Six graduate students have been awarded fellowships from McMaster’s MacData Institute to pursue interdisciplinary research projects that use data science approaches to help shed light on a number of areas of research.

From harnessing data to inform earlier diagnosis of brain tumours, to analysing data to better understand infectious disease epidemics, the fellows* will be working with McMaster faculty on projects that apply data analysis, collection and curation methodologies to a number of areas of research.

“Over the past decade, there has been an explosion in different types of data and in the amount of data available to researchers,” says Paul McNicholas, Director of the MacData Institute and a Canada Research Chair in Computational Statistics, who was recently named to the Royal Society of Canada.

“This fellowship program is bringing together talented graduate students with expertise in data science and McMaster faculty from a range of disciplines to work on projects that use different types of data in novel ways and which will help lead to new insights or innovations in a number of research areas.”

The program, which provides funding for fellows to work on a multidisciplinary project involving two or more Faculties, is intended to provide students with practical experience, and to facilitate an exchange of knowledge in data science across Faculties.

The students will be co-supervised by an expert in data science as well as by a faculty member from another field of research.

The fellowship program is part of the MacData Institute’s mission to promote innovative research and training programs related to data science and to work with and support Faculties and institutes on data issues and initiatives.

A call for membership of the MacData Institute was circulated in December. The call is open to all faculty and librarians, and will remain open until the end of February. Interested faculty and librarians can join here.

New MacData Fellows projects include:

Identifying and treating brain tumors in adults

Working under the supervision of McNicholas and McMaster’s Sheila Singh, of the McMaster Stem Cell and Cancer Research Institute, Tyler Roick, a PhD student in the School of Computational Science and Engineering, will apply advanced computational statistics techniques to identify sub-types of Glioblastoma – the most common brain tumor in adults. Roick will then use this information to help develop a predictive model to aid in earlier diagnosis of the disease and allow for more beneficial treatment.

Improving the interpretation of medical images for better diagnosis

Medical images are often very difficult for physicians to interpret, leading to a diagnostic error rate that is estimated to be as high as 30%. Lauren Smail, a graduate student in McMaster’s Department of Psychology, Neuroscience and Behaviour (PNB), will use machine learning to classify thousands of ultrasound images, and conduct qualitative interviews with physicians to understand how doctors read and interpret medical images. Smail will use this information to develop an education toolkit to teach medical trainees learn how to read scans. Smail will be supervised by Ranil Sonnandara from McMaster’s Department of Surgery and Sue Becker from the Department of PNB.

Finding ways to incorporate new kinds of data into health research studies

New data collection technologies like electronic sensors and cell phones have created many rich sources of data that could be incorporated into existing health research data sets to provide new insights.  Peter Tait, a graduate student in the School of Computational Science and Engineering, will develop a methodology to analyze incomplete data from accelerometers – a tool used to assess children’s physical activity – into the Health Outcomes and Physical Activity in Preschoolers (HOPP) study. This methodology will be implemented in easy-to-use software that can then be used directly by clinicians and scientists. Tait will work under the supervision of both McNicholas and Joyce Obeid from the Department of Pediatrics.

Gaining insight into brain development of infants

Infants produce “general movements” and “spontaneous movements” which when evaluated by a clinician, can lead to early insight into the neurodevelopmental health of the infants and can be used to predict Cerebral Palsy and potentially Autism Spectrum Disorders. Omar Nassif, a graduate student in Electrical and Computer Engineering will use machine learning to dig deeper into the data on infant movements to gain physiological insights and develop an easily accessible method of infant evaluation for clinicians. Nassif will work under the supervision of Vickie Galea of McMaster’s School of Rehabilitation Science and Jim Reilly from the Department of Electrical & Computer Engineering.

Predicting emergency room visits and hospitalization among older adults

InterRAI is an assessment tool used by clinicians in health care settings in over 20 countries, providing a range of detailed patient information on disease, function and care characteristics and forming the basis for electronic medical records in home care and long-term care across Canada. Angelina Pesevski, a graduate student is the School of Computational Science and Engineering, will work with McNicholas and McMaster’s Andrew Costa, an InterRAI fellow and member of the Network of Excellence in Acute Care to compare the performance of conventional statistical approaches and modern methods such as machine learning in predicting emergency visits and hospitalization among older adults based on InterRAI data.

Better understanding the origins, transmission and dynamics of infectious diseases

 Recent advances in genome sequencing technology have led to the creation of large amounts of genetic data which has had an overwhelmingly positive effect on disease research, but also poses challenges in terms of accessing, integrating and analyzing this avalanche of data. Working with Hendrik Poinar from McMaster’s Ancient DNA Centre and Brian Golding from the Department of Biology, Katherine Eaton, a PhD student in the Department of Anthropology, will develop a computational framework to mine repositories for infectious disease sequencing projects – specifically sequencing projects of the plague bacterium, Yersinia pestis –  and transform raw sequencing data into completed genomes for analysis. This framework could greatly assist with characterizing the origins, modes of transmission and dynamics of infectious disease epidemics.

*The MacData Institute Graduate Fellows are pictured above. From left to right: Katherine Eaton, Peter Tait, Lauren Smail, Tyler Roick, Angelina Pesevski and Omar Nassif.