>>>

I'm Colin Zhang, a process engineer at TSMC Kumamoto working on etching processes for semiconductor manufacturing. I received my B.S. in Mechanical Engineering from UC Santa Barbara in 2020, my M.E. in Chemical Engineering from The University of Osaka in 2023, and my Ph.D. in Engineering from The University of Osaka in 2026. My doctoral research focused on additive manufacturing, numerical simulation, and machine learning, with particular emphasis on 3D bioprinting.

Throughout my academic journey, I developed expertise in Python and machine learning. My doctoral research leveraged machine learning to optimize 3D bioprinting processes. This work involved predicting cell viability affected by the extrusion process in 3D bioprinting through numerical simulations, support vector regression (SVR), and multilayer perceptron (MLP) models. Another focus area was to develop a novel printing performance evaluation framework that combines the Hausdorff distance (HD) with convolutional neural networks (CNNs) to assess geometric fidelity automatically. Additionally, generative AI models such as conditional variational autoencoders (CVAEs) were explored to generate printing outcome predictions based on ink rheological properties and printing parameters.

Outside work, you'll often find me planning my next mountain climbing adventure or exploring new places in Japan. I'm constantly curious about learning new things, whether through watching online courses or getting lost in a good book on a quiet evening. These pursuits keep me balanced and bring fresh perspectives to both my work and everyday life.

Doctoral Research

3D bioprinting is advancing rapidly in the field of tissue engineering, enabling the creation of complex, biologically engineered structures with high precision. Key challenges in the process include maintaining cell viability during extrusion, achieving high shape fidelity, and selecting suitable ink formulations and printing parameters. My doctoral research addressed these challenges by combining ink rheology, numerical simulation, and machine learning. The work progressed from predicting as-extruded cell viability, to evaluating printability automatically with the Hausdorff distance and convolutional neural networks, and finally to generative AI that predicts construct morphology and optimizes ink rheology in silico. This combination of computational methods and experimental validation reduces trial-and-error experiments and offers a promising path toward more effective bioprinting solutions for tissue regeneration and therapeutic applications.

GraphicalAbstract1

Figure 1. Graphical workflow for bioprinting optimization: Rheological testing provides ink properties for numerical simulation of needle extrusion, followed by live/dead cell viability analysis. A machine learning framework integrates cell type, ink viscosity, and extrusion parameters to predict and optimize cell viability outcomes. Created with BioRender.com. Reproduced from Zhang, C. et al., Virtual Phys. Prototyp. 19, e2400330 (2024), under CC BY-NC 4.0.

GraphicalAbstract2

Figure 2. Overview of the AI-powered printing performance (printability) evaluation framework. Rheological properties of alginate (ALG) and hyaluronate (HA) composite inks are modeled by support vector regression (SVR) to predict rheological parameters. Regression-based MLP models are employed to predict the Hausdorff distance (HD) and cell viability. The HD metric quantitatively assesses the geometric fidelity between printed and designed structures, while a regression-based CNN directly predicts HD values from grid images. Created with BioRender.com. Derived from author-prepared materials used in the pre-acceptance manuscript version of Zhang, C. et al., Biofabrication 18, 015015 (2026).

GraphicalAbstract3

Figure 3. Overview of the β-CVAE framework for optimizing deposition morphology in photo-induced crosslinking 3D bioprinting. (a) Construct images acquired under varying irradiation intensity, crosslinker content, and rheology are encoded into a disentangled latent space z conditioned on rheological and printing parameters c for conditional construct image generations, with transfer learning extending to additional blueprint designs. (b) In silico closed-loop optimization in which the decoder generates candidate constructs, a CNN evaluates shape fidelity (printability) against the reference blueprint, and Bayesian optimization iteratively updates the rheological inputs (η0, m, n, tan(δ)) based on SVR rheological models until convergence. Adapted from Zhang, C. et al., Virtual Phys. Prototyp. 21, e2671497 (2026), under CC BY 4.0.

