Perceptual and Adaptive Learning in Medicine

Medical practice requires pattern recognition in the interpretation of many types of clinical presentations, such as recognizing key features in a patient’s narrative to diagnosing cardiac problems based on such features as heart sounds, an EKG tracing, clinical laboratory results, or a cardiac ultrasound scan. Although the guidelines for making such interpretations are taught in the course of medical school and residency, proficiency in clinical interpretation is primarily gained via experience, which frequently involves clinical practice over many months or years, since the actual presentations of features varies widely from person to person.

We have created a large number online training modules (perceptual and adaptive learning modules, or PALMs) that incorporate both perceptual learning (PL) and adaptive response-time sequencing (ARTS) in order to accelerate both accuracy and fluency (automaticity) in interpreting clinical information. Importantly, we have found that PALM training leads to dramatic improvements in both short- and long-term proficiency.

These PALMs are currently being employed in a variety of health-related settings, from training medical students to active practitioners, making their development of both applied and theoretical importance. The availability of a large number of trainees using PALMs make these modules ideal for projects aimed at identifying factors that optimize learning technologies. A few examples of possible research questions to pursue are —

Variations in presentation mode that might affect learning and retention:

  • Are multiple-choice answers vs. “click on” a particular feature approaches equally useful?
  • How well does learning from each approach generalize to the other approach?
  • Does intermixing approaches benefit learning speed, retention, and/or generalization?

Factors that predict optimal sequencing and spacing:

  • What is the optimum spacing of category exemplars?
  • Does this vary with learning difficulty?
  • Can we predict optimal long-term “refresher” times from an individual’s PALM learning outcomes?

Individual differences in types of perceptual learning performance:

  • Does perceptual learning performance correlate with that for declarative learning?
  • Do individual differences in perceptual learning performance vary with the nature of the learning stimuli (e.g., visual vs. auditory)?
From the Histopathology PLM
From the Wrist X-Ray PLM

Examples of Learning Modules


  • Philip J. Kellman Philip J. Kellman
  • Everett Mettler Everett Mettler
  • Tim Burke Tim Burke


  • Sally Krasne Sally Krasne (UCLA)
  • face Erik Dutson (UCLA)
  • face Cheryl Hein (UCLA)

Selected Publications