Ohio State Navbar

Sign In

RSS Feed BMI News

Two Student submissions from The Ohio State University chosen for the 2015 AMIA Student Design Challenge 


AMIA recently invited teams of graduate students from different scientific disciplines and various backgrounds to propose creative solutions to specific healthcare problems. Student teams were asked to envison new ways for clinicians and patients to engage with data-analytics systems. Zach Abrams and Satyajeet Raje have both been chosen to present their research at the 2015 Annual Symposium in San Francisco. Only 8 teams from across the nation were selected as finalists; Congratulations Zach and Satya!

Accelerating Biomedical Informatics Research with Interactive Multidimensional Data Fusion Platforms; Satyajeet Raje, Justin Dano, Marques Mayoras, Kuhn, Bradley Myers, Tyler Kuhn

A prototype data fusion platform was presented that enables users to perform automated interpretation and ad-hoc integration of multiple datasets in real-time. The prototype platform design illustrates interactive features that are highly visual and intuitive and do not require programming expertise designed to cut down data pre-processing time significantly. The data fusion platform provides researcher with an intuitive interface to perform several preprocessing tasks. The prototype implementation allows users to automatically interpret to find “matches” across datasets as well as perform ad-hoc “joins” to integrate datasets in a real-time environment. The platform is expected to evolve into an end-to-end solution for all data preprocessing tasks.

KaryoViz: Designing A Karyotype Visualization Platform for Clinical Decision Support;  Zachary B. Abrams, Satyajeet Raje

Data visualization is a critical step in helping researchers better understand the structure and meaning of their data. In the domain of medical genomics cytogenetic data is frequently used to provide visualization at the chromosomal level. Karyotypes are syntactic descriptions of chromosomes, derived from a microscopic visual examination of chromosomes. Karyotype analyses help detect chromosomal structural and numerical defects that can serve as genetic indicators of disorders or diseases. These tests are more readily available than the more specialized sequencing technologies and are thus a more accessible means of assessing a patient's genetics for diagnosis and/or therapeutic planning purposes.

It has long been observed that  karyotype analysis can be used as a diagnostic tool in medicine since many diseases have distinguishable patterns of chromosomal aberrations1. The International System for Human Cytogenetic Nomenclature (ISCN)2 is a domain-specific language that records these chromosomal defects through visual microscopic inspection of chromosomes. However, the kayotype information in this format is difficult difficult to analyze using existing computational methods by virtue of their syntactic variability, information density and potential for human error. Thus, the karyotype data remains severly underutilized. Prior attempts have parsed ISCN karyotypes but have failed to accurately map them to a structured language biological model3.

In our previous work, we propose, 1: a Loss-Gain-Fusion (LGF) biological model (AMIA citation) that provides a standardized machine-readable representation of the karyotype information and 2: a computational cytogenetic platform that transforms ISCN-encoded karyotypes through a process of parsing and mapping into their standardized representations using the biological model. Thus, the platfrom provides an automated pipeline that can extracts biologically important information from text-based karyotype data in a form that is ready for reuse. The karyotype data can then be utilized for several clinical as well as applications such as visualization, analysis, etc.

In this paper, we highlight the data processing and visual component of this pipeline. We propose a visualizaion tool, called KaryoViz, that will primarily allow clinicians to input the syntactic representation of a karyotype and analyze it in an intuitive interface. The users would be able to navigate through diffferent granularity of the information extracted from the karyotype data. It would also allow the users to visually browse through the Mitelman database, which is the the largest public repository for karyotype data4. Thus, the tool would be instrumental in augmenting clinical decision support by reusing the karyotype data, which is otherwise be inaccessible to clinicians and researchers alike.​


Posted on 21-Sep-15 by Buster, Judith
Trackback Url  |  Link to this post | Bookmark this post with: