Using iPhones to Identify Skin Cancer in Sweden

Post image for Using iPhones to Identify Skin Cancer in Sweden

[This is a preview of some of the exciting mHealth research being presented at the upcoming Medicine 2.0 Congress in September. This abstract and others are candidates for the iMedicalApps-Medicine 2.0 mHealth Research Award]

By: Alexander  Börve, PhD, Carin Sandberg, MD,  Karin Terstappen, PhD,  John Paoli, PhD
Finalists, iMedicalApps-Medicine 2.0 Research Award

Background

40,000 people get skin cancer every year in Sweden, and this number is increasing at 7% a year which is similar to other western countries. Not all forms of skin cancer are fatal; however, nearly all skin cancers can be cured if discovered early enough.

Nevertheless, skin cancer can be difficult to diagnose by a doctor who is untrained in skin diseases, while a trained dermatologist can diagnose skin cancer at a much higher rate.

In Sweden, 80% of doctors have a smartphone. Physicians have started to realize that these smartphones can be used as medical tools. Third party smartphone “add ons” and apps have made it possible to “ask a specialist” through smartphone communication. We looked at the possibility of using the built-in digital camera in a smartphone, a customized dermoscope (a special microscope to look at the skin), and a smartphone application to carry out “teledermoscopic” evaluations of suspicious skin lesions.

Study method

In a study conducted in 2011, we included 69 skin lesions (pigmented and non-pigmented) that were going to be operated on or biopsied to rule out the possibility of the lesion being cancerous. A dermatologist first looked at a visiting patient face-to-face and examined the suspicious lesion with a dermascope. Before performing an operation or taking a biopsy, the dermatologist took a clinical digital photograph with an iPhone4, a dermascopic digital photograph with an iPhone4 as well as the “add on” dermascope.

The images were uploaded to the iPhone application “iDoc24” and a brief text describing the patient history, the lesion appearance, a change in size and the colour or shape was included. Each case was encrypted and sent to a secure internet platform connected to the iDoc24 app for further independent assessment by two dermatologists that were trained in dermoscopy. The suspected diagnosis and a management decision was provided by the face-to-face dermatologists and later compared to the histopathological report.

Furthermore, the level of diagnostic difficulty and the image quality was assessed.

Study results

The diagnostic accuracy of the dermatologist meeting the patient face-to-face was comparable to that of the two teledermoscopists. Adequate management decisions were provided by the dermatologist in the clinic and by the first teledermoscopist for all 69 lesions and for 68 out of 69 lesions (98.6%) by the second teledermoscopist. One melanoma was missed by the second teledermoscopist.

Discussion

In our study there was one malignant melanoma missed.  To prevent this in a real clinical setting it is important that all dermatologists are adequately trained and that where there are difficult cases they can request another dermatologist’s opinion. In breast cancer screening which is another area of diagnostic difficulty, it is normal that there are two radiologists diagnosing to prevent missed breast cancers.

The level of diagnostic difficulty of the cases was rated as high by the two teledermoscopists in 61% and 87% of the cases respectively, as compared to 54% of the cases when scored by the dermatologist who saw the patients face-to-face. This could possibly be explained by a certain degree of diagnostic insecurity when not being able to meet the patient in real life.

Similarly, the image quality was rated as excellent or sufficient in all cases by the dermatologist who took the photographs, but in 84% and 94% of the lesions by the respective teledermoscopists.

Nevertheless, the use of photographs of skin lesions suspicious of skin cancer acquired with a smartphone camera and a customized dermoscope, and then sent to dermatologists through a smartphone app for teledermoscopic evaluation, seems feasible due to the comparable diagnostic accuracy and the adequacy of the management decisions provided despite the diagnostic difficulty of the cases.

A dermatologist cannot always diagnose skin cancer correctly in a clinical setting. If they are in doubt, they remove the lesion as a precaution. Sometimes lesions that are least suspected of being a skin cancer is actually a malignant melanoma. Ultimately, many lesions that are not skin cancer are removed due to the suspicion of skin cancer.

From an economical point of view it is therefore favorable if the ratio of “not skin cancer” to “skin cancer” is low. This ratio can possibly be lower if dermatologists with dermoscopy training are used. Our study has indicated that this ratio can have a similar rate with teledermoscopy as of face-to-face examination with a dermascope by a dermatologist.

Authors

Alexadner Börve is a PhD student in telemedicine and an orthopaedic surgeon from Sahlgrenska University Hospital in Gothenburg, Sweden. Alexander is an advisor to Health 2.0 Europe and is the Health 2.0 Stockholm chapter leader. He is currently on leave to finish his PhD and work on his start up iDoc24- Ask the dermatologist today!

John Paoli is a PhD and dermatologist, from Sahlgrenska University Hospital in Gothenburg, Sweden. His current research focuses on the prevention, early diagnosis and optimal therapy of melanoma and non-melanoma skin cancer. Dr. Paoli is currently Vice President of the Swedish Association of Dermatology and Venereology, President of the Swedish Association for Dermatologic Surgery and Oncology and a Board Member of the International Dermoscopy Society. He is also a certified Mohs Micrographic Surgeon. He is an expert in dermoscopy.

Carin Sandberg is a dermatologist and head of the dermatology department at Sahlgrenska University hospital, Gothenburg, Sweden. Her research interest is in skin cancer. She is an expert in dermoscopy.

Karin Terstappen is a PhD and dermatologist, at the dermatology clinic Skaraborg Hospital, Sweden. She is head of the regional skin cancer centre in Västra Götalands regionen, Sweden. Her research focuses on early detection of skin cancer and its implications for the west coast population, where early diagnosis and life quality of cancer patients is a priority. She is an expert in dermoscopy

 

Discussion ( 2 comments ) Post a Comment
  • Excellent article. What was the Level of the melanoma that was missed? If Level 1, then it’s not such a big deal. If it was a thick melanona — that is more significant.

    Which attachment for dermatoscopic photos did you use?

    Best wishes,

    David

Comment on this discussion

Your email is never published nor shared.