Supervisor: Prof. Alexander Effland, Prof. Marieta Toma

• Obtain each cell class based on fluorescent microscopic pathology slides.
• Compute spatial proximity between cells using transformed distance metrics.
• Identify and count neighboring cell-type pairs within a dynamic distance threshold to study spatial interactions across the tissue microenvironment.

Supervisor: Prof. Alexander Effland

• The model’s labels are only the proportions of the pattern classes of the global image
• Did weakly supervised semantic segmentation based on test-time augmentation for neural networks
• Proposed an energy-based optimization model to refine the segmentation masks

Supervisor: Dr. Guangjian Liu

• Developed a fully automated process for blood cell segmentation and classification
• Calculated and measured morphological characteristics of blood cells
• Explored morphological differences between cancer cells and normal cells, based on large image data

Supervisor: Prof. Haifeng Zhao, Ling Huang

• Extracted rock outlines from sample photos from Apollo 14
• Wrote automatic scripts to simulate subsurface layered structuret, and obtained ground penetrating radar signals
• Reconstructed the upper surface, average error of reconstructed model was wihtin 5 cm

Supervisor: Prof. Haifeng Zhao, Liang Jiang, Wenwang Huang

• Calculated microstructural and distribution state of each γ′ precipitates
• Predicted mechanical properties from microstructural features, R2 values reached at 0.89
• Made Influence of microstructural features on mechanical properties transparent