AI holds immense potential to drive productivity and efficiency gains for healthcare organizations. Just look at recent FDA listings; there are over 521 individual AI-enabled medical devices as of October 2022[1].
For healthcare providers, AI medical imaging can increase the accuracy and detectability of diseases, leading to earlier diagnosis and better patient outcomes. For medical personnel, AI can also be useful to help overcome chronic overworking and burnout:
- Nurses and doctors are under immense pressure. The US Surgeon General highlights how 35%-54% of nurses and physicians[2] reported symptoms of burnout such as exhaustion, depersonalization, and anxiety in 2019. This happened before the start of the COVID-19 pandemic, which will have magnified this problem.
- During the COVID-19 pandemic, the number of nurses and physicians reporting burnout increased to 62.8%[3] in 2021, which set a record. For comparison, this figure was only 38.2% in 2020.
- Sadly, these staffing problems will only get worse. Nearly 0.5 million healthcare sector workers were lost between Feb 2020 to Nov 2021[4]. A further 33% of physicians, APP’s, and nurses plan to reduce their working hours according to survey data from July to December 2020 by Mayo Clinic[5]. The final contusion for care providers is that 20% of physicians, and 40% of nurses plan to stop practicing altogether[6].
This is where AI can offer hope, despite the obvious and glaring challenges. Several AI ecosystems are now available for healthcare providers to integrate algorithms and deploy workflow enhancements for clinicians.
In this blog, TeraRecon focuses on the role of AI in medical imaging and its modalities – such as X-ray, CT scans, and MRI. The Eureka Clinical AI platform aims to reduce healthcare burdens for clinicians, and empower healthcare professionals to deliver better quality care while reducing burnout.
Importance of Adjudicating AI Results
It is important to remember that AI is not designed to replace physicians. Instead, AI wants to serve as the radiologist’s companion.
Radiology image adjudication processes lower the risk of variation and image defects. This is performed by ‘readers’ who use quality assurance (QA) to reduce variations in image interpretation.
As clinical decision makers, radiologists are accountable for what ends up in the final patient record. The accuracy of patient records is vital for coordinating integrated care. Therefore, it makes sense to put the decision back in the hands of radiologists through a joint image adjudication partnership between computers and clinicians.
AI Center of Excellence Starts with AI Consolidation
Healthcare providers should consider creating an Artificial Intelligence Center of Excellence (CoE).
As an example, the UK government aims to be a world leader in medical AI and machine learning by 2025 . As of 2019, five AI CoE’s exist to assist the UK NHS in medical imaging and digital pathology. Furthermore, outside of the healthcare industry, 37% of large companies in the US already have an AI CoE.
The AI CoE consists of technical experts that know how to advise, guide, and oversee AI projects within a clinical environment. Below are several tasks and benefits to establishing an AI CoE for healthcare providers:
- Outline organizational problems, and identify ways in which AI can solve those problems
- Determine AI maturity level, and what is needed to level-up AI maturity
- Plan and develop IT infrastructure to support the use of medical imaging and AI algorithms
- Build implementation roadmaps for AI radiography projects; ensure the best tools and technologies are used, and measurable KPIs are established.
- Coordinate and orchestrate AI development and adoption across clinical sites and departments
- Standardize AI practices and processes in pursuit of scaling up and out
Benefits of AI Consolidation via the Eureka Clinical AI platform
Consolidation is about taking multiple disconnected solutions and bringing them together in one enterprise-wide ecosystem.
TeraRecon excels at aggregating multiple AI algorithms and visualization tools and making them easily accessible through a single pane of glass. As our number of AI partners grows, we hope to become your go to partner for all clinical and operational AI needs.
How do we achieve this via the Eureka Clinical AI platform?
- Full support and interoperability with PACS, VNA, RIS, EMR, DICOM, HL7 reporting systems and frameworks
- Procure multiple AI solutions through a single, trusted vendor
- Combine multiple customer-developed, TeraRecon-developed, and third-party AI algorithms and orchestrate them via a single platform
- Reduce the complexity of onboarding and maintaining multiple AI solutions
- Readily accessible tools for physicians across desktop and mobile
- Standardize the underlying technology infrastructure foundation to streamline long-term maintenance and scalability
- Monitor usage, performance, and productivity metrics for each AI algorithm in a single dashboard
- Universal viewer for all algorithms to reduce learning curve and streamline AI adoption among physicians
- A curated library of proven, workflow-ready clinical AI algorithms with extensibility for future use cases
- All of the above is included in one simple subscription to the Eureka Clinical AI solution
Enabling Self-Made, Home-Grown AI Medical Imaging Solutions
We urgently need more AI developers to create medical AI algorithms and machine learning models.
TeraRecon is making strides in its goal to empower AI algorithm developers through the Eureka Clinical AI platform:
- The Intuition AI Data Extractor transforms advanced visualization data into a format that is readily accessible for training AI
- Intuition and Eureka AI integration enables in-house AI development capabilities for academic medical centers. The end result is new AI algorithms being productized and operationalized for use in clinical environments.
