The time when artificial intelligence may be the key tool for detecting skin cancer is near, experts say, without undervalued the factor of human labor.

“Humanity has really matured and we have now surpassed the phase of the advertising campaign and the big media headlines,” says Ivy Lee, a Los Angeles -based dermatologist chairing the Augmented Intelligence Committee.

Work in 2017 in Nature Magazine summarizes that a model of artificial intelligence that examined 129,450 clinical images exceeded 21 dermatologists’ performance in diagnosis of skin cancer.

Although the above is promising, the work points out that the dermatologist is able to ask questions to the patient about his problems, to examine, and to gain a clearer perception of the patient’s clinical picture through the touch, says Veronica Rotemberg, a dermatologist at the Memorial Sloan Kettering Center.

“There is much more about clinical medicine than looking at a photo.”

In today, experts point out how these artificial intelligence models can help the daily lives of doctors and patients.

“We wonder how we can make use of it practically and essentially in the real world,” says Lee.

“Overall, I’m optimistic,” he adds.

“We know we have to do the best for more people.”

More information, fewer unnecessary biopsies

With a handheld device AIwhich a doctor “scans” over lesions on the skin for analysis via visual spectroscopy, can give a “stigma” for the three most common skin cancers – basal cell carcinoma, coil cell carcinoma and melanoma.

This can help reduce unnecessary dermatologists for biopsies in sub-served areas, says Mitchell A. Kline, a clinical assistant professor at New York-Presbyterian Weill Cornell Medical Center.

The devices are intended for general physicians and other health professionals who do not specialize in dermatology.

Dermasensor has funded research on such a product that was published in the Journal of Primary Care & Community Health and showed many promising early data on its use by primary care doctors.

Dermasensor did a detailed research in detecting skin cancer, but also had a high percentage of false positive results and does not record all types of lesions, according to Roxana Daneshjou, Assistant Professor of Biomedical Science and Dermatology at Stanford University.

However, its greatest concern, repeated by other experts, is whether such models have been tested or designed to function properly in different skin tones.

Other markets on the market are promoting Skin Analytics, an approved product in the European Union as a medical device, and Nevisense, another US -based company whose device has won the approval of the Food and Drugs Service before buying melanoma.

The next level of skin control

Sloan Kettering and other institutions use three -dimensional scanners to photograph and watch patients with a complex risk of melanoma.

Patients have access to the images when they are examining them so they can either eliminate their stress or indicate suspicious damage for their doctor to examine them more carefully. Today, artificial intelligence is used as part of the three -dimensional body scan to performance and their reconstruction.

Version 2.0 of this device would entail the utilization of artificial intelligence to automatically detect skin cancer.

Some companies are already offering services such as this, but experts note that they have not seen publicly available data to prove their effectiveness or regulatory regime in the United States or abroad.

Daneshjou adds that there may be companies that are doing tests and then apply for approval by the FDA.

“I think it’s going.”

Lee also says that there could be a future where these whole body scans will become widely accessible – and that would be positive.

“Anything about improving access to screening and high quality care is something we have to insist on,” he says.

This would require powerful scientific evidence showing that they are safe, financially beneficial and fair, and have been made after proper regulatory interventions to ensure that they are viable and expandable.

It takes time

Experts point out that we should not rely on applications and chat bot to diagnose skin cancer.

‘It’s actually quite problematic because [αυτές οι εφαρμογές] They have not overcome any regulatory barrier and some of them claim to be able to diagnose skin cancer and there is no real data available to support their allegations, “says Daneshjou, who wrote a work on Jama Dermatology in 2024 with Rotemberg.

Mobile applications and chatbots can be “dangerous because we often target these tools as patients because we cannot have access to health care,” Lee says, adding that confidentiality and data safety should also be concern.

Experts are also not the use of chatgpt and other famous technological models to analyze skin lesions.

“What we have found from previous work is that these models do not do so well in the diagnosis of skin cancer,” says Daneshjou.

Rotemberg says that products addressed directly to the consumer remain “away” than being useful for patients, although they are “a very promising sector”.

The best option is to do an annual skin control with a dermatologist, Lee says, noting that if one observes something different from the usual skin, he will be resolved immediately to a specialist.