The Department of Imaging Science (DISI) and Innovation dissolves the traditionally “siloed” barriers between innovation, research and clinicians and has dual responsibilities to pursue both clinical innovation and externally funded research. Capitalizing on Geisinger’s extensive imaging, phenomic and genomic datasets, DISI aims to become a national leader in interdisciplinary innovation and research initiatives related to medical imaging.
Brandon K. Fornwalt, MD, PhD is the founding chair of the Department of Imaging Science and Innovation at Geisinger Health System. Dr. Fornwalt grew up in South Carolina where he attended the University of South Carolina as an undergraduate in mathematics and marine science. He then worked in a Free Medical Clinic for a year before starting an MD/PhD program at Emory and Georgia Tech in Atlanta, GA. After finishing his degrees in 2010, he spent a year doing his internship in pediatrics at Children's Hospital Boston before joining the faculty at the University of Kentucky. After four years on faculty at the University of Kentucky, Dr. Fornwalt moved to Geisinger. Dr. Fornwalt was awarded the NIH Director’s Early Independence Award and has several completed and ongoing funded projects in cardiovascular imaging and heart disease.
We are currently recruiting faculty at all levels (assistant, associate and full professor) for the Department of Imaging Science & Innovation. Apply here.
Haggerty, Christopher PhD
Jing, Linyuan PhD
Suever, Jonathan PhD
Fielden, Samuel PhD
Arbabshirani, Mohammad PhD
Hefazi-Torghabeh, Mehyar MD
Samad, Manar PhD
Hamlet, Sean MS
Wehner, Greg BS
Nevius, Chris BS
Pulenthiran, Arichanah BS
VanMaanen, David MS
Gange, Jennifer – Sr. Administrative Assistant to the department
Haggerty, Allyson MBA – Research Operations and Administration
1. Quantifying how the heart deforms during contraction, and understanding what that can tell us about patients with heart disease.
How efficiently the heart is pumping blood is one of the best known markers of heart health. With cardiac magnetic resonance imaging (MRI), there are many sophisticated ways of measuring this heart function in great detail. For example, DISI researchers specialize in displacement sensitive MRI, known as “DENSE”, which can quantitatively measure the motion of heart tissue with very high resolution. By collecting data from medical imaging, and linking these images to outcomes data from Geisinger’s EHR, they hope to learn how this detailed motion data can help physicians make better predictions about heart health.
2. Detection of sub-clinical heart disease for patients identified with rare genetic variants associated with inherited heart conditions.
Geisinger is among the leaders in “GenomeFIRST” medicine—using information for a person’s genetic sequence to clinically screen for diseases such as cancer and heart disease. With many inherited heart diseases, not everyone with these genetic risk factors will develop disease, but it is often difficult to distinguish these “healthy” people from those who will develop heart disease, but do not yet have symptoms. Our teams are collaborating with researchers and physicians in Geisinger’s Clinical Genomics Department to develop novel advanced imaging approaches that can provide important insights to these challenging clinical evaluations.
3. Identifying early signs of heart disease in obese children.
Obesity is a major health problem around the world, and Pennsylvania is no exception. The negative consequences of obesity for heart health are well known, but increasingly these problems are being found to originate with childhood obesity. DISI researchers are working with children from our local communities to measure early signs of heart disease with childhood obesity and identify opportunities and strategies to prevent and/or reverse such changes. Importantly, this work was selected for a national press release at the American Heart Association scientific sessions in 2015, where the article received coverage by over 50 news outlets including Time, >The Washington Post, The Chicago Tribune and NBC News.
4. Accurate Segmentation of Lung Fields on Chest Radiographs using Deep Convolutional Networks
Accurate segmentation of lung fields on chest radiographs is the primary step for computer-aided detection of various conditions such as lung cancer and tuberculosis. The size, shape and texture of lung fields are key parameters for chest X-ray (CXR) based lung disease diagnosis in which the lung field segmentation is a significant primary step. Although many methods have been proposed for this problem, lung field segmentation remains a challenge. In recent years, deep learning has shown state of the art performance in many visual tasks such as object detection, image classification and semantic image segmentation. In this study, we propose a deep convolutional neural network (CNN) framework for segmentation of lung fields. The suggested framework outperforms state of the art registration-based segmentation by a significant margin. To our knowledge, this is the first deep learning based study of lung field segmentation on CXR images developed on a heterogeneous clinical dataset.
5. Unsupervised Quantification of Abdominal Fat from CT images using Greedy Snakes
Adipose tissue has been associated with adverse consequences of obesity. Total adipose tissue (TAT) is divided into subcutaneous adipose tissue (SAT) and visceral adipose tissue (VAT). Intra-abdominal fat (VAT), located inside the abdominal cavity, is a major factor for the classic obesity related pathologies. Since direct measurement of visceral and subcutaneous fat is not trivial, substitute metrics like waist circumference (WC) and body mass index (BMI) are used in clinical settings to quantify obesity. Abdominal fat can be assessed effectively using CT or MRI, but manual fat segmentation is rather subjective and time-consuming. Hence, an automatic and accurate quantification tool for abdominal fat is needed. The goal of this study is to extract TAT, VAT and SAT from abdominal CT in a fully automated, unsupervised fashion using energy minimization techniques. To our knowledge, this is the first study of its kind on such a large and diverse clinical dataset. Our algorithm obtained a 3.4% error for VAT segmentation compared to manual segmentation. These personalized and accurate measurements of fat can complement traditional population health driven obesity metrics such as BMI and WC.