Guide-linked quiz · Medium Level

AI in Dermatology Quiz

Review dermatology image AI, its diagnostic-triage role, skin-tone representation challenges, image inputs, and clinical validation limits.

Related guide pathsAI In Dermatology
Question 1 of 8

What did the landmark 2017 Stanford dermatology study demonstrate?

Preview every question8 questions

Answers and explanations appear as you take the quiz.

  1. What did the landmark 2017 Stanford dermatology study demonstrate?
    • A neural network classified skin cancers as accurately as 21 board-certified dermatologists
    • AI could perform skin surgery
    • Skin cancer cannot be detected from images
    • Dermatologists are always wrong
  2. What is a major fairness problem documented in dermatology AI?
    • Models work only on cartoon images
    • Many models perform worse on darker skin tones due to light-skin-dominated training data
    • Models can only see in black and white
    • Models refuse to analyze freckles
  3. What does dermoscopy add to a standard skin photograph?
    • A blood pressure reading
    • A 3D hologram
    • Magnification and cross-polarized light revealing sub-surface pigment and vessels
    • An X-ray of the bone underneath
  4. What spurious shortcut have dermatology models been caught learning?
    • Counting the patient's freckles
    • Reading the date stamp on the photo
    • Detecting the camera brand
    • Flagging lesions photographed next to a ruler or skin marker as malignant
  5. What is the 'gold standard' these models are usually validated against?
    • Biopsy-confirmed diagnoses
    • A patient's self-assessment
    • The color of the lesion
    • A coin flip
  6. What is the most likely near-term clinical role for dermatology AI?
    • Replacing all dermatologists immediately
    • Triage — helping decide which lesions need urgent specialist review or biopsy
    • Prescribing antibiotics automatically
    • Performing the biopsy itself
  7. What is the purpose of building datasets such as Fitzpatrick17k and Diverse Dermatology Images?
    • To replace biopsy confirmation in clinical care
    • To collect only images of light skin
    • To support evaluation and training with a wider range of skin tones and reduce representation gaps
    • To convert dermoscopic photographs into X-rays
  8. What can AI-assisted total-body photography help flag for clinician review over time?
    • Changes in a patient’s blood pressure
    • The presence of a bone fracture
    • Whether a biopsy result is benign
    • New or changing skin lesions

Keep testing yourself

More quizzes picked for your level.

View all quizzes