Oregon State University



Event Details

PhD Oral Preliminary Examination – Anahita Sanandaji

Monday, September 26, 2016 10:00 AM - 12:00 PM

Developing Training Strategies for 3D Volume Segmentation by Analyzing the Underlying Human Perception and Cognitive Tasks
3D volume segmentation is a fundamental process in many scientific and medical applications. Producing accurate segmentations, in an efficient way, is challenging, in part due to low imaging data quality (e.g., noise and low image resolution), and ambiguity in the data.  Automatic algorithms do exist, but there are many use cases where they fail. The gold standard is still manual segmentation or review.  Unfortunately, even for an expert, manual segmentation is laborious, time consuming, and prone to errors.  Existing 3D segmentation tools are often designed based on the underlying algorithm, and do not take into account human mental models, their lower-level perception abilities, and higher-level cognitive tasks.

We propose to analyze manual segmentation as a human-computer interaction paradigm to gain a better understanding of both low-level (perceptual) actions, and higher-level tasks and decision making process. We initially employed formative field studies using our novel hybrid protocol that blends observation, surveys, and eye-tracking. We then, developed and validated data coding schemes to discern segmenters' low-level actions, higher-level tasks, and overall task structures. We could successfully identify workflow patterns and different segmentation strategies utilized by expert versus novice segmenters. Based on our gained knowledge, we hypothesize that the majority of the experts' skill sets can be characterized by spatial sub-skills such as understanding 3D structures and ability to predict the 2D contours of a 3D shape using an arbitrarily oriented slicing plane. To test this hypothesis, we propose developing a measurement instrument to evaluate segmenters' spatial abilities and how they build mental model of 3D structures. The measurement instrument will also help us evaluate the effectiveness of our tool-agnostic training strategies. The ultimate goal of our research is to develop a training pedagogy to help humans do segmentation in a more efficient and accurate way, while separating learning the tool set from learning how to segment.

Major Advisor: Cindy Grimm
Committee: Eugene Zhang
Committee: Carlos Jensen
Committee Ronald Metoyer
GCR: Leonard Coop

Kelley Engineering Center (campus map)
Nicole Thompson
1 541 737 3617
Nicole.Thompson at oregonstate.edu
Sch Elect Engr/Comp Sci
This event appears on the following calendars: