This AI system can empower doctors in diagnosing skin diseases

Researchers have developed a deep learning-based artificial intelligence (AI) algorithm that can accurately classify cutaneous skin disorders, predict malignancy, suggest primary treatment options, and serve as an ancillary tool to enhance the diagnostic accuracy of clinicians. With the assistance of this system, the diagnostic accuracy of dermatologists, as well as the general public, was significantly improved, said the study, published in the Journal of Investigative Dermatology.

Skin diseases are common, but it is not always easy to visit a dermatologist quickly or distinguish malignant from benign conditions. “Recently, there have been remarkable advances in the use of AI in medicine. For specific problems, such as distinguishing between melanoma and nevi, AI has shown results comparable to those of human dermatologists,” said lead investigator Jung-Im Na from Seoul National University in South Korea. Most prior studies have been limited to specific binary tasks, such as differentiating melanoma from nevi.

“Our results suggest that our algorithm may serve as an Augmented Intelligence that can empower medical professionals in diagnostic dermatology,” Na added.

Using a “convolutional neural network,” a specialised AI algorithm, the research team developed an AI system capable of predicting malignancy, suggesting treatment options, and classifying skin disorders. They collected 220,000 images of Asians and Caucasians with 174 skin diseases and trained neural networks to interpret those images.

They found that the algorithm could diagnose 134 skin disorders and suggest primary treatment options, render multi-class classification among disorders, and enhance the performance of medical professionals through Augmented Intelligence.”Rather than AI replacing humans, we expect AI to support humans as Augmented Intelligence to reach diagnoses faster and more accurately,” Na said.

The researchers caution that AI cannot definitively interpret images, that it is not trained to interpret even when the problem presented is straightforward. For example, an algorithm trained only to differentiate between melanoma and nevi cannot differentiate between an image of a nail hematoma and either a melanoma or a nevus. If the shape of the hematoma is irregular, the algorithm may diagnose it as melanoma.

They also pointed out that the algorithm was trained and tested using high-quality images and its performance is generally suboptimal if the input images are of low quality. “We anticipate that the use of our algorithm with a smartphone could encourage the public to visit specialists for cancerous lesions such as melanoma that might have been neglected otherwise, however, there are issues with the quality or composition of photographs taken by the general public that may affect the results of the algorithm, Na said.