George Howard

Pranav Kulkarni, Brajesh K. Lal, Georges Jreij et al.

Accurate characterization of carotid plaques is critical for stroke prevention in patients with carotid stenosis. We analyze 500 plaques from CREST-2, a multi-center clinical trial, to identify radiomics-based markers from B-mode ultrasound images linked with high-risk. We propose a new kernel-based additive model, combining coherence loss with group-sparse regularization for nonlinear classification. Group-wise additive effects of each feature group are visualized using partial dependence plots. Results indicate our method accurately and interpretably assesses plaques, revealing a strong association between plaque texture and clinical risk.

Thomas Hardjono, George Howard, Eric Scace et al.

One of the significant issues in the music supply chain today is the lack of consistent, complete and authoritative information or metadata regarding the creation of a given musical work. In many cases multiple entities in the music supply chain have each created their own version of the metadata for a musical work, often by manually re-entering the same information or through scraping data from other sites. In such cases, the effort to synchronize or to correct the information becomes manually laborious and error-prone. Furthermore, confidential information regarding the legal ownership of the musical work is often commingled in the same metadata, making the entire database proprietary and thus closed. In this paper we explore an alternative model for creation metadata following the open access paradigm found in other industries, such as in book publishing, library systems and in the automotive parts supply chain. The vision is to create a new music metadata layer for creation metadata that is open, scalable and provides an authoritative source of information that is available to all entities in the music supply chain globally.