ALERT | Awareness and Localization of Explosive-Related Threats

ALERT and Gordon-CenSSIS Scholars – Research Projects

Tunnel Detection
Prof. Carey Rappaport, Electrical and Computer Engineering

Description:  Underground tunnels present both military and homeland security threats since smugglers use them as transit routes for trafficking weapons, explosives, people, drugs, and other illicit materials. Detecting and imaging the presence of tunnels in any given region of ground is possible because the air that fills them is materially quite different from anything else underground. The Spotlight Synthetic Aperture Radar (SL-SAR) has been used due to its ability to scan large areas of terrain in a short amount of time, which is ideal for tunnel detection. In order to obtain strong and distinct target signals, Underground Focusing, based on ray refraction at the ground surface is being considered. This presents a challenge since the technique requires an estimation of the ground characteristics, and the random roughness of the soil surface tends to distort the reconstructed image of the analyzed geometry. This project will explore the impact of the ground surface roughness in Underground Focusing SAR imaging for tunnel detection applications. The study will be computational, simulating incident plane wave interactions with the ground with and without tunnels. Shallow tunnels in dry soil can be easily imaged. We continue to explore ways of imaging the deep tunnels targets in moist clay.

Advanced Imaging Technology (AIT) Experiments
Prof. Carey Rappaport, Electrical and Computer Engineering

Link: https://alert.northeastern.edu/project/r3-a1/

Link: https://alert.northeastern.edu/project/r3-A2/

Description: The Advanced Imaging Technology project began Fall, 2010 with the goal of developing an improved multi-modality portal-based passenger screening system.  Millimeter-wave, x-ray backscatter, infrared sensing, and Terahertz sensing are being considered.   Initial concentration in mm-wave imaging makes use of optimal antenna placement, model-based inversion, superior frequency specification, and custom designed radar hardware.  A specially-built hardware platform has been designed and built.  Consisting of mechanical and electrical subcomponents it will facilitate reconfigurable sensor placement in order to develop a multi-static imaging radar system.

This project will work on improving the radar hardware, computer modeling, image reconstruction algorithms and/or collection of experimental data for concealed object detection.

Radar Breast Cancer Imaging
Prof. Carey Rappaport, Electrical and Computer Engineering

Description: The breast cancer imaging project combines Microwave Nearfield Radar Imaging and Digital Breast Tomosynthesis (DBT) to develop a system that will provide an important public service to our society, addressing the most common of women’s cancers. DBT is a clinically tested, three dimensional X-ray imaging technology developed by our Massachusetts General Hospital research partners. Despite DBT’s enhanced tumor detection capability, the 1% radiological contrast between cancer tissue and healthy, commonly-occurring fibroglandular tissue limits the unequivocal characterization of tumors.

Our research will fuse DBT and microwave imaging. The contrast between fibroglandular and cancer tissue is significant – of the order of 10% — for microwave radar frequencies. However, detecting breast cancer using only microwave imaging has had poor success because the breast is often highly heterogeneous with fibroglandular tissue randomly interspersed in the adipose background. And since the microwave contrast between fibroglandular and fatty tissue is even greater than between fibroglandular and tumor tissue, the microwave scattering becomes quite cluttered, resulting in poor image target reconstruction. Our approach consists of using the high resolution X-ray-based DBT images to obtain the spatial background configuration of the breast tissues and couple it with simultaneously observed microwave measurements to perform a three dimensional reconstruction. By computationally modeling the propagation and scattering of the incident field throughout the DBT-imaged fibroglandular geometry, the problem is reduced to determining whether all regions of high dielectric contrast represent normal fibroglandular or cancer tissue.

Standoff Detection of Potential Suicide Bombers
Prof. Jose Martinez-Lorenzo, Electrical and Computer Engineering & Mechanical and Industrial Engineering

Research Webpage: https://alert.northeastern.edu/project/r3-b1/

Description: We are developing a new radar system concept capable of detecting explosive related threats at standoff distances. The system consists of a two dimensional aperture of randomly distributed transmitting/receiving antenna elements and a set of Passive Reflecting Surfaces (PRS) positioned in the vicinity of the target. A 3D imaging algorithm, based on novel compressive sensing techniques, is used in this work. Preliminary results show that images having a resolution of 7.5 mm in cross-range and 30mm in range can be achieved at 10-40m range, when the radar works at 60GHz center frequency and has 6GHz bandwidth. Research funded by the Department of Homeland Security.

