AI could revolutionise dermatology, but key hurdles remain

5 December 2018 (Last Updated December 5th, 2018 14:05)

The use of artificial intelligence (AI) in the field of dermatology could provide significant benefit to both patients and dermatologists alike. Not only does AI in dermatology have potential in alerting people when they may need to see a doctor (usually through the use of a smart device), it can also be used to create an educational resource for medical students, as well as a confidence boost to physicians making a differential diagnosis.

AI could revolutionise dermatology, but key hurdles remain

The use of artificial intelligence (AI) in the field of dermatology could provide significant benefit to both patients and dermatologists alike. Not only does AI in dermatology have potential in alerting people when they may need to see a doctor (usually through the use of a smart device), it can also be used to create an educational resource for medical students, as well as a confidence boost to physicians making a differential diagnosis.

GlobalData believes that implementing AI into a dermatological setting could result in greater diagnostic capabilities and optimised patient care outcomes.

Convolutional neural networks and teledermatology

One notable application of AI in dermatology has been the training of a convolutional neural network (CNN) to more accurately diagnose melanoma and less often misdiagnose benign moles.

This method has been shown to demonstrate greater sensitivity and specificity than 58 international dermatologists, as reported by the European Society for Medical Oncology (ESMO) in May 2018. A CNN is trained by taking an image and converting it into a series of numbers while retaining the visual information found in the original image.

This process is iterated many times over so that the CNN can compare a new image to the most visually similar images from a pre-established library of images that capture the same pathology. Through machine learning, the CNN can teach itself from what it has learned to improve performance.

In the study published by ESMO, it was revealed that when calibrated to identify benign moles to the same extent as the 58 dermatologists, the CNN correctly identified 95% of melanomas from a sample of 100 images, whereas the dermatologists identified 86.6% of melanomas.

Another example of integrating AI in dermatology is the appearance of teledermatology companies, such as First Derm. These companies produce applications that users can download to their smartphone and take pictures of a suspected condition, which are then reviewed by an online dermatologist who will provide a suspected condition and possible treatment choices for that particular skin condition.

This can help patients in deciding whether they need to see a doctor in person, with the app directing the patient to relevant dermatologists or clinics. As a result, the patient can have peace of mind and dermatologists are not inundated with unnecessary patient visits that could easily be solved by over-the-counter medication.

Challenges of implementing AI in dermatology

However, the implementation of AI in dermatology is not infallible and does present several issues that need to be addressed. Access to medical data sets needed to train AI can be difficult to attain at the scale necessary for training due to the need for patient consent and the potential requirement of approval from relevant ethics boards.

Moreover, the clinical datasets used for training are clinical images taken with a dermascope under controlled conditions; comparison of clinical data sets to images taken with a mobile phone in poor lighting conditions may lead to misdiagnoses.

Another issue for AI may be the inability to distinguish between visually similar conditions with different causes, such as rashes caused by infection compared to rashes caused by a drug reaction. Lastly, there is the potential for racial disparity in machine learning for skin cancer screenings.

This was seen in the study published by ESMO, where there were far fewer images of patients that were not Caucasian, raising concerns that the mortality rate for African Americans developing melanoma will remain higher than in Caucasians even with the advent of AI in dermatology, despite Caucasians having a much higher risk of developing melanoma.

To navigate the issue of patients not consenting to have their data stored, smartphone applications often employ data-sharing opt-in schemes for patients or having the patient’s image dumped after analysis so that no one else sees it. First Derm also allows free access to their services in exchange for patient data. With regards to issues surrounding lighting and picture size, some AI solutions are able to use the skin surrounding a lesion as a reference for the actual condition, meaning images taken by patients will not have to be of dermascope quality.