Tairan Song

Ph.D. Student, MIDI Lab
Department of Computational Mathematics, Science & Engineering (CMSE)
Michigan State University
Advisor: Dr. Adam Alessio
Email: songtai1@msu.edu | LinkedIn | CV | GitHub

I am a Ph.D. student in the MIDI Lab at Michigan State University, advised by Dr. Adam Alessio. My research lies at the intersection of medical image analysis, statistical modeling, and deep learning, with a primary focus on CT-based kidney cancer imaging.

I develop machine learning methods that are robust to real-world clinical heterogeneity, including variability in contrast timing, acquisition protocols, and anatomical context. My work emphasizes radiomics-based modeling, multi-organ contextual feature representations, and clinically deployable pipelines for oncologic imaging and prognosis.

Selected Research

• Kidney Cancer Grade Prediction from Heterogeneous Contrast-Enhanced CT
  Proposed a radiomics-based modeling framework for renal tumor grading that is robust to highly variable contrast timing in real-world abdominal CT. By incorporating quantitative descriptors from surrounding anatomical regions (e.g., normal kidney and major vessels), the approach improves grading stability without relying on explicit phase labeling or normalization.
• De Novo Identification of Expressed Somatic Mutations from Single-Cell RNA Sequencing
  Contributed to the development of the RESA computational framework for high-precision identification of expressed somatic mutations from scRNA-seq data. The work integrates statistical modeling and empirical validation to improve mutation detection sensitivity and precision across diverse datasets.

Publications & Preprints

Song, T., Noyes, S. L., Muterspaugh, R., Munavar Ali, M. A., Zhou, S., Robert, C., Lane, B. R., Lim, E., Alessio, A. (2025). Predicting Kidney Cancer Grade from Highly Variable Contrast Timing CT. Submitted to SIIM 2026. Under review.

Zhang, T., Jia, H., Song, T., Lv, L., Gulhan, D. C., Wang, H., … & Shen, N. (2023). De novo identification of expressed cancer somatic mutations from single-cell RNA sequencing data. Genome Medicine, 15(1), 115.

Awards

NSF NRT-IMPACTS Fellowship – Michigan State University (2023). Funded by the National Science Foundation Research Traineeship Program (DGE-1828149).

Teaching

Teaching Assistant

• CMSE 202: Computational Modeling and Data Analysis II (Spring 2025)

Contact

songtai1@msu.edu
tairan_song@outlook.com
Michigan State University, East Lansing, MI, USA

A complete version of my curriculum vitae can be downloaded here (PDF).