Non-invasive 'virtual' biopsy for skin cancer diagnosis

Researchers have developed a new imaging method that may enable a dermatologist to scan the skin with a non-invasive ‘virtual’ biopsy to determine if it contains any cancerous cells.

Similarly, the method may allow surgeons trying to determine whether they have removed all of a breast tumour to eventually rely on an image captured during surgery rather than wait for a pathologist to process the excised tissue.

Developed by Stanford Medicine researchers, the method uses lasers to penetrate tissue and create a high-resolution, three-dimensional reconstruction of the cells it contains.

From this virtual reconstruction, they can make cross-sectional images that mimic those generated by a standard biopsy, in which a sample of tissue is sliced into thin layers and placed on a slide to be examined under a microscope.

The new method could generate rapid results on biopsies taken elsewhere in the body, and also provide more information than current diagnostic approaches.

“We’ve not only created something that can replace the current gold-standard pathology slides for diagnosing many conditions, but we actually improved the resolution of these scans so much that we start to pick up information that would be extremely hard to see otherwise,” said Adam de la Zerda, PhD, an associate professor of structural biology and the senior author of the article describing the method.

Transforming diagnosis

The method was developed by Yonatan Winetraub, PhD, a former graduate student in the de la Zerda lab who now leads his own research lab at Stanford focusing in part on virtual biopsies.

“This has the potential to transform how we diagnose and monitor concerning skin lesions and diseases in the clinic,” said co-author Kavita Sarin, MD, PhD, an associate professor of dermatology.

For nearly a decade, de la Zerda and his colleagues have been studying a different way of seeing inside the body, called optical coherence tomography (OCT). Typically used by ophthalmologists to image the back of the eye, OCT scans measure how lightwaves from a laser bounce off a tissue to create a rendering of its insides (similar to the way ultrasound uses soundwaves to visualise organs).

As de la Zerda and Winetraub enhanced the OCT scans so they would work in organs other than the eye — developing both new hardware to collect data and new processing methods — they needed a way to verify the accuracy of their scans, so they sent the tissues they were scanning with OCT to pathologists to create H&E images.

“We kept improving and improving the quality of the image, letting us see smaller and smaller details of a tissue,” de la Zerda said. “And we realised the OCT images we were creating were really getting very similar to the H&Es in terms of what they could show.”

Artificial intelligence

The higher resolution of the OCT images opened the door to using the method to diagnose disease without producing H&Es. But de la Zerda and his colleagues thought clinicians would be more apt to use OCT if the images looked familiar.

“Every physician in a hospital is very much used to reading H&Es, and it was important to us that we translate OCT images into something that physicians were already comfortable with — rather than an entirely new type of image,” de la Zerda said.

Winetraub turned to artificial intelligence to help convert OCT scans into flat images resembling H&E slides.

For 199 skin biopsies collected at Stanford Hospital, Winetraub carried out an OCT scan before pathologists created H&E slices. He and his colleagues developed a way of putting molecular tags on the surface of the biopsies so they could be sure exactly where in the OCT scan each H&E slice came from. Then, Winetraub paired up 1005 of these H&E images with the corresponding OCT images and entered them into an artificial intelligence algorithm which could learn how to create accurate H&Es from the raw OCT data.

“The uniqueness of this work lies in the method we developed to align OCT and H&E image pairs, letting machine-learning algorithms train on real tissue sections and providing clinicians with more accurate virtual biopsies,” Winetraub said.

The researchers fine-tuned the AI program by showing it an additional 553 pairs of H&E and OCT images before testing it out on new OCT images. When three Stanford dermatologists analysed random assortments of true H&E images and those created from the OCT scans, they could detect cellular structures at a similar rate. Any number of H&E images can be created from a single OCT image, virtually slicing the three-dimensional reconstruction in any direction.

Toward non-invasive biopsies

When a dermatologist notices an unusual looking spot on a person’s skin, they currently have two options to determine if it poses a risk: wait and see whether it grows bigger or cut it off and send it to a pathologist for testing.

De la Zerda and Winetraub now see a third path — scanning a potentially cancerous mole with OCT and analysing the virtual H&E images.

“Imagine if we could give physicians the ability, right there in the room with the patient, to take out an OCT camera and — rather than slice the patient up in dozens of places — image the cells inside each mole,” de la Zerda said.

Similarly, surgeons removing breast tumours currently send removed tissue to pathologists to process over several days and determine whether any cancerous cells were missed. Around 20% of breast cancer patients require a second surgery to remove more cells. If H&E images could be produced from an OCT camera in the operating room to instantaneously detect whether cancer cells remained, subsequent surgeries could be avoided.

More work is needed to move the approach toward these applications, but the researchers are confident that their approach will give clinicians a new way to carry out biopsies.

The study has been published in Science Advances.

Image credit: iStock.com/ipopba