Tyler O'Neal, Staff Editor GOVERNMENT

WPI wins grant to help New York City youths at risk for human trafficking

The team will use data analytics and optimization tools to identify and recommend resources

A research team led by professors at Worcester Polytechnic Institute (WPI) will use data analytics and optimization to determine the most efficient use of shelters and services for homeless youths in New York City. Their goal is to disrupt the "supply-side" of human trafficking networks by reducing the vulnerability of those most at risk of exploitation.

Renata Konrad, associate professor at the Foisie Business School at WPI, has received a $535,565 grant from the National Science Foundation's (NSF) Special Initiatives program for the three-year project. Andrew Trapp, also associate professor, is a co-principal investigator on the project. The study will build on previous research led by Konrad using analytics to develop tools to understand and address human trafficking networks.

"To disrupt human trafficking, we need to look at the beginning of the supply chain--at-risk homeless youths," Konrad said. "The question is, can we stop the trafficking process before it happens with shelters and services for homeless youths?" Renata Konrad, associate professor, Foisie Business School at WPI{module INSIDE STORY}

Konrad noted the challenge associated with estimating the number of homeless youths in New York City, and said that not all of those who are homeless will be trafficked or exploited.

The office of New York City Mayor Bill de Blasio, the mayor's Youth Homelessness Task Force, and the Coalition for the Homeless have committed to support the project.

Under the grant, the researchers will first design surveys and gather information about the numbers and needs of homeless youths ages 16 to 24 in New York City. Then the researchers will use that data to inform mathematical models regarding the prevalence of youth homelessness and use optimization to project how the capacity of shelters and services could be deployed to cost-effectively meet those needs. Finally, the researchers will recommend how best to roll out public resources.

"The models we develop can be used to optimize the benefit-cost ratio," Trapp said. "The costs related to providing food and shelter, including building shelters, as well as medical and psychological care, and employment training. And the benefits are rehabilitated lives, less time incarcerated, more productive jobs, and tax revenues going back to society because people are having more stable jobs."

Assaf Naor receives Ostrowski Prize in Higher Mathematics

Tyler O'Neal, Staff Editor GOVERNMENT

The Czech-Israeli mathematician Assaf Naor has been awarded the international Ostrowski Prize in Higher Mathematics 2019. The Ostrowski Prize is worth 100,000 Swiss Francs and named after Alexander M. Ostrowski, a professor of mathematics who taught at the University of Basel.

Assaf Naor, a professor of mathematics at Princeton University (USA), receives the Ostrowski Prize 2019 in recognition of his pioneering achievements at the interface of the geometry of Banach spaces, the structure of metric spaces and algorithms.

Since the mid-1990s, geometric methods have played an influential role in designing algorithms for computational problems that a priori have little connection to geometry. Assaf Naor is the world's leading researcher in this field, building a long-term cohesive research program. He has discovered and applied deep results from the theory of Banach spaces and quantitative metric geometry to solve long-standing algorithmic questions. Prof. Dr. Assaf Naor{module INSIDE STORY}

One particular focus of Assaf Naor's research is the optimal partition of graphs. A graph is a set of nodes together with a set of edges, which are paired connections between the nodes. In graph theory, a cut is a partition of the set of nodes of a graph. The determination of optimal cuts is an NP-complete problem. Therefore, a number of proposed heuristics exist to find an approximation of optimal cuts in a short time. Assaf Naor investigates polynomial-temporal approximation methods, which find the cut that divides a graph into two equally sized parts and thereby divides as few edges as possible (Sparsest Cut Problem).

Assaf Naor, born 1975, is a Czech-Israeli mathematician. He received his doctorate from the Hebrew University in Jerusalem in 2002, under the supervision of the Israeli mathematician Joram Lindenstrauss. After positions at Microsoft Research, the University of Washington and the Courant Institute of Mathematical Sciences, he was appointed a professor of mathematics at Princeton University in 2014.

Prize with Basel history

The Foundation A. M. Ostrowski for an international prize in higher mathematics was established by Alexander Markovich Ostrowski (1893-1986), a former professor of mathematics at the University of Basel. Since 1989, the foundation awards every other year a prize for outstanding achievements in the field of pure mathematics and in the foundations of numerical mathematics.

The jury consists of one representative of each of the following institutions: the University of Basel, the University of Jerusalem, the University of Waterloo (Canada), the Royal Netherlands Academy of Arts and Sciences, the Royal Danish Academy of Sciences and Letters. The prize is awarded irrespective of politics, nationality, religion, or age.

The Ostrowski Prize is awarded for the 16th time this year. In 2017, it was conferred to the US mathematician Akshay Venkatesh. The award ceremony will take place at the University of Basel in the coming months.

Dartmouth team develops new machine learning approach that detects esophageal cancer better than current methods

Tyler O'Neal, Staff Editor GOVERNMENT

Researchers at Dartmouth's Norris Cotton Cancer Center have created a deep learning model that can accurately identify cancerous esophagus tissue on microscopy images without the time-consuming manual data input required for current methods

Recently, deep learning methods have shown promising results for analyzing histological patterns in microscopy images. These approaches, however, require a laborious, high-cost, manual annotation process by pathologists called "region-of-interest annotations." A research team at Dartmouth and Dartmouth-Hitchcock Norris Cotton Cancer Center, led by Saeed Hassanpour, Ph.D., has addressed this shortcoming of current methods by developing a novel attention-based deep learning method that automatically learns clinically important regions on whole-slide images to classify them.

The team tested their new approach for identifying cancerous and precancerous esophagus tissue on high-resolution microscopy images without training on region-of-interest annotations. "Our new approach outperformed the current state-of-the-art approach that requires these detailed annotations for its training," concludes Hassanpour. Their results, "Detection of Cancerous and Precancerous Esophagus Tissue on Histopathology Slides Using Attention-Based Deep Neural Networks" will publish in JAMA Network Open in early November 2019. {module INSIDE STORY}

For histopathology image analysis, the manual annotation process typically outlines the regions of interest on a high-resolution whole slide image to facilitate training the computer model. "Data annotation is the most time-consuming and laborious bottleneck in developing modern deep learning methods," notes Hassanpour. "Our study shows that deep learning models for histopathology slides analysis can be trained with labels only at the tissue level, thus removing the need for high-cost data annotation and creating new opportunities for expanding the application of deep learning in digital pathology."

The team proposed the network for Barrett esophagus and esophageal adenocarcinoma detection and found that its performance exceeds that of the existing state-of-the-art method. "The result is significant because our method is based solely on tissue-level annotations, unlike existing methods that are based on manually annotated regions," says Hassanpour. He expects this model to open new avenues for applying deep learning to digital pathology. "Our method would facilitate a more extensive range of research on analyzing histopathology images that were previously not possible due to the lack of detailed annotations. Clinical deployment of such systems could assist pathologists in reading histopathology slides more accurately and efficiently, which is a critical task for the cancer diagnosis, predicting prognosis, and treatment of cancer patients."

Looking ahead, Hassanpour's team is planning to validate their model further by testing it on data from other institutions and running prospective clinical trials. They also plan to apply the proposed model to histological images of other types of tumors and lesions for which training data are scarce or bounding box annotations are not available.