Does the algorithm provide a diagnosis for my skin lesion?

Please be sure that the result of the algorithm is NOT A CONCLUSIVE DIAGNOSIS. The algorithm is not intended to be directly used for skin disease diagnosis. In-person examination of dermatologists is much more accurate. The algorithm could provide relevant documents for the images. However, in-person physical examination and history taking are required to get a correct diagnosis.

How accurate is the algorithm?

The algorithm could provide more accurate information than random internet searches, and the accuracies of the algorithm were equivalent to that of general physicians. For the internet community images (Reddit), the accuracy of the algorithm was 39.2% (1st result; Top-1 accuracy) / 67.2% (1st ~ 3rd results; Top-3 accuracy).

How should I use the algorithm?

The result of the algorithm can help to obtain medical information about the signs and symptoms of highly relevant diseases. It can deliver more accurate clues than a search engine to find medical information. Please seek a doctor’s advice in addition to using this app before making any medical decisions.

What does ‘High Risk’ mean?

If there is a lesion with a repeated result of high risk, it is more appropriate to see a dermatologist rather than a general practitioner. In the actual examination, diagnosis is made by considering various factors such as changes in lesions. Please be aware that making a medical decision based only on the result of the App can lead to overdiagnosis. 

Is it Free?

The use of the algorithm is FREE. 

In what setting does the algorithm work best?

The algorithm can give the most accurate advice on Asian skin lesions. The algorithm was trained and validated on lesions from the images of the white population. However, it has not yet been validated for the black population and may show uncertainty.

Are the submitted photographs stored?

The photograph is transferred for cloud analysis, however, we do NOT store your images. Simple having big data does not improve the performance of the algorithm. We have established a skin image dataset by annotating data published on the Internet, not user data.

How many disease classes have been validated?

A total of 133 disease classes have been validated. The accuracy and balanced accuracy of the algorithm were comparable with those of dermatologists in the experimental setting.  

ABNOM, Abscess, Acanthosis nigricans, Acne, Actinic keratosis, Acute generalized exanthematous pustulosis, Alopecia areata, Amyloidosis, Androgenic alopecia, Angiofibroma, Angiokeratoma, Angular cheilitis, Atopic dermatitis, Basal cell carcinoma, Becker nevus, Blue nevus, Intraepithelial carcinoma (Bowen disease), Bullous disease, Cafe au lait macule, Cellulitis, Condyloma, Confluent reticulated papillomatosis, Congenital nevus, Contact dermatitis, Dermatofibroma, Drug eruption(Viral exanthem), Eczema herpeticum, Epidermal cyst, Epidermal nevus, Erythema ab igne, Erythema annulare centrifugum, Erythema multiforme, Erythema nodosum, Exfoliative dermatitis, Fifth disease, Folliculitis, Freckle, Furuncle, Granuloma annulare, Guttate psoriasis, Hand eczema, Hemangioma, Herpes simplex, Herpes zoster, Hypertrophic scar, Idiopathic guttate hypomelanosis, Impetigo, Inflammed cyst, Ingrowing nail, Insect bite, Juvenile xanthogranuloma, Keloid, Keratoacanthoma, Keratoderma, Keratosis pilaris, Lentigo, Lichen nitidus, Lichen planus, Lichen simplex chronicus, Lichen striatus, Livedo reticularis, Lupus erythematosus, Lymphangioma, Malignant melanoma, Melanocytic nevus, Melanonychia, Melasma, Milia, Molluscum contagiosum, Morphea, Mucocele, Mucosal melanotic macule, Mucous cyst, Nail dystrophy, Neurofibroma, Neurofibromatosis, Nevus depigmentosus, Nevus spilus, Nummular eczema, Onycholysis, Onychomysosis, Orgarnoid nevus, Ota nevus, Palmoplantar pustulosis, Paronychia, Perioral dermatitis, Pigmented progressive purpuric dermatosis, Pityriasis alba, Pityriasis lichenoides chronica, Pityriasis lichenoides et varioliformis acuta, Pityriasis rosea, Poikiloderma, Pompholyx, Porokeratosis, Poroma, Portwine stain, Prurigo nodularis, Prurigo pigmentosa, Psoriasis, Pustular psoriasis, Pyoderma gangrenosum, Pyogenic granuloma, Riehl melanosis, Rosacea, Scabies, Sebaceus hyperplasia, Seborrheic dermatitis, Seborrheic keratosis, Skin tag, Squamous cell carcinoma, Staphylococcal scaled skin syndrome, Steatocystoma multiplex, Striae distensae, Subungal hematoma, Syringoma, Telangiectasia, Tinea corporis, Tinea cruris, Tinea faciale, Tinea pedis, Tinea versicolor, Urticaria, Urticarial vasculitis, Urticaria pigmentosa, Varicella, Vasculitis, Venous lake, Verruca plana, Viral exanthem(Drug eruption), Vitiligo, Wart, Xanthelasma, Xerotic eczema