July 2022

Creating a table-top imaging device capable of fast and accurate skin cancer diagnosis 

Professor Ioan Notingher is a Professor of Physics specialising in Biophotonics at the University of Nottingham. He has been awarded a British Skin Foundation PhD Studentship entitled "Development of multimodal optical microscopy for imaging tumour margins during skin cancer surgery".

Above image: Professor Notingher (left), Professor at the University of Nottingham, and Dr. Sandeep Varma (right), Consultant Dermatologist at the Queen's Medical Centre, Nottingham.

What is squamous cell carcinoma of the skin?

Whilst skin cancers vary in both appearance and location, they can be grouped into two major types: melanoma, and non-melanoma skin cancer (NMSC). Between the two, NMSCs are less aggressive but significantly more common, comprising some 90% of skin cancer cases in the United Kingdom. Of the NMSCs, the two most common types are basal cell carcinoma (BCC) and squamous cell carcinoma (SCC). SCCs are less common (approximately 23% of NMSC cases) but pose a greater risk to human life because in some cases, they can spread to the lymph nodes or other parts of the body. They can occur anywhere on the skin but are often found in sun-exposed areas such as the head, neck, ears, and lips. SCCs can vary in appearance, but often appear as scaly or crusty areas or lumps, and can sometimes feature an ulcer which is sore or tender, and in some cases can bleed easily.

NMSC treatments

Both BCCs and SCCs are usually treated by removing them surgically under a local anaesthetic. In some circumstances, especially large or recurrent lesions, a technique called Mohs micrographic surgery is used. During surgery, a sample of skin is removed and checked for any residual tumour. To do this, the skin is cut into sections, colour-coded with dyes and rapidly frozen and examined under a microscope to look for any remaining skin cancer cells. This is known as frozen section histology. Successive horizontal layers of the skin sample are cut, and each layer is checked under microscope for cancer. If cancer is detected, further sections of skin are removed from the patient until no more cancer is detected. Though Mohs surgery is very effective at removing NMSCs with as little damage to healthy skin as possible, it can be expensive as the frozen section histology technique requires histopathology technicians and specialist surgeons trained to interpret these skin tissue samples. Further, the time delay between skin tissue removal and result (cancer found/not found) can sometimes prevent patients from achieving a same-day result.

How can this be improved?

We wish to improve the effectiveness of Mohs surgery by developing techniques to diagnose skin tissue samples faster, more reliably, and without requiring extra tissue processing. To do so, we are developing a prototype table-top device capable of imaging complete Mohs samples within 30 minutes. This device can currently diagnose BCCs using a Fast Raman Technique, and has indicated promising, repeatable results in a study of skin cancer patients at Nottingham. Additionally, this device will have the capability to be operated by clinical personnel after only a few hours of training.

The Fast Raman technique 

Our Fast Raman technique uses two imaging processes: auto-fluorescence imaging and Raman micro-spectroscopy (RMS). RMS works by shining light (usually from a laser) onto the skin layer and measuring the resultant spectrum, which is a graph of how the light’s intensity varies at different frequencies. Some photons of lights from the laser will interact with molecular excitations within the skin sample and become scattered. These photons will appear as spikes on the spectrum and contain information on molecular excitations of the material, allowing us to determine the sample’s chemical composition. The pattern of the spikes from a skin cancer has a specific signature that allows us to determine whether any cancer is present in the sample imaged.

One issue, however, is that RMS is still a relatively slow process; and can only image a 1mm x 1mm area in about 5 hours. This makes a pure-RMS technique unsuitable for skin cancer surgery where we need to diagnose a considerably larger area in only 30 minutes. This is where we use auto-fluorescence imaging, which creates less detailed images but is significantly faster. As part of the Fast Raman technique, we apply image recognition techniques to scan the auto-fluorescence images for areas of likely cancer, and then selectively perform RMS measurements of only these regions of interest. This enables us to accurately determine whether the skin sample contains any NMSC without needing to perform an RMS measurement of the entire surface.

What we plan to do

At present, our prototype device is underway with small-scale clinical studies to reliably determine its accuracy when diagnosing BCCs. Although we have made good progress in detecting BCC cells, we have not yet developed the device to pick up SCC cells. Along with larger-scale clinical trials, we aim to further expand the prototype’s capabilities to also diagnose SCCs. In doing so, we intend to provide a more accurate, faster, and less expensive alternative to frozen section histology techniques when removing SCC.  We wholeheartedly thank the generosity of British Skin Foundation donors, who have enabled us to continue to push this research forward to further improve the care of those being treated for skin cancer.

Professor Ioan Notingher, Professor at the University of Nottingham

Donate to fund more research like this    Find out about our research     Find out more about squamous cell carcinoma

By donating to skin disease research you are helping us to find treatments and cures for common conditions like eczema, acne and psoriasis through to potential killers like melanoma skin cancer. Thank you.