- TR Analytics can track and monitor AI algorithm performance, and other metrics during development to achieve validation for clinical utility
- Ingest AI results into the market leading Intuition advanced visualization software to render photorealistic presentations of 3D models and structures with Glow Rendering and Neuro Perfusion Maps
- Choose your preferred deployment configuration with on-prem single and multi-node
Integrating Workflows with PACS and Reporting Systems
The Eureka Clinical AI platform is vendor agnostic and all ingested data is standardized; this is what drives interoperability on the platform.
This enables the integration of clinician workflows directly with PACS, RIS, and EMR systems, and sharing of AI results via DICOM, HL7 and GSPS communication standards. Furthermore, physicians can interpret AI results and explore before committing them to the enterprise system.
Previously, this would involve uploading images to a vendor neutral archive (VNA) system to standardize the format, which is now regarded as insufficient[1] for modern healthcare needs.
Coordinating Integrated Patient Care Among Multi-Specialty, Multi-Disciplinary Doctors
Consider this example involving a stroke victim[1]. The timeline demonstrates just how many teams are involved in coordinating patient care:
- Paramedic to take the patient from home to hospital
- Remote diagnostic machine in ambulance relaying paramedic records to neurologist in real-time
- Joint assessment between paramedic and neurologist, patient is sent to emergency department
- Triage and communication with specialist stroke experts in preparation for patient arrival
- Radiologist performs CT scans and angiography and sends to surgical team for minimally invasive mechanical thrombectomy procedure
In total, that was 6 specialist teams for a single patient. The ability of these teams to collaborate is vital in protecting the lives of patients.
TeraRecon strives to help orchestrate multi-specialty teams of doctors by removing digital barriers to collaboration. Some examples of this include:
- Full access to patient lists on a mobile device
- Integrated tools for clinical measurement, treatment planning, and diagnosis
- Automated large vessel occlusion (LVO) and intracerebral brain hemorrhage (ICH) detection in images, with notifications sent to physicians in real-time
- View, accept, reject, or modify image findings via desktop; control the final patient record
- View CT and MR slices with phase scrolling using touch controls on mobile
- HIPAA compliant instant messaging, group chats, and doctor-to-patient communications
- Full access to medical insights on any device through to the point-of-care
Most importantly, clinicians retain full control over what ends up in the patient record thanks to TeraRecon’s patented confirmation, rejection, and adjustment tools.
AI is the Long-Term Partner Healthcare Providers Need to Transform Medical Imaging
The Eureka Clinical AI platform is your consolidated foundation to host multiple AI medical imaging algorithms:
- End-to-End AI Strategy - We empower AI developers to develop, run, and execute all types of AI
- On-Demand AI Orchestration - Healthcare organizations can access pre-selected clinical AI algorithms and care area suites that deploy quickly and are easy to orchestrate.
- Integration and Interoperability - Share AI results with your preferred reporting system via DICOM and HL7 communication standards.
- Intuitive Integrated Care - Dual support for desktop and mobile devices, in-app messaging and custom alerts for medical personnel, sharing AI results to optimize care delivery.
- Physician-Centered AI - Physicians can step in at any stage to accept, reject, or adjust AI findings before they are committed to EMR or PACS.
- Actionable Insights, Confidence in AI - Simultaneously track AI algorithm performance, productivity, and utilization. Boost confidence in AI with improved visibility and monitoring.
We invite you to contact us for a free trial, and test drive the AI-powered future of clinical practice.
Alternatively, reach out to one of our local teams in North Carolina, California, Massachusetts, Japan or Germany for expert advice and guidance.
- https://www.fda.gov/medical-devices/software-medical-device-samd/artificial-intelligence-and-machine-learning-aiml-enabled-medical-devices
- https://www.hhs.gov/sites/default/files/health-worker-wellbeing-advisory.pdf
- https://www.ama-assn.org/about/leadership/reducing-physician-burnout-must-be-urgent-national-priority
- https://ritms.rutgers.edu/news/why-health-care-workers-are-quitting-in-droves/
- https://www.mcpiqojournal.org/article/S2542-4548(21)00126-0/fulltext
- https://www.mcpiqojournal.org/article/S2542-4548(21)00126-0/fulltext
- https://www.longtermplan.nhs.uk/nhs-aims-to-be-a-world-leader-in-artificial-intelligence-and-machine-learning-within-5-years/
- https://www.gov.uk/government/news/artificial-intelligence-to-help-save-lives-at-five-new-technology-centres
- https://hbr.org/2019/01/how-to-set-up-an-ai-center-of-excellence
- https://www.bridgeheadsoftware.com/2015/06/its-time-to-retire-the-vna/
- https://www.philips.com/a-w/about/news/archive/standard/news/articles/2022/20221025-how-philips-is-enabling-the-future-of-stroke-care.html
No Comments Yet
Let us know what you think