Skip to Accessibility Tools Skip to Content Skip to Footer
Two phones showing an app that patients could potentially use to get advice about suspicious looking moles

An App to Screen for Skin Cancer?

What do you think of the idea of an app to screen for skin cancer? A friend recently sent me an article with the following headline: “Hoping to deliver free skin cancer screening worldwide, two software developers used artificial intelligence to create an app to detect skin cancer in real-time.” (Skin Cancer Detection Using Artificial Intelligence – iQ by Intel).

Artificial Intelligence body scans

My immediate thought was to not even read the article. I’m the person who said “no” when the PA in my dermatologist’s office told me she could perform my twice-yearly body scan. I only trust my doctor (who just left the practice, by the way). But curiosity led me to read about Dr. Hazel, an app that uses Artificial Intelligence (AI) to analyze a photo of a suspicious mole. The app tells the person if the mole looks like it could be cancerous by sorting through a database of images to classify it.

Potential for false negatives

According to the developers, Dr. Hazel can scan thousands of images in seconds. As of now, they report an accuracy rate of 85%, which they predict will increase as more data is collected. The goal is to provide free skin cancer screening for everyone. But, as the developers note, the worst thing that could happen (which is pretty bad) is that the app would tell a person they don’t have cancer when in fact they do. So the app isn’t being released into the public’s hands right now.

More development is needed

The initial plan is to have health care providers – primary care physicians, nurses, and pharmacists — use Dr. Hazel to provide a quick and inexpensive diagnosis of suspicious moles. The developers do hope to eventually have patients use the app themselves. They also plan to include instructions on what to do if the findings suggest that the mole is potentially a skin cancer.

In sum, there’s more work to be done on Dr. Hazel.

Here are my initial reactions:

Preferring a trained health care professional

I want a trained health care professional to diagnose any suspicious moles that I have. I want to go to someone who is trained to look at a mole and determine whether or not it needs to be biopsied. So for now at least, I’ll wait for my dermatologist to examine me.

How reliable is the image database?

But how do I feel about the possibility of my dermatologist using Dr. Hazel to analyze a mole? I’m still not sure. I worry about the images that are in the current database, the images that my mole will be compared to in order to determine if it needs to be biopsied. I’m concerned that just because my mole doesn’t look like one contained in the database, it could be considered not at risk of being cancerous, and wouldn’t be biopsied. Perhaps, as time goes on and more images are included in the database used by Dr. Hazel, I may feel more comfortable.

Is it better than nothing?

But what about the potential use of Dr. Hazel in remote areas, in those places with few or no dermatologists? My first gut reaction is that Dr. Hazel is better than nothing, better than not showing the suspicious mole to someone. But I feel that even if Dr. Hazel categorizes a mole as probably not cancerous, patients should be told (repeatedly) that they should see a dermatologist if the mole changes. I worry that someone will be told that a mole isn’t cancerous when in fact it is. And as of now, I’m thinking that I want a trained dermatologist to examine a mole, determine whether or not it could be cancerous, and if the mole is deemed okay, to tell the patient to follow up in 6 months or sooner if it changes.

I’m wondering, what do you think of the idea of an app that screens for skin cancer? Do you think you’d use it? Would you want your doctor to use it? Would it ever replace a yearly body scan by the physician? What do you think is the best use of such an app? Please send your thoughts and your opinions.

This article represents the opinions, thoughts, and experiences of the author; none of this content has been paid for by any advertiser. The team does not recommend or endorse any products or treatments discussed herein. Learn more about how we maintain editorial integrity here.