Publications

  1. Colin Zhang, Kelum Elvitigala, and Shinji Sakai. Generative AI-guided in silico closed-loop optimisation of deposition morphology for 3D bioprinting applications. Virtual and Physical Prototyping 21, e2671497 (2026).
    https://doi.org/10.1080/17452759.2026.2671497. Open Access
  2. Colin Zhang, Kelum Elvitigala, and Shinji Sakai. AI-powered printability evaluation framework for 3D bioprinting using Hausdorff distance metrics. Biofabrication 18, 015015 (2026).
    https://doi.org/10.1088/1758-5090/ae288c. Subscription Access [Publisher Preview] [Accepted Manuscript]
  3. Colin Zhang, Kelum Elvitigala, Wildan Mubarok, Yasunori Okano, and Shinji Sakai. Machine learning-based prediction and optimisation framework for as-extruded cell viability in extrusion-based 3D bioprinting. Virtual and Physical Prototyping 19, e2400330 (2024).
    https://doi.org/10.1080/17452759.2024.2400330. Open Access
  4. Mitsuyuki Hidaka, Masaru Kojima, Colin Zhang, Yasunori Okano, and Shinji Sakai. Experimental and numerical approaches for optimizing conjunction area design to enhance switching efficiency in single-nozzle multi-ink bioprinting systems. International Journal of Bioprinting 10, 4091 (2024).
    https://doi.org/10.36922/ijb.4091. Open Access
  5. Wildan Mubarok, Colin Zhang, and Shinji Sakai. 3D Bioprinting of Sugar Beet Pectin through Horseradish Peroxidase-Catalyzed Cross-Linking. ACS Applied Bio Materials 7, 3506-3514 (2024).
    https://doi.org/10.1021/acsabm.4c00418. Subscription Access

Conferences

  1. The 57th Autumn Meeting of the Society of Chemical Engineers, Japan, Hiroshima University, September 13-16, 2026
    Presentation Title: Rheology-Informed Generative AI for In Silico Optimization of Morphology Fidelity in Extrusion-Based 3D Bioprinting
    [Excellent Presentation Award]
    Colin Zhang, Kelum Elvitigala, and Shinji Sakai
    Excellent Presentation Award certificate from the SCEJ 57th Autumn Meeting
  2. The 90th Annual Meeting of the Society of Chemical Engineers, Japan, Tokyo University of Science, March 11-14, 2025
    Presentation Title: AI-Driven 3D Bioprinting System for Optimizing Printing Performance
    Colin Zhang, Kelum Elvitigala, and Shinji Sakai
  3. The 55th Autumn Meeting of the Society of Chemical Engineers, Japan, Hokkaido University, September 10-13, 2024
    Presentation Title: Predictive modeling of cell viability in extrusion-based 3D bioprinting using machine learning [Featured Presentation (22 out of 1130)]
    Colin Zhang, Yasunori Okano, and Shinji Sakai
  4. The 53rd Autumn Meeting of the Society of Chemical Engineers, Japan, Shinshu University, September 14-16, 2022
    Presentation Title: A Numerical Investigation of Stresses, Printing Efficiency, Printability, and Cell Viability in Nozzle Printheads for 3D Extrusion Bioprinting
    Colin Zhang, Mitsuyuki Hidaka, Masaru Kojima, Shinji Sakai, and Yasunori Okano

Course Certificates

Contact Information

You can reach me by executing the following Python script, which uses ASCII-to-string conversion:

"".join(chr(i) for i in (107, 111, 114, 105, 110, 46, 106, 97, 112, 97, 110, 64, 103, 109, 97, 105, 108, 46, 99, 111, 109))

Or, the JavaScript version which can be run directly in your browser's console:

String.fromCharCode(107, 111, 114, 105, 110, 46, 106, 97, 112, 97, 110, 64, 103, 109, 97, 105, 108, 46, 99, 111, 109);