COMPUTER VISION

While human beings have always strived to advance as a species, today’s technological innovations allow us to make unprecedented leaps—especially in the field of medicine.

The Italian expression “tali sono gli occhi, tale è il corpo” roughly translates into “such are the eyes, such is the body.” But it was in ancient Greece where Hippocrates conducted early research on the human eye, as he believed that it offered a window into health and well-being. To this day, scientists and physicians continue to refer to the Corpus Hippocraticum.

That same desire for discovery and progress has given way to digital health technologies that have led to advancements like AI medical diagnosis solutions, which have dramatically improved our quality of life. Combining ancient wisdom with modern innovations, Jan Hlaváček—an engineer who has been passionate about medical software since he was 16 years old—understands how technology can be harnessed to improve life on a global scale. In fact, Hlaváček, as COO, leads Aireen, a provider of AI-based screening medical devices, based in the Czech Republic.

“It will be a big challenge to ensure the necessary level of healthcare,” says Hlaváček. “As there is more and more work for healthcare professionals, there is an increasing need for more medical expertise. Healthcare should innovate itself. It’s very important.”

Developed hand in hand with the ophthalmology experts at Military University Hospital in Prague, Aireen’s solution aims to facilitate diagnosis of diabetic retinopathy, age-related macular degeneration, glaucoma, and Alzheimer’s disease.

“As there is more and more work for #healthcare professionals, there is an increasing need for more #medical expertise. Healthcare should innovate itself. It’s very important” – Jan Hlaváček, Aireen via @insightdottech

AI and Computer Vision in Healthcare Applications

From a patient’s first interaction with the healthcare system—whether with a family physician or an optometrist—this platform serves as an invaluable diagnostic resource.

Aireen’s solution, trained with more than 1.5 million fundus camera images—which capture the back of the eye—can provide over 99% sensitivity when analyzing a new retina image. This level of accuracy represents a substantial and compelling contribution to the diagnosis. Since it’s being generated with mathematical rigor, it is free from the emotional conditioning typical of the process led by human beings.

Healthcare providers need only connect a fundus camera to the Aireen solution, which is available on the cloud or easily installed on-premises, to leverage the power of AI in their diagnoses.

The software respects the privacy of sensitive data, with encryption practices that ensure the inviolability of the data to a reasonable extent. And this is without the need to install other devices, apart from the fundus camera and a standard computer—or installed on-premises in larger medical facilities.

In the case of cloud services, you can upload the image coming from the camera using a simple, dedicated web interface and press the button to start the analysis. In a short time, the system generates a report that can be downloaded by the practitioner. That’s all.

Optimizing AI-Based Diagnostic Solutions

The solution is built upon artificial intelligence, computer vision, and Intel^® deep-learning technology, which enables users to analyze the fundus noninvasively, and painlessly screen patients based on digital scans of the retina. The image is then processed by proprietary algorithms that leverage the Intel^® OpenVINO^™ toolkit.

OpenVINO’s ability to optimize AI models for high-speed, efficient performance on Intel^® hardware brings a new dimension to healthcare applications. It enables solutions like Aireen’s to offer diagnostic accuracy in a matter of minutes, demonstrating the benefits of AI applications in medical diagnostics and healthcare.

“This is just the beginning of a transformative journey where AI, enhanced by OpenVINO, paves the way for a future where healthcare is more accessible, personalized, and effective,” says Anisha Udayakumar, AI Evangelist at Intel. “The synergy of AI and Intel technology in healthcare is not just about innovation; it’s about creating a healthier, more hopeful world for all.”

Furthermore, Intel software is helping play a critical role in other AI-based medical diagnostic solutions. Segmentation is essential for isolating specific areas of interest in medical images, which is a key step in many diagnostic processes. By further optimizing these models with techniques like quantization, OpenVINO significantly improves model efficiency and performance. This is highly relevant in healthcare, as it demonstrates the application of AI in processing complex medical images efficiently, aligning with the broader theme of using AI models to enhance diagnostic accuracy and speed in healthcare, Anisha Udayakumar explains.

Enhancing Medical Diagnostics

The solution also leverages the DICOM standard, which guarantees interoperability between the local device and the solution. DICOM—Digital Imaging and Communications in Medicine—is the international standard for medical images and related information. It defines the formats for medical images that can be exchanged with the data and quality necessary for clinical use. At the end of the process, the system deletes the provided image to respect the privacy of sensitive data.

“The DICOM protocol is implemented directly in fundus scans,” says Hlaváček. “It allows us to automatically receive data from the scans to our app. We can also send a report to the application, which subscribes to it. Plus you can create a request for examination, with a unique ID for all data, which is produced during the workflow.”

To be clear, the goal is not to serve as a substitute or a replacement for medical expertise but rather to complement it. Aireen’s solution enhances practitioners’ diagnostic skills and frees their time to come up with treatment plans, respond to psychological challenges, and engage authentically and empathetically with their patients.

Digital Health Transformation: The Time Is Now

According to the latest research issued by the World Health Organization and other institutions, more than 500,000,000 people have been diagnosed with diabetes worldwide, and this alarming number is expected to grow by 50% over the next 15 years. Underscoring the significance of Hlaváček’s innovation, diabetic retinopathy and age-related macular degeneration are the most frequent cause of blindness among European descendants.

Bearing that in mind, it is not only timely but also urgent to facilitate diagnosis on a broad scale. Aireen’s solution, together with Intel technology, holds tremendous promise. As a CE certified medical device, it is already in daily use among Czech Republic and Slovakia ophthalmology clinic customers.

“Seeing the fusion of AI with healthcare, particularly in innovative solutions like Aireen’s AI-based Retina Eye Diagnostic tool, is truly a source of excitement for me,” says Anisha Udayakumar. “Aireen’s application of AI for more accurate and quicker diagnostics not only demonstrates the immense potential of this technology but also highlights the revolutionary role of Intel’s OpenVINO toolkit in healthcare technology.”

And as the Aireen platform provides convenient access to early diagnosis, leading to treatment can help address the urgent shortage of ophthalmologists across the world.

If Hippocrates were alive today, he would, no doubt, be entranced by our medical advances and modern technologies, and would be eager to build upon the Corpus Hippocraticum for future generations.
 

This article was originally published on February 18, 2022.

This article was edited by Leila Escandar, Editorial Strategist for insight.tech.