They are also dependent on the radiology department, medical institution and ongoing requirements. The data, the volume of work and the pressure of the radiology role are as varied as the patients that rely upon them, but the potential of AI to help transform these pressures and patient care is clear. Seamless AI integration with PACS RIS technology and minimal IT integration complexity will allow for the radiology department to transform potential and redefine collaborative patient care alongside deeper medical capability.