John R. Pani - US grants

This project is a multidisciplinary study of cognition in the practice of histology, the microanatomy of tissues. Histology is a core course in the biological, premedical, and medical school curricula, and pathologists use its methods to diagnose a variety of medical disorders. Instruction in histology centers on the recognition of tissues in microscopes. This is a very challenging task that takes years to master. Much of the basis for this difficulty is that the tissue in microscope slides differs substantially in appearance from the whole tissue. As a result, ability in histology requires the development of high level skills of visual recognition and reasoning. This project will investigate the current and optimum methods for gaining expertise in basic histology through a variety of experimental studies. The specific aims of the proposed research are the following: 1) Determine the information for human perceivers in microscope slides. Two studies will generate models of the use of diagnostic information in microscope slides for individuals who have completed the first year course in histology. 2) Assess the importance of visuospatial understanding of microanatomy for identification of tissues in microscope slides. An experiment will be conducted to test the hypothesis that the ability to identify structures and tissues in microscope slides will vary directly with the degree to which people have a clear visuospatial understanding of the microanatomy of the whole tissue. 3) Assess the importance of visuospatial understanding of the slice transformation for identification of tissues in microscope slides. An experiment will be conducted to test the hypothesis that the ability to identify structures and tissues in microscope slides will vary directly with the degree to which people have a clear visuospatial understanding of the transformation that relates 3D anatomy to 2D anatomical sections. The results of this research will lead to a detailed understanding of cognition in histology and to advances in theories of visual symbol systems used in the interpretation of images generated by human technology (e.g., x-ray, mri, aerial photographs). This project has the potential to generate substantial change in the methods for developing skill in histology at both the undergraduate and the medical school levels. The study of virtual reality methods of learning in conjunction with a rigorous study of histological reasoning will lead directly to the development of software that may significantly accelerate the development of skill in this challenging and important field.

The interpretation of imagery acquired through technical means is a cognitive skill that is central to nearly every advanced discipline. Reading x-ray, CT, and MRI images is a mainstay in the modern practice of medicine. Interpreting microscope slides is fundamental in histology and pathology. Interpreting a variety of aerial and satellite imagery is standard practice in geology, navigation, and environmental science, as well as in the gathering of national intelligence information. Numerous studies of cognition across the broad array of different types of technical imagery suggest that image interpretation is extremely challenging. Expertise takes many years to acquire, and even highly trained experts continue to make errors. Relatively little is known about what makes the interpretation of technical imagery so challenging. Even less is known about how to improve training methods to make the acquisition of expertise more efficient. This project will advance basic cognitive theory with regard to the fundamental nature of cognition in technical image analysis. The broader impact of this work will be twofold. First, the project will develop empirical techniques for study of technical image analysis that can be applied across different domains of imagery. These techniques will isolate the sources of error at different levels of expertise and lead to the development of instructional techniques to make training more efficient. Second, new training methods that involve the use of 3D computer graphics will be developed to increase the efficiency of training in image analysis and to generate more robust skills in image interpretation. This project will develop instructional techniques that use interactive 3D graphics to encourage a deeper understanding of the nature and variety of technical images in a domain. These techniques will make training in image analysis substantially more efficient and lead to skills that generalize more readily to novel images.