Cataract surgery goes digital

More than half of the 1,500 delegates interviewed in the latest ESCRS Clinical Survey believe a more integrated digital operating room has the potential to improve efficiency and workflow. What remains to be seen is just how - and by how much - automating the OR and the entire surgical process might improve quality and efficiency.

Currently available digital integration platforms touch every step of the surgical process, from patient registration and diagnosis to surgical planning to follow-up. Data is automatically captured and forwarded to populate devices at the next step, increasing efficiency and reducing the chance of transcription errors - such as automated patient registration data updates to ensure record continuity.

Biometry and other data flow directly from measuring devices to decision support portals, automatically populating intraocular lens (IOL) power calculation formulas. Images and other data, such as corneal astigmatism and asphericity, are available at a keystroke for treatment and surgical planning, such as fine-tuning IOLs to meet specific patient anatomic and refractive needs.

The operating room, too, has seen more automation. Often checked manually, lenses are also automatically scanned for accurate documentation and to ensure an exact match to the surgical plans. Information - such as the axis of orientation for toric Lenses - can be directly loaded into microscopes that project it onto the surgical image in real time.

Follow-up exam data supports quality improvement after surgery, such as customising personal A-constants to improve the refractive accuracy of future lens implants.

AI’s three domains

Artificial intelligence (AI) has rapidly incorporated into ophthalmology in three major domains—big data, imaging, and roboticization, said Professor Béatrice Cochener-Lamard in her Ridley Medal Lecture at the 40th Congress of the ESCRS. She has researched AI solutions in ophthalmology with partners for 20 years at LaTIM (Laboratory of Analysis and Treatment of Medical Imaging), Inserm unit 1101 (the French National Institute of Health and Medical Research), and OPHTAI. They developed patented algorithms on photos (retinophotography, OCT), videos (cataract, simulator, tear film), or multimodal approaches for automated recognition or grading of eye diseases.

“The targets are detection, follow-up, prognosis, and gain in medical efficiency,” she noted. With several diagnostic and screening devices already approved by regulators around the world, AI has radically changed ophthalmic imaging, setting up to improve workflows as well.

One example is the PEARL-DGS Calculator, which applies both AI and linear algorithms to predict postoperative spherical equivalents for IOLs, Prof Cochener-Lamard said. Another is WeOptimeye, which uses AI at every step of the cataract workflow - including managing surgical scheduling and tailoring IOL power calculations and lens choices to specific patient needs.

At Carl Zeiss Meditech, AI supports a training module that monitors and helps analyse every step of cataract surgery by comparing it with a reference video, said Dr Euan S Thomson, president of ophthalmic devices, head of the digital business unit, and the company’s CEO. Future plans for the Zeiss information platform include a database able to support AI research. Google’s Sunny Virmani MS noted improving data quality is also critical, adding Google and other organisations are building expertise in collecting and standardising all kinds of imaging data from both the anterior and posterior segments.

However, Dr Thompson shared that one barrier to using AI applications in clinical practice is regulators will only approve them for use with data collected on the device that provided the training data set used to develop the programme – which makes it difficult to develop applications capable of using data from any source.

Prof Cochener-Lamard also identified bioethics, data protection, and distrust by clinicians who see AI as a threat to their profession. She suggested addressing this scepticism by encouraging them to think of AI as a way to help answer questions while remembering they must formulate those questions and supply answers.

What of the cost?

Cost is another challenge to implementing the digital OR. Two-thirds of ESCRS delegates surveyed see cost as a barrier, with one-quarter reporting no access to such technology. But it won’t stop the digital OR from becoming a reality, said Dr Eric D Donnenfeld.

“I think it is inevitable we will all be using digital ORs in the future, just like we are now doing digital medical records,” Dr Donnenfeld said. “First of all, it’ll be mandated by the government, and secondly, the economies of scale will make it better, and data entry will be seamless.

“When this happens, [the digital OR] will be our greatest ally for improved outcomes and fewer mistakes, so it will pay for itself. Cost is an issue today, but in the long run, this is actually going to pay for itself many times over.”

The ESCRS will have a major symposium on digitizing the operating room at the 41st Congress in Vienna in 2023.

Béatrice Cochener-Lamard MD, PhD is Professor and Head of the Ophthalmology Department in Brest University Hospital, France, member of LaTIM (Inserm 1101 – Laboratory of research in analysis and treatment of medical imaging), and dean of faculty of medicine and medical sciences at the UBO University. beatrice.cochener-lamard@chu-brest.fr

Euan S Thomson PhD is president of ophthalmic devices, head of the digital business unit, and CEO of Carl Zeiss Meditec Inc.

Sunny Virmani MS is group product manager in ophthalmology and dermatology at Google. svirmani@google.com

Eric D Donnenfeld MD is clinical professor of ophthalmology at New York University in New York City, US; trustee of Dartmouth Medical School in Hanover, New Hampshire, US; and in private practice in New York State. ericdonnenfeld@gmail.com