This collaborative research project led by the colleagues at the University of Delaware aims to develop and advance tools for using image-data appearing in scientific publications, in addition to text, in order to support beneficial, targeted access to the biomedical literature. 
UIC's principal investigator, Computer Science associate professor, G. Elisabeta Marai will be responsible for the image analysis algorithms to distinguish between gel and plate images, the image analysis algorithms to distinguish between Electronic Microscopy vs Fluorescent Microscopy vs Light Microscopy and X-Rays, and the design of the visual interface to present results to end users. She will lead the design, development and implementation of the feature extraction and selection tools and techniques related to this research project.

Incorporating Image-based Features into Biomedical Document Classification

Cancer treatment courses which rely on imaging and spatially-dependent therapy involve making multiple treatment decisions (e.g., radiotherapy alone, radiotherapy plus chemotherapy, induction chemotherapy) over time. These decisions depend on complex factors, including the tumor location with respect to sensitive organs and its response to treatment, laboratory data, toxicity, anticipated side effects and survival probability. In this project, we design and develop novel statistical methodology for dynamic and personalized treatment decisions with specific application to head and neck cancer radiotherapy planning. We extend the state of the art by developing methods which take into account both imaging data features (such as radiation dosage and spatial distribution) and nonspatial data (such as demographics and toxicity) in the development of predictive models for anticipated outcomes and of the optimal treatment rules. We introduce novel similarity metrics over these hybrid, high-dimensional and sometimes incomplete patient data, and we extract informative, relevant and non-redundant features to identify similar patients. We further combine statistical models for outcome prediction with local models built over cohorts of patients. These ensemble models are used to further improve the reinforcement learning process to account for patient-specific preferences regarding side effects and survival outcomes. Specifically, we extend q-learning, a type of reinforcement learning, to account for competing, multiple outcomes and different preferences on the probability of short versus long-term outcomes. We validate our precision-driven model against known toxicity outcomes using a retrospective head-and-neck cancer dataset. 
Our specific aims are to: 
- Create quantitative methodology for extracting and selecting spatial relationships from medical images as well as non-imaging features, and generate a hybrid similarity metric which can identify similar cohorts of patients. 
- Create an ensemble model that combines global static models with dynamic local models over similar cohorts of patients with locally-sparse data. 
- Create dynamic reinforcement learning methodology which uses spatial and non-imaging features and metrics and accounts for patient preferences of multiple treatment outcomes. 
The project defines novel reinforcement learning techniques which account for multidimensional outcomes and patient preferences. Previous examples of reinforcement learning in the statistical literature have utilized examples with a small number of well-defined features, which is not applicable in the case of head and neck cancer. This project creates novel hybrid techniques for extracting informative and relevant features and similarity from imaging and nonimaging data. The project further introduces novel predictive modeling using a combination of global and dynamic local models over multiple outcomes. The proposed techniques leverage Big Data that is not only heterogeneous but also high-dimensional and locally sparse, in order to derive a novel precision-medicine approach. 
The empirically-derived treatment rules developed in this project have the potential to improve the standard of care (i.e., treatment plans chosen by the tumor treatment board) and the quality of life of surviving patients. The methods developed in the proposal may be used to derive optimal treatment strategies across not only a variety of spatially-dependent cancer diagnoses, but also other chronic conditions including mental health disorders, substance abuse diseases, or diabetes, that require making multiple decisions that must weigh the tradeoffs between efficacy and toxicity.

QuBBD: Precision E-Radiomics for Dynamic Big Head & Neck Cancer Data

Deep learning is often referred to as a &lsquo;black box&rsquo;, because, aside from selecting model and the model hyperparameters (like the learning rate, activation function parameters), users don&rsquo;t have access to what happens inside the model. We can know that the model works by measuring its accuracy on some testing data set, suggesting that the model must have &lsquo;learned&rsquo; something meaningful from the training data. But how can we peek inside and &lsquo;see&rsquo; what the model &lsquo;sees&rsquo;? 
Biologists have recently begun to explore how deep learning approaches could be used to address questions involving diverse biological datasets. For this community, producing an accurate model is important, but also insufficient, since the larger goal is to contribute to our understanding of complex biological systems and phenomena. Visualization could help biology researchers &lsquo;peak&rsquo; inside the black box and relate deep learning mechanics to existing biological knowledge. However, most visualization techniques for deep learning have been developed for images and text, so many of the encodings and layouts are designed for input data with fundamentally different perceptual properties than biological data sets, such as gene sequence data or gene expression data. 
This research explores current approaches in visualization for deep learning and how these approaches may (or may not) be applied to biological data. This work is conducted with researchers at Argonne National Labs studying deep learning for cancer research and genomics.

