In many fields of medicine, complex devices such as X-ray machines or magnetic resonance imaging (MRI) scanners have been developed to give clinicians additional images of the patient and their ailment. Often, these images are critical for an accurate diagnosis, and therefore selection of the correct treatment, especially when they reveal important details hidden below the surface of the body, such as the presence of a tumour.

Dermatology, the study of pathologies of the skin, is no exception. A visit to the clinical-dermatology shelves of an academic bookshop or library shows that dermatologists are experts at visual recognition: most of the books are thick tomes full of highly detailed colour photographs of skin lesions, and the pink and purple images of pathology slides from skin biopsies. Dermatologists are superbly skilled at recognizing any one of the hundreds of different forms of lesions resulting from skin diseases.

Yet dermatology is unusual among medical professions in that, until very recently, no imaging technology was available that could readily reveal details of sub-surface skin tissue at a clinically useful resolution. Ultrasound machines have insufficient resolution, and optical microscopes cannot see deep enough. As a result, dermatologists have no option but to make a visual assessment of the lesion on the skin surface. If unsure of their diagnosis, they will usually take a biopsy – physically remove a piece of skin – which is sent away to a pathology lab for analysis. Unfortunately, this is not pleasant for the patient, often leaves a permanent scar and can take days or weeks to produce a result.

Now, an emerging imaging technology based on laser scanning, known as optical coherence tomography or OCT, is giving the dermatologist a new and powerful way to “see beneath the surface”.

OCT is not entirely new: it was originally invented and developed by James G Fujimoto and co-workers at Massachusetts Institute of Technology in 1991 (Lasers Surg. Med. 11 419) and has since been successfully applied to imaging the layers of the retina in the eye. It is safe to say that OCT has revolutionized ophthalmology – there are thousands of OCT retina scanners in use today, taking millions of images of the layers of the retina every year. However, the race is now on to apply OCT to other medical applications.

OCT can potentially be used wherever disease resides in the surface layers of tissue that an optical probe can reach, for example cancer of the gastrointestinal tract and oral cancers. Another application is cardiovascular disease, where OCT probes inside arteries are used to inspect fatty plaques adhering to the inner walls of the vessels.

Numerous companies have sprung up, conceived and launched by enterprising physicists, engineers and entrepreneurs, with the common goal of developing OCT probes for these diverse and exciting new imaging applications. One prime example is the UK company Michelson Diagnostics Ltd, founded in 2006 by the author and four colleagues, and which is highly active in applying OCT to the field of dermatology.

OCT utilizes the phenomenon of the coherence of laser light to extract fine detail from within highly scattering turbid media. Inside every OCT machine is an optical interferometer – as used to great effect by the Nobel-prize-winning physicist Alfred Michelson in 1887 for his investigations into the fundamental speed of light, the baffling results of which were pivotal in driving Einstein to develop his theory of relativity. The interferometer is indeed a triumph of modern physics with many applications, and Michelson Diagnostics was named in its honour.

Accurate measurements

A typical, simple OCT device is shown schematically in figure 1. A laser beam is split into two by the beamsplitter, with one part directed towards a reference mirror and the other into the biological tissue. The reflected light from each path is recombined and, if coherence is maintained and the optical pathlengths are close, interferes at the detector. Translating the reference mirror in a direction perpendicular to the tissue produces interference fringes at the detector, enabling the magnitude of the coherently reflected light from the tissue to be accurately measured.

Crucially, any photons that are multiply scattered on their way out of the tissue and reach the detector will lose their coherence, and thus will not produce interference fringes at the detector. While these multiply scattered photons outnumber the reflected coherent photons by many orders of magnitude, they are of no value for imaging because they have lost both directional and intensity information from the feature producing the reflection. By detecting only coherent photons in the interferometer, OCT in essence “filters out” the noise from the unwanted multiply scattered photons, allowing details to be seen – almost like seeing through fog.

In order to build up an image of the subsurface detail, the OCT sensing location within the tissue is scanned, laterally by translating the beam position across the tissue surface and vertically by adjusting the reference mirror position to probe different depths.

A biological fit

The characteristics of OCT make it ideal for imaging surface biological tissue. Figure 2 shows how OCT fits in the landscape of imaging technologies. It has greater resolution than ultrasound or MRI, but far lower penetration into the tissue. On the other hand, OCT has lower resolution than optical microscopy, but it can “see” considerably deeper.

OCT is extraordinarily sensitive: a signal-to-noise ratio of 100 dB or more is typical for commercial instruments, but ultimately the penetration depth is limited. The OCT signal drops off exponentially with depth, so OCT images are conventionally shown using a grey scale based on the logarithm of OCT signal strength, so that a uniformly scattering medium appears in the image as a region with the grey level decreasing linearly with depth. It follows that even very large increases in the incident light power produce only minor improvements in the imaging depth. For all practical purposes, this limits the imaging depth in most biological tissues to between 1 and 2 mm.

This limitation in penetration depth is a penalty worth taking, however, in exchange for the impressive imaging resolution available from OCT compared with more deeply penetrating modalities such as ultrasound or MRI. The lateral resolution is simply a property of the numerical aperture of the imaging optics. The axial resolution is controlled by the properties of the light source and interferometer, such as the laser coherence length. Most OCT systems are capable of axial and lateral resolutions of the order of 5–20 µm, which is sufficiently high to reveal clinically important details of the tissue microstructure in surface layers (epithelia), although not of individual cells. For comparison, the resolution of ultrasound is typically 100–200 µm, while that of MRI is typically 1000 µm.

