May 23, 2019

Why you shouldn’t trust today’s skin cancer apps for diagnosing melanoma

Healthcare

4 min read

The number of smartphone apps that claim they can identify skin cancer has grown hugely. These apps assess smartphone photographs of skin lesions using proprietary algorithms, without the involvement of a dermatologist. Studies have shown that the use of smartphone apps for diagnosing skin cancer is not recommended. Here’s why.


1. You cannot trust the findings

Studies indicate that skin cancer apps have poor diagnostic accuracy for melanoma. The number of cases correctly identified as melanoma (also called sensitivity) ranges from 0 to 70%. The cases that were correctly identified as not being melanoma (also called specificity) range from 30 to 93.7%.1

To put things in perspective: when patients do a self-examination of the skin, reported sensitivity increases from 25 to 93% and specificity ranges from 83 to 97%2. Conclusion: it’s better to trust your own findings than those of an app.

This is why medical professionals only use medical-grade equipment and tools (medical CE and/or FDA approved) and thoroughly investigate whether a diagnostic support app or system is clinically and scientifically validated.


2. Apps don’t pick up every symptom

Without specialist input, apps may not recognize rare or unusual cancers. They may not detect all red-flag symptoms3  like e.g. scaly, crusted, ulcerated areas or melanomas which do not produce pigment (amelanotic melanomas). This could increase the number of false negatives and delay treatment.

What’s more, with apps, patients tend to focus more on moles that look suspicious to them (a non-professional), while ignoring more worrisome lesions a doctor would pay attention to. Screening only a few of the most visible moles can create a false sense of security.


3. Photographs don’t show and tell

When screening for skin cancer, dermatologists take special dermoscopic images of the skin, using a dermatoscope. Dermoscopy has proven to provide highly accurate diagnoses because it can reveal subtle details in malignant skin lesions. Dermoscopic images can unveil e.g. blue-white pigmentation or asymmetries that suggest melanoma. These clues can hardly be seen in photos (clinical images) alone.

And that’s the issue with skin cancer apps. These apps use standard photos taken with a smartphone camera instead of dermoscopic images. The quality of the images may be influenced by lighting, angles of imaging, hair, etc.4  Also, apps usually cannot detect changes in structure or size of lesions over time, which is an important indicator when screening for skin diseases.

What’s more, not all smartphone cameras can provide good-quality, high-resolution pictures with sufficient color reproduction. Needless to say, the quality of the image will have a significant impact on the accuracy of the diagnosis.


4. No compliance with medical regulations

Researchers say that skin cancer apps vary in quality and that some have not been tested properly to show that they work and are safe5. A 2013 University of Pittsburgh studyfound that three out of four unnamed skin cancer apps wrongly classified 30% of cancerous growths as harmless.

In 2015, the Federal Trade Commission took action against two melanoma detection apps for deceptively claiming their mobile apps could detect symptoms of melanoma, even in its early stages, without having scientific evidence to support such claims7. Apps must have substantial scientific backing in order to make claims that suggest increased detection or risk analyses of melanoma.

Throughout the world, medical device regulations say that any application that claims to help diagnose a condition is a medical device. Therefore, it must comply with the applicable regulations and standards. In Europe, this means that the CE mark must be displayed on the app or on the developer’s website. In the US, the app needs to be cleared by the FDA.


5. Apps can cause anxiety

As skin cancer apps have a moderate-to-high sensitivity but only moderate specificity, they might increase the risk of unnecessary removal of pigmented skin lesions and create more dermatologist visits, which would do harm not only to the patients – it would also be expensive to society.

Nevertheless, there certainly might be potential for smartphone applications. Skin cancer apps can educate on prevention, screening, and the importance of early detection. They can aid consumers in assessing the risk of a skin lesion and could potentially improve the quality of patients’ skin self-examinations.

Still, if a lesion is changing in shape or size or just doesn't seem right, go and see your GP or dermatologist regardless of what any smartphone app tells you.


Interested in more? Read What is the best way to look at skin cancer images?


Wolf JA, Moreau JF, Akilov O et al. Diagnostic inaccuracy of smartphone applications for melanoma detection. JAMA Dermatol. 2013; 149: 422.
2 Hamidi R, Peng D, Cockburn M. Efficacy of skin self-examination for the early detection of melanoma. Int. J. Dermatol. 2010; 49: 126–34.
3 Three major failings in some apps used for the diagnosis of skin cancer, University of Birmingham, July 5, 2018
4 Maier T, Kulichova D, Schotten K et al. Accuracy of a smartphone application using fractal image analysis of pigmented moles compared to clinical diagnosis and histological result. J. Eur. Acad. Dermatol. Venereol. 2015; 29: 663–7.
5 Skin cancer: smartphone diagnostic apps may offer false reassurance, warn dermatologists, Jacqui Wise, BMJ 2018; 362 doi: https://doi.org/10.1136/bmj.k2999 (Published 05 July 2018), BMJ 2018;362:k2999
6 Diagnostic Inaccuracy of Smartphone Applications for Melanoma Detection, Joel A. Wolf, BA; Jacqueline F. Moreau, BA; Oleg Akilov, MD; et al, JAMA Dermatol. 2013;149(4):422-426. doi:10.1001/jamadermatol.2013.2382
7 https://www.ftc.gov/news-events/press-releases/2015/02/ftc-cracks-down-marketers-melanoma-detection-apps

Stay in touch

Receive the latest news about our service & products 

we will not share your e-mail address with 3rd parties