Argonne Leadership Computing Facility

Vis for Deep Learning for Biology

Exploring big cities like Chicago can be a daunting experience when the tourist has to gather information from multiple sources regarding places, directions, transportation, safety, and weather.  We propose a large display solution (kiosks) which aims to encourage people to explore the city by showing cultural attractions, special events and other points of interest (POI) on a map along with multi-modal transportation options to take them to their destination. POI&rsquo;s and events are updated frequently based on the time and safety information. Kiosks are intended to be located throughout the city especially locations with the highest density of points of interest.

Screen shot of the &ldquo;Exploring Chicago&rdquo; application prototype.

Exploring Chicago

The Englewood Data Hub (EDH) is a project that brings together both crowdsourced and public data surrounding the Englewood area into a single place. 
The <a href="https://englewooddatahub.org/">EDH Resource Directory</a> makes use of the crowdsourced service and school data to display an interactive map with their locations. Its user interface (UI) allows for an exploratory experience that offers filtering of services by names and type and displays detailed information about the services available in Englewood. 
The <a href="https://englewooddatahub.org/analytics">EDH Analytics Tool</a> combines the crowdsourced service and school data with public data sets to display visualizations; specifically, the public datasets used are 2010 Census data along with Vacant Lot and Crime data from the Chicago Data Portal. 
The Englewood Data Hub has been developed by a group of researchers at the Electronic Visualization Laboratory, University of Illinois at Chicago under the direction of Prof. G. Elisabeta Marai. 
Team members include EVL research experience undergraduate students (REUs): Andrew Burks, Joshua Castor, and Isabel Lindmae. 
This project is funded by the The Joseph and Bessie Feinberg Foundation and the University of Illinois at Chicago.

Englewood Social Service Finder and Englewood Analytics

Head and neck squamous carcinomas affect all ethnicities and age groups, accounting for significant mortality and therapy-related
side effects. Over 50,000 new cases are diagnosed each year in the United States, leading to large, rich repositories of patient data. For each of these cases, oncologists need to anticipate survival, oncologic, and side effect outcomes associated with treatment strategies in order to select a treatment which balances efficacy and toxicity. However, despite the wealth of data available, risk prediction algorithms for cancers are rudimentary and incorporate only a handful of patient characteristics, largely due to a lack of computational methodologies. 
This project will develop a computational methodology that supports the construction of a validated precision model of patient-specific outcomes for head and neck cancers, based on demographics, toxicity, and complex imaging data. The methodology will be the first in the field to include both complex imaging and non-imaging data, while taking into account large-scale biological and clinical correlates. A web-based environment will further allow practitioners to interactively explore and communicate the model mechanics and outcomes to stakeholders. The approach is innovative through its leverage of big imaging data repositories and through its unique blend of computational modeling principles from bioengineering, statistics, and computer science (medical imaging and visualization). 
From a clinical perspective, this integrative approach is novel in the field of cancer therapy. The resulting model and web-based
environment will mark a significant advance in biomedical computing because it will be able to identify, for the first time, specific subgroups who are at risk for distinct oncologic, toxicity, and survival profiles. Furthermore, the currently clinical practice paradigm for head and neck treatment is stage-driven. With validation, this novel approach can identify patients based not on clinical staging, but on precision modeling using cohort data and similar sets of patients. The methodology is applicable to clinical practice involving human subjects and has the potential to change the standard of care and outcomes of treatment in the field.

SMART-ACT: Spatial Methodologic Approaches for Risk Assessment and Therapeutic Adaptation in Cancer Treatment

The <a href="http://hci.cs.uic.edu/">Human-Computer Interaction (HCI) Laboratory</a> at the University of Illinois at Chicago (UIC) is led by assistant professor Debaleena Chattopadhyay of the Department of Computer Science. The HCI Lab designs computer technologies to transform human experiences and deliver positively impactful societal change. 
UIC HCI is an affiliate laboratory of the Electronic Visualization Laboratory (EVL) providing a human centered approach to the development of tools and technologies related to advanced visualization, virtual reality, visual analytics and networked collaboration. 
The HCI Lab&rsquo;s mission is to support individual and social activities and improve everyday life, designing, building and evaluating ubiquitous computing technologies. The design of these systems draws from research on how people work, play, and communicate in day-to-day life. 
The UIC HCI Lab is located on the UIC campus at: 
Science & Engineering Laboratory West Room 1032 
900 West Taylor Street 
Chicago, IL 60607