Processing of Physiologic Optical Images and Signals for the Development of an Intraoperative Burn Surgery Diagnostic Device
Prof. Jose Martinez-Lorenzo, Electrical and Computer Engineering & Mechanical and Industrial Engineering

Description: As a part of this project, we are using machine learning techniques to differentiate the wound bed from the devitalized burn tissue and healthy skin using IR cameras and multi/hyper-spectral data. Additionally, we are studying signal features that will help to assess the following: 1) the severity of the burn; 2) oxygen and/or hemoglobin content of the patient’s healthy skin and wounded tissues; 3) fluid resuscitation status of the patient; and 4) cardiovascular status of the patient. Funded by SpectralMD and BARDA.

Multi-Modal Breast Cancer Detection
Prof. Jose Martinez-Lorenzo, Electrical and Computer Engineering & Mechanical and Industrial Engineering

Description: This projects is focused on addressing the problem of false positive results for breast cancer detection by supplementing conventional x-ray mammography scans with microwave technology to increase the contrast between healthy and cancerous breast tissue. In addition, our team is working to implement non-iodizing ultrasound technology to this screening system.

Parallel Computing with Graphics Processors
Prof. David Kaeli, Electrical and Computer Engineering

Research Website: http://ece.neu.edu/groups/nucar/index.html

Description: Many applications are limited due to their associated computational barriers.  Graphics processors (GPUs) have become an attractive accelerator to overcome many of these barriers.  In this project, we work with a range of applications from material science, environmental health and biomedical imaging that can significantly benefit from leveraging data-parallel GPUs.  Students will work with application domain scientists/engineers, and move their applications to either CUDA or OpenCL.

Big Data Analytics with Parallel Computers
Prof. David Kaeli, Electrical and Computer Engineering

Research Website: http://ece.neu.edu/groups/nucar/index.html

Description: Given advances in our ability to collect data about our world, we need to develop new methods for accelerating the processing of this data to better understand patterns and trends that affect our world.  Typical applications include human health informatics and human-computer interaction (e.g., Facebook traces).  In these projects we leverage cluster computing to accelerate these workloads.  We will leverage the Hadoop framework as part of this project.

Secure Computing Systems and Software
Prof. David Kaeli, Electrical and Computer Engineering

Research Website: http://ece.neu.edu/groups/nucar/index.html

Description: Cybersecurity has become a first rate design tradeoff for both computer hardware designers and software developers. Current computing systems remain vulnerable to malware attacks, exploiting poorly written software, or exploiting hardware or architectural features to gain access to sensitive information. This project will immerse students in challenges related to side-channel analysis, architectural vulnerabilities, and encryption standards. Students will work with both hardware/software simulators and actual platforms as they design attacks and develop remediation approaches to thwart attackers.

Green Remediation by Solar Energy Conversion into Electrolysis in Groundwater
Prof. Akram Alshawabkeh and Dr Lily Rajic, Civil and Environmental Engineering 

Research Website: http://www.northeastern.edu/protect/research/p5/

Description: The project’s long-term goal is to develop novel, sustainable, solar-powered and environmentally-friendly electrochemical technologies for remediation of contaminated groundwater. We develop electrochemically-induced oxidation and reduction processes to degrade individual and mixtures of contaminants and investigate the pathways of degradation and their associated effects on groundwater toxicity to ensure the efficiency of the electrochemical treatment.

Robotics and Imaging
Prof. Hanumant Singh, Electrical and Computer Engineering

Research Website: http://web.whoi.edu/singh/

Description: Students in our lab will be working on a variety of projects to do with Robotics. These include underwater robots, land robots (including an autonomous car and golf carts), and aerial robots, and will be focused on hardware, software and systems level issues.

Optical Imaging
Prof. Charles Dimarzio, Electrical and Computer Engineering

Research Website: http://www.ece.neu.edu/fac-ece/dimarzio/

In-vivo microscopy has the potential to reduce the need for painful and expensive biopsies.   Optical techniques are required to “section” the tissue so that only one layer is visible in an image.  Confocal microscopy is one technique for sectioning, and has found application in dermatology.  We are exploring another technique called structured illumination microscopy.   The student will collect images and process them to demonstrate the ability to section tissue-like phantoms.