Joe has always been fascinated by the incredible feats performed by the nervous system – especially the control of movement.  He sees the nervous system as “a master puppeteer,” activating muscles, which pull on bones, which rotate about joints.  This system can learn to perform incredible feats -- from launching bodies into the aerial acrobatics of gymnastics to producing intricate melodies on a violin.  

This interest led him to his Ph.D. work on neuroscience and musculoskeletal biomechanics at the Harrington Department of Bioengineering at Arizona State University.  In graduate school, he developed experimental techniques and created computational models to better understand form and function in spinal motorneurons, the "final common pathway" of motor control. During this work, he became convinced that understanding the nervous system required anatomically-accurate computational modeling at multiple scales.  

But he also found that experimental reconstructions of neuronal morphology were rare.  So, he began developing computational algorithms to analyze reconstructed neuronal morphologies and to synthesize (algorithmically generate) "virtual" morphologies.  His ultimate goal was to generate virtual neurons which were statistically indistinguishable from the neuronal populations on which they were based. 

At the Blue Brain Project, he has continued this work, studying and generating different types of cortical neurons.  As the BBP scales up to larger and more extended models of the nervous system, there is an ever-increasing need for this work.  Algorithmic synthesis of neurons offers unlimited scalability: the ability to generate any number of virtual neurons, on demand, while creating the natural range of morphological variability.