COURSE P657 – SPRING 2017 Indiana University, Bloomington, Indiana
https://iu.instructure.com/courses/1607749
https://www.github.com/francopestilli/pestilli-teaching-2017
Tuesdays,& Thursdays 8:00-9:15AM, Room 286, Psychology Building.
Franco Pestilli, [email protected] Office: Room P359 Psychological and Brain Sciences. Office Hours: Tuesdays 9:20-10:20AM by appointment.
Historically, anatomical data were collected using camera lucida drawings in post-mortem brain to qualitatively describe the shape, extent and connectivity of the brain structures. Modern neuroimaging methods allow quantifying the precise location of the brain structures. These measurements can be performed post-mortem as well as in-vivo. They provide information about both microscopic brain tissue composition as well as macroscopic anatomical organization. For example, neuroimaging data is routinely used to measure white matter anatomy and tissue-composition, the profile of cortical brain maps, layers and their tissue organization. This is a transformative advance in the practice of neuroanatomy, with the potential to open many new avenues of understanding.
Digital neuroanatomy brings potential for strong scientific advances. It is allowing the creation of normative models of anatomical architecture, as well as the integration of anatomical information with data from both molecular and genetic measurements. However, in addition to the many benefits, the new digital methods come with modern challenges. Overcoming these challenges requires a change in perspective on neuroanatomy. Digital neuroanatomy requires setting unprecedented standards for data acquisition, preprocessing, and quality control. Digital neuroanatomy also encompasses the adoption of methods from modern statistics and machine learning, and the growing acceptance that the models that we develop require new quantitative validation methods. Current approaches to validation compare estimates from in-vivo neuroimaging to ex-vivo histological preparations. Modern digital data require integrating validation in the process of scientific investigation; digital validation is a continuous evaluation process by which anatomical structures are reported with associated validity scores.
The class will cover articles from the modern era of digital neuroanatomy, some history, methods, applications, neuroinformatics, big data neuroscience.