Combating the Radiologist Shortage with Artificial Intelligence.

Medical imaging is crucial to global healthcare. It is the backbone that allows us to diagnose acute and chronic medical conditions directly, affecting patient morbidity and mortality.

However, there is a global shortage of radiologists and that’s made it difficult to meet the medical needs of society. The shortage has led to delayed diagnoses and even undetected medical conditions that have postponed appropriate medical treatment and impacted patients’ survival.

To meet the high volume of diagnostic imaging studies required, radiologists work intensely for long hours. This makes them more prone to human error and is a main cause of a high radiologist burn-out rate too.

Now innovation in medical imaging technology and AI is combating this crisis, with industry giants and disruptive start-ups contributing to the fight. I spoke to experts at Qure.ai, Zebra Medical Vision, GE Healthcare and Kheiron Medical to find out more.

Chiranjiv Singh, Chief Commercial Officer at Qure.ai, told me how their AI technology is used to work alongside radiologists to help identify conditions and prioritize workloads.

It’s common practice for screenings to be looked at by two radiologists, which takes time. With AI, you can free-up one of these radiologists (who are in short supply) to see patients, look at other scans, conduct scans and do biopsies.

That’s exactly the advantages that Qure.ai are offering with their qXR - which can detect a total of 29 findings on a chest x-ray - and their qER which identifies critical abnormalities in CT scans.

 These have been effective and lifesaving tools used in the emergency room setting.

 “Some hospitals don’t have the hardware or a trained neuroradiology specialist - especially not 24-seven. We provide the ability to detect bleeds and critical findings within minutes, rather than the hours taken for someone to read that same exam. This can make a serious difference to the patient outcome.

 The qER, which is FDA 510(k), also features a cleared triage tool that prioritizes head CT scans with critical abnormalities such as a bleed, fracture, mass effect or midline shift on the radiology worklist for priority review.

 Multiple published studies have found the Qure.ai technology to be equally or more accurate than radiologists in diagnosis. Radiologists can now lean on AI to help them prioritise which examinations they read, increase their output and diagnose accurately.

 Now partnered with Nanox, an innovative medical imaging company, Qure.ai’s AI-powered diagnostics can be enjoyed by a larger network of radiologists and clinicians – combating the radiologist shortage on a larger scale.

With similar advantages, GE Healthcare have released the first AI tool to quickly identify and help prioritize critical cases like Pneumothorax (a type of collapsed lung).

Their Critical Care Suite automatically inferences frontal chest images upon acquisition in the x-ray system for the presence of quality errors and critical findings. The x-ray AI notification for a suspicious critical finding arrives in PACS at the same time as the original DICOM images which enables worklist prioritization and may lead to potentially shorter Radiologist turn-around times for these prioritised cases.

This technology has been proven to detect 96% of large pneumothoraxes and 75% of small, difficult pneumothoraxes.

While this increased accuracy is a benefit, X-ray Product Manager, Alec Baenen, focused on the value that its efficient workflow provides, helping radiologists prioritise critical cases faster.

“Our AI technology helps enable prioritization of critical findings so that the radiologist faces with a reading queue of dozens of images can immediately assess the most critical cases . This can help ensures that potentially critical or life-threatening conditions, like Pneumothorax, are not left un-resolved or untreated and potentially save somebody's life.”

Kheiron Medical are applying AI to help diagnose breast cancer quicker and more accurately with their Mammography Intelligent Assessment (Mia). This is something that their unique team of cancer survivors, surgeons and radiologists are passionate about.

Mia is a decision-making algorithm which tells radiologists whether to recall a woman for breast screening. According to a spokesperson at Kheiron Medical, this is something that many radiologists identify as the most important decision in a screening service.

Kheiron Medical are set to run clinical trials with the East Midlands Radiology Consortium (EMRAD), which a spokesperson at Kheiron claimed to be

“one of the most ambitious and biggest retrospective clinical studies ever done on an AI.”

EMRAD, an established collaboration of seven NHS Acute Trusts in the East Midlands, is the largest collaborative network of hospitals in the UK to share a singular image sharing platform, supported by GE.

The Mia technology is known to work at the same level of accuracy, or better than, what’s expected of a breast screening radiologist. The technology has also received awards like Best New Radiology Software at the 2019 AuntMinnie Europe Awards and Best AI Product In Health at the CogX 2019 Innovation Awards.

Having also recently been awarded NHSX funding, Kheiron Medical want their technology to impact the world as they continue to test Mia on integration with multiple technologies and different levels of breast densities.

Unlike the other companies I’ve mentioned which focus on specific areas of imaging, Zebra Medical Vision apply their AI solutions on a wider scale to bone health, oncology as well as to acute medical conditions.

I spoke to Dr. Orit Wimpfheimer, Clinical Director at Zebra Medical Vision, about their analytics platform which allows healthcare institutions to identify patients at risk of disease and offer improved, preventative treatment pathways to improve patient care.

 Their Analytics Engine receives imaging scans from various modalities and automatically analyzes them for multiple clinical findings, in full synergy with radiology workflow. Dr. Wimpfheimer believes that the ability to offer all-in-one solution is what makes Zebra Medical Vision unique.

“While many AI companies create solutions for very specific domains, it is unwieldy for a radiology department to deal with so many AI vendors simultaneously. As such, we hope to be the one company radiologists turn to for all of their AI needs.”

Their analytics platform can identify patients at high risk of cardiovascular, lung, bone and other diseases. This facilitates preventative care programs, proper risk adjustments and allocation, as well as conformance with quality standards.

This means that focus of the clinician can be provided to the right patients, at the right time - overall saving cost and improving patient care.

There are plenty of advantages to AI in medical imaging, however, like in other industries, there has been media suggestions that adoption of this machine-led way of working will take away jobs from radiologists.

Myself and every expert that I spoke with for this article couldn’t disagree more. There are not enough radiologists in the world anyway, so what jobs are there to be taken? AI should not be seen as a threat, but as a tool that can be used to support radiologists, helping them work more accurately and efficiently. Chiranjiv of Qure.ai said it best:

 “We’re decades away from having an excess supply of radiologists with AI taking over their jobs. AI is a tool to help radiologists improve care for the patient. After all, that is what’s most important to everyone in healthcare - how are we improving outcomes for our patients?”

 This isn’t to say that AI doesn’t have hurdles to overcome before it achieves widespread adoption in medical imaging. I could probably write a whole new article on the challenges AI faces in accessing, collecting and reviewing data – but that’s for another day.

What is clear, is that these innovations are already making the lives of our radiologists easier and empowering better patient outcomes. As wider adoption of AI spreads throughout medical imaging, the shortage of radiologists will become less of an issue and patients will reap the rewards.

I would love to hear your thoughts on the article or, hear about any another AI company that is exciting you in the MedTech space!

Please feel free to leave a comment below, send me a LinkedIn message or, contact me at [email protected]