Challenges. Tumor spheroids (and organoids) have become an instrumental tool in cancer research. These self-organized, three-dimensional (3D) systems can recapitulate phenotypic and functional traits of patient tumors in vivo, thereby serving as a powerful testing bed to study tumor heterogeneity, interactions with the environment (e.g., extracellular matrix), and responses to external stimuli (e.g., chemotherapy, radiation). Fully harnessing spheroids' utility, however, is stymied by lack of high-throughput analysis methods. Conventional bright-field microscopy, although widely used to monitor spheroids in culture, fails to capture detailed cellular organizations; advanced fluorescent microscopy can resolve individual cells, but its imaging throughput is restricted by the small field-of-view (FOV) and the scanning mechanisms involved. Innovations. We aim to advance a new volumetric imaging microscope (VIM) for single cell analyses in tumor spheroids. Specifically, we will explore integrating Fourier ptychographic microscopy (FPM) with diffraction tomography. FPM is based on a spatially coded-illumination technique, collecting low resolution image sequences while changing the position of a point-light source. These images are then numerically combined in the Fourier space, which allows FPM to achieve both wide field-of-view and high spatial resolution in 2D images. We reason that full 3D microscopic images can be recovered by accounting for optical diffraction during the numerical reconstruction. Approaches. Aim 1. System development. We will build a VIM system featuring: i) a new numerical algorithm to reconstruct 3D volumetric images; ii) a new light-illumination strategy to speed up the data acquisition; iii) microfluidic cartridges optimized for spheroid culture and drug treatment; and iv) multicolor imaging capacity for molecular detection. The complete VIM will resolve individual cells constituting a spheroid at high resolution (lateral, 0.4 µm; axial, 1 µm) in a large imaging volume. Aim 2. Treatment monitoring with tumor spheroids. We will test VIM's practical utility: VIM-enabled spheroid imaging will reveal earlier than bulk imaging whether a spheroid is responsive or resistance to drug treatment. To generate a tumor model, we will use primary GBM cells from patients. GBM spheroids will be grown and treated with drug (temozolomide) inside microfluidic cartridges. We will use the VIM to monitor how single cells change their phenotypes under treatment, and correlate these changes with treatment outcomes. Impact. The VIM will be a transformative tool for cancer research, empowering researchers with rich data sets and substantially advanced analytics. Immediate applications include better monitoring of anticancer drug responses in 3D cell culture, analyzing cellular heterogeneity, and prospectively detecting cellular fate under various physiological conditions. These outcomes will strengthen the clinical and scientific utility of tumor spheroids in cancer research.

We will advance a new, volumetric imaging microscope (VIM) for high-resolution, 3-dimensional (3D) tomography. Specifically, we will integrate computational optics, microfluidics, and machine learning approaches; it will make the VIM capable of resolving a large number of individual cells in 3D culture. We will use the VIM to analyze how single cells in a tumor spheroid change their phenotypes under drug treatment.