Michele Guindani

How the Brain Works
Professor Guindani’s research proposes principled statistical approaches to study how brain regions activate when conducting specific tasks and how interactions among brain regions change dynamically over the course of an fMRI experiment. The brain’s ability to respond to stimuli and external events changes between individuals and explains individuals’ behaviors and decision-making processes. “My research tries to unveil the determinants of human behaviors through the analysis of brain imaging data,” he says. For example, he has studied individuals whose brains show a stronger response to food-related cues, which could contribute to our understanding of the neurobiological basis of vulnerability to obesity.

Microbiome and Global Change
“Many conditions, from obesity to anxiety, appear to be associated with the type, amount and distribution of microbes inside and outside the human body,” says Professor Guindani. “Similarly, microbial communities in our environment hold an essential role in governing the functioning of our planet.” His research interrogates data at different scales to investigate the role of microbes and their impact on different eco-systems, such as ocean temperature, agricultural outcomes and remediation of pollution. In one study, he investigated the role of fat and fiber in the association between a specific diet and the risk of colon cancer.

Medical Diagnoses
Professor Guindani develops statistical methods that help with medical diagnoses by leveraging information contained in clinical images and combining it with clinical and genomic data. “For example, the spatial distribution of enhancement levels within a lesion image can help capture the complex patterns underlying the intrinsic heterogeneity observed in tumor pathology,” he says. “Adrenal lesion diagnosis remains a challenge for radiologist.” In a recent project, he helped develop a methodology that can capture spatial patterns in patients’ scans and yielded patterns from unsupervised learning that described malignant and benign majority subtypes when combined with pathology information.

Michele Guindani​

I leverage machine learning and statistical methods to overcome challenges created by the high-dimensional, multimodal, and multiscale nature of microbiome data.”

(949) 824-5968
DBH 2232