·¬ÇÑÉçÇø

·¬ÇÑÉçÇø

²Ñ²¹°ù²âÌý³¢±ð²õ±è±ð°ù²¹²Ô³¦±ð

Mary Lesperance
Position
Professor
²Ñ²¹³Ù³ó±ð³¾²¹³Ù¾±³¦²õÌý²¹²Ô»åÌý³§³Ù²¹³Ù¾±²õ³Ù¾±³¦²õ
Credentials

Ph.D. from U of Waterloo; Professional Statistician P.Stat.

Contact
°¿´Ú´Ú¾±³¦±ð:Ìý¶Ù°Õµþ-´¡449

Dr. Mary Lesperance needs only 3 words to explain why mathematics and statistics are so important to the field of biomedical science. Evidence-based medicine.

The information acquired from clinical studies can have a profound impact on healthcare decisions and policy-making. However, the strength of all research is built upon solid study design and the proper analyses of the data. Certainly, without statistics, data is simply a set of meaningless numbers. This is where Dr. Lesperance's expertise comes into play.

Dr. Lesperance is a professor of statistics at the ·¬ÇÑÉçÇø, and a statistician whose skill lies in developing complex mathematical models and innovative statistical tools and software capable of analyzing large and complex sets of data. As the Founding Director of UVic's Statistical Consulting Centre, Lesperance is also responsible for providing both internal and external clients with statistical advice.

As for future aspirations, Lesperance hopes to see people using the statistical tools and software that she creates. "I'd like to see people using the results of my work," says Lesperance, "to make an impact in the broader community."

The driving force that keeps Lesperance plugging away at her projects might leave the more mathematically timid out there somewhat baffled.

Statistics is just fun.

Interests

  • Statistical Inference
  • Mixture models
  • Biostatistics
  • Statistical methods for genomics
  • Industrial Statistics

Courses

  • Fall 2024:
  • Spring 2025: Ìý
  • Summer 2025:

Current Projects

Semiparametric mixture models

One of Lesperance's areas of expertise lies in theoretical model development, particularly semiparametric mixture models - statistical models that allow researchers to accurately analyze data without employing assumptions that are inherent to the more traditional models. Thus, semiparametric mixture models are ideal when dealing with the complex data sets that often accompany the technological advancements that define cutting-edge science.

Transient Ischemic Attacks and Stroke

Lesperance is currently involved in a large-scale, four-year $9.6 million research project headed up by neurologist, Dr. Andrew Penn, called, Reducing stroke burden with a hospital-ready biomarker test for rapid TIA triage. The goal is to develop a rapid, diagnostic tool to assist Emergency Room physicians to differentiate between transient ischemic attacks (TIAs) - essentially mini-strokes - and other more benign conditions - such as migraines - that often present with a similar set of symptoms but dictate a much different treatment regime.

Statistical Genomics

Lesperance is also collaborating with Dr. Caren Helbing, a molecular biologist who is studying the effects of wastewater contaminants on bullfrog development. Along with the assistance of 2 graduate students, Lesperance is developing the statistical methodology that is key to accurately analyzing the study's massive sets of RNA sequencing data.

Models for palliative care

Lesperance has already demonstrated her ability to create meaning out of numbers by developing a practical, nomogram tool that is in use at hospices around the world. Based on a set of variables that include age, gender, type of disease, and functional status, the nomogram tool helps palliative care clinicians predict a patient's remaining lifespan, and ultimately, assists patients and their families with end-of-life planning and preparations.

Selected Publications

  • Lesperance M., Reed W.J., Stephens M.A., Tsao C., Wilton B.. 2016. .

    An is available with instructions on github.

  • Lesperance, M.L., Nathoo, F.S., Sabelnykova, V.Y. 2015. "A joint model for interval-censored functional decline trajectories under informative observation times," Statistics in Medicine.
  • Ichu, T.-A., Han, J., Borchers, C., Lesperance, M., and Helbing, C.C. 2014 "Metabolomic insights into system-wide coordination of vertebrate metamorphosis", BMC Developmental Biology, 14(5), electronic version is the complete one, doi:10.1186/1471-213X-14-5.
  • Lesperance, M., Saab, R., Neuhaus, J. 2014. "Nonparametric estimation of the mixing distribution in logistic regression models with random intercepts and slopes" Journal of Computational Statistics and Data Analysis, 71, 211-219.
  • Smith, S., Truong, P., Lu, L., Lesperance, M., Olivotto, I.A. 2014. "Identification of patients at very low risk of local recurrence after breast conserving surgery," International Journal of Radiation Oncology*Biology*physics, 89(3), 556-62.