Beijing, China, August 28, 2025- MathWorks, a leading global developer of mathematical computing software, announced today that a research team from The Chinese University of Hong Kong (CUHK) has accelerated the workflow of biomedical image processing using MATLAB, Medical Imaging ToolboxTM, and Image Processing ToolboxTM. With the help of MathWorks software, researchers efficiently segmented and analyzed trillion voxel level images, tasks that previously required high-end computing infrastructure and extensive manual programming. This breakthrough enables researchers to process massive datasets in a very short amount of time, paving the way for real-time diagnosis and potentially significantly improving clinical decision-making capabilities.

CUHK has developed effective methods for exploring and mapping biological structures and molecular compositions. The core of this method development lies in image processing, which requires the ability to flexibly handle multidimensional images and large voxel datasets. For this purpose, a team led by Professor Li Ximing, Clinical Assistant Professor of the Department of Chemical Pathology at the Chinese University of Hong Kong School of Medicine, extensively utilized the functions in MATLAB, Medical Imaging ToolboxTM, and Image Processing ToolboxTM. The Cellpose algorithm and blockedImage function in MATLAB enable researchers to complete image processing and segmentation in a single script, thereby accelerating the segmentation and analysis of trillion voxel images. By using BlockedImage, large images can be parsed into smaller stacks, eliminating the need for expensive high-end computers and reducing programming time and errors.

A mouse whole brain dataset of approximately 800 GB was segmented using cellpose in MATLAB and scatter plots were plotted for each cell center point

We chose MATLAB because we wanted to complete the entire process in a single environment, "Professor Li said. Its programming environment provides simplified processes, comprehensive documentation, and reliable technical support, enabling us to effectively handle large datasets. BlockedImage and Cellpose in MATLAB are very compatible with our workflow, as we also want to perform image processing and segmentation simultaneously in a single script, in addition to other classic image processing algorithms. This workflow makes the previously arduous task of cell segmentation possible without requiring a significant amount of manpower. ”

In the latest study published in Nature by CUHK, they have two 3D image datasets analyzed using cellpose in MATLAB. One of them has 10 trillion voxels and 28 channels, representing approximately 1 million cells, and requires segmentation and cell typing analysis. The second dataset consists of approximately 800 GB of mouse whole brain images, which require global segmentation of neuronal bodies and registration to the Allen Brain Atlas.

Professor Li's team used cellpose in MATLAB to segment images that had undergone threshold processing and background subtraction, and obtained cell masks using blockedImage. They then analyzed the molecular expression profiles of each cell. After obtaining a 3D 28 fold image of the cell mask, the immunostaining intensity of 25 selected markers can be analyzed, which are used for cell type classification. All operations are completed in a single script in MATLAB.

CUHK hopes to further promote this technology in clinical applications, as real-time image processing will bring more efficient patient diagnosis. Dan Bo, Chief Technology Officer of MathWorks China's medical industry, said, "The processing of large images, especially high-resolution pathological micro sections, is a huge challenge for storage and computation. The solution proposed by CUHK adopts the blockedImage function for processing large images, which extracts mask ROI regions from low resolution images, greatly reducing computational complexity; Subsequently, based on the blockedImage framework for large-scale image processing, Cellpose's AI model was used for segmentation and processing. This turns the originally daunting challenge into reality. I am very honored that MATLAB can contribute to such high-level biomedical innovation research. ”

The comprehensive support provided by MathWorks' bioscience experts and academic support engineering team is crucial for the success of the project. Li Qingjie, the General Manager of Education Industry at MathWorks China, emphasized that this is a typical case of the MathWorks education team providing scientific research support to campus wide licensed partner universities, fully reflecting our five in one technical support concept in teaching and research projects in Chinese universities: 'Dao (simplifying complex problems with computational thinking), Fa (using AI for interdisciplinary data-driven paradigms), Shu (skills), Wu (toolbox and technology), and Ren (local and headquarters technical teams)'. We look forward to collaborating with more researchers in the future

About The Author

This is reported by Top Components, a leading supplier of electronic components in the semiconductor industry

They are committed to providing customers around the world with the most necessary, outdated, licensed, and hard-to-find parts.

Media Relations

Name: John Chen

Email: salesdept@topcomponents.ruThis is reported by Top Components, a leading supplier of electronic components in the semiconductor industry. They are committed to p with the most necessary, outdated, licensed, and hard-to-find parts.

Media Relations Name: John Chen

Email: salesdept@topcomponents.ru