Congratulations to Jocelyn Gandelman and collaborators on the recent publication!
Machine learning reveals chronic graft-versus-host disease phenotypes and stratified survival after stem cell transplant for hematologic malignancies. Gandelman JS, Byrne MT, Mistry AM, Polikowsky HG, Diggins KE, Chen H, Lee SJ, Arora M, Cutler C, Flowers ME, Pidala J, Irish JM**, Jagasia M**. Haematologica 2018 PMID: 30237265. Pubmed. DOI. **co-corresponding authors.
► This peer-reviewed research used a novel machine learning workflow to analyze semi-quantitative organ scores and reveal hidden patient subtypes within graft vs. host disease that are stratified for clinical outcome. In addition, a clinically useable decision tree was designed to capture these groups with a short yes/no questionnaire or smart phone app. This research was led by medical student Jocelyn Gandelman and conducted in collaboration with her co-mentor, Madan Jagasia. The approach was based in part on MEM and other tools created by Kirsten Diggins and colleagues (see Methods, Nature Methods, and Current Protocols articles).
Gandelman et al. Figure 1 and 2