Figure 3 shows an OCT image of a skin cancer “nest” approximately 2 mm in diameter. It can be readily seen that OCT is capable of revealing the fine detail of skin-tissue microstructure, including the principal skin layers such as the dermis and epidermis, skin appendages such as blood vessels and hair follicles, and tiny skin cancers and their borders. The imaging depth is greater than 1 mm and the lateral field of view is 6 mm. Such characteristics make OCT of intense interest to dermatologists.

But for any new medical-device technology to be successfully adopted by clinicians, its performance must be good enough to deliver clear clinical benefits to the patient. It goes without saying that it must be affordable, and above all it must be easy to use and to integrate into the clinical pathway. Early OCT systems that were developed for skin-imaging applications were none of these things: they were large, cumbersome machines that were too difficult to use and too slow for use in a busy dermatology clinic (a typical dermatologist sees a new patient every 10 minutes).

To improve this situation, Michelson Diagnostics has developed a commercial OCT scanner called VivoSight that has a light, ergonomic probe that is easy to manipulate, and which provides high- quality images at video frame rates. As such, a suspicious skin lesion can be scanned in a matter of seconds by a doctor or physician’s assistant with the minimum of training. The company’s patented technology utilizes an array of laser beams, each focused at a different depth, to produce a blended mosaic OCT image with higher resolution than is possible with a single laser beam. Four beams give a doubling in resolution compared with that achieved using a single beam, and the system’s lateral resolution of 7.5 µm has been verified by measurements at the UK’s National Physical Laboratory (Proc. SPIE 6847 68472Q).

The image in figure 3 provides an example of the remarkable detail that can be achieved by the VivoSight OCT scanner. This breakthrough in image quality, coupled with an easy-to-use probe design, has excited a lot of interest among leading dermatologists, who are now applying it to many pressing clinical challenges.

Dermatology applications

Rapid progress has been made since development of the VivoSight OCT scanner was completed in 2010. The system received a CE mark, allowing its use in Europe. Then in early 2011, FDA clearance for medical use in the USA was obtained, enabling clinicians to scan patients in clinical trials and then in their clinical practices for an ever-increasing variety of applications.

It has been tremendously exciting to witness and be part of this explosion of interest. In just three years, more than 90 clinical papers on OCT in dermatology have been published and currently, new papers are appearing at the rate of one or two per month. In late 2011 dermatologists in Germany started to routinely scan their patients with VivoSight, and by mid-2013 more than 2000 patients had been scanned in 10 German clinics. As such, Michelson Diagnostics has established an office in southern Germany and is expanding commercial operations in Switzerland and Australia.

So what are dermatologists using the OCT scanners for? The main application today is diagnosis and guiding treatment of non-melanoma skin cancer (NMSC) and the premalignant condition known as actinic keratosis. This common disease affecting mainly white elderly people is not life-threatening, but may progress to the skin cancer squamous cell carcinoma and should therefore be diagnosed as early as possible. The earlier the treatment, the less invasive and traumatic the cure. Our modern addiction to holidays in the sun has resulted in a surge of people suffering from NMSC and actinic keratosis. According to published sources, as many as 58 million Americans suffer from actinic keratosis, with at least 3.5 million new cases of NMSC in the US annually, a number that continues to rise every year.

The standard pathway for suspected NMSC is to take a biopsy and send the sample to a laboratory for analysis. If it is positive, the tumour is excised along with a 5 mm clear border around the visibly apparent lesion (in an attempt to ensure that it is all removed), leaving the patient with a permanent scar. Less invasive treatments that leave no scar are available, using drugs or photodynamic therapy, but these are not often used because surgery has a more certain outcome.

Now OCT scanners have come to the rescue of the patient. The dermatologist can quickly scan all of the patient’s lesions and identify which ones are cancerous, and if any are, whether the tumour is sufficiently shallow that it will likely respond to non-invasive, non-scarring treatments. Furthermore, the dermatologist can perform follow-up scans to verify whether or not the tumour has been successfully treated and has disappeared or whether it requires further treatment.

OCT can also be used to map the lateral margins of a tumour prior to surgery so that the amount of skin removed can be minimized, while also reducing the risk of missing an outlying cancer growth that would require a second operation to remove. Clinicians are using OCT for all of these applications and more in the clinic on a routine basis, and clinical data that show its effectiveness are mounting up in the literature.

Many further applications are emerging, within both clinical applications and dermatological research. These include, for example, diagnosis of nail fungus and other nail conditions, scleroderma, psoriatic arthritis, skin parasites, blistering diseases and perhaps even melanoma. OCT could potentially result in game-changing improvements to the diagnosis and treatment of these and many other skin conditions. The future is bright for OCT imaging in dermatology.

• This article originally appeared in the Physics World Focus on Medical Imaging.

Related stories

• Optical imaging targets pre-clinical studies
• OCT aids early diagnosis of skin lesions
• All-optical: OCT and PAT in one
• High-resolution OCT explores the gut