The vast majority of medical computer-based training (CBT) systems aim at problem-oriented case based training. A crucial issue in the design of CBT systems is the selection of appropriate cases for a representative collection of medical cases. Another issue is the presentation of those case collections to the users. With simple numbered list (e.g. Case 1, Case 2, ...) goal-oriented case selection driven by learning objectives and clinical problems is not feasible. In this paper, we investigate interactive information visualization techniques, which promise a more sophisticated case selection. We analyze, compare, and empirically evaluate their appropriateness for case selection in medical CBT systems. Whereas analysis results show superiority of interactive information visualization techniques, the empirical evaluation reveals preference of the target group for simple tabular representations.