Software design and integration for a dual-modality OC-PAM imaging device

Today one of the main open challenges in cancer treatment is the presence and growth of drug-tolerant persister cells. Innovative studies across multiple scientific fields, such as medicine, biology and biomedical engineering, are being conducted to deepen our understanding of these cells. This thesis presents a contribution to the development of a dual-modality imaging system within the context of the REAP European project, which aims to utilize multiple contrast-enhanced optical techniques to detect and characterize rare cancer cells in therapy-resistant breast cancer models using genetically engineered organoid-based murine mammary tumors. Specifically, this work focuses on the design and implementation of the software architecture required to acquire images using photoacoustic microscopy (PAM) and optical coherence microscopy (OCM) modalities with micrometric resolution. Achieving this result involved integrating multiple software packages and hardware components provided by project partners, ensuring precise system synchronization, and optimizing image quality. For this purpose, an innovative microring resonator (MRR) was employed as the sensor for PAM, necessitating a tunable laser and photodiode for signal readout. For OCM, which operates in the spectral domain to maximize acquisition speed, a spectrometer captures interference fringes generated by the combination of light from the sample and reference arms. A field programmable gate array (FPGA) controls the scanning pattern of the galvanometers and triggers all components. Python was chosen as the primary programming language to ensure cross-platform compatibility without specific licenses and to maintain readable, modifiable code. The software architecture features a graphical user interface (GUI) for easy selection of acquisition parameters and a set of functions leveraging various hardware APIs to operate the system. Both modalities include preview options to verify the correct sample positioning. Moreover, custom-made scripts reconstruct and process the images, applying averaging for PAM and k-linearization for OCM as well as other de-noising techniques, using GPU for accelerated parallel processing. At last, integration with 3D Slicer allows for visualization of acquired images and volumetric data. Multiple tests were done to validate the software by performing mock acquisitions and measuring the inputs and outputs of the different components. Preliminary imaging sessions were conducted to evaluate image quality and overall system usability for its intended purpose. Software performances for scanning routines were evaluated in terms of speed, guaranteeing an acquisition rate of 30 kHz, which is the maximum possible within the limits of the hardware. The processing pipeline for OCM was also validated on data from the first system's acquisitions and external datasets, demonstrating the software capability to reconstruct, visualize and save full volumes in less than 2 minutes. In the future, this system will enable fast and advanced studies, unlocking new possibilities such as real-time visualization and analysis of persister cells during in vitro investigations.