The visualization of privacy filters for sharing sensor-based health data

This dissertation explores interactive data visualization techniques that can enable patients to understand how privacy filters can govern the sharing of sensed health data with their social networks, how the customization of granular privacy filters influences data segmentation of their Personal Health Records (PHRs), and how privacy preferences can be mapped to secure in-place health data. A user study was conducted to evaluate these visualization techniques. Experiments were conducted to demonstrate the feasibility of securely sharing sensed health data in social networking applications.;Individuals deliberately share their personal health data with a number of audiences that can include the following:;• Patients that want to keep their family, friends and caregivers apprised of their health status and medical events.;• Patients that blog about their health.;• Individuals that want to augment their doctors' otherwise fragmented records of their health data.;• Individuals that share their health data in anonymized form for inclusion in aggregated data stores.;• Individuals that share their personal fitness or wellness monitoring data.;This sharing is done through a variety of methods, including patient portals to PHRs, health information exchanges, personal health blogs, and other social networking applications.;With the increasing use of sensors to monitor patients with chronic illnesses and in personal fitness monitoring applications, it's important to provide individuals with methods to understand what personal health data could potentially be shared and with whom. Unlike health data in-place from sources such as laboratory results or doctor's findings, which can be controlled using per-record privacy settings, sensors are dynamic by nature and require privacy filters configured in advance of sensor generated data. This dissertation provides the results of a user study conducted to explore visualization techniques designed to enable patient awareness of how these privacy filters govern the sharing of their health data.;To maintain security for personal health data, it's important that privacy preferences be continued with data that is transferred from PHRs. This can be accomplished using a variety of access control and encryption techniques. This dissertation demonstrates the feasibility of continuing privacy preferences using Attribute-Based Encryption (ABE).