Project Summary / Abstract This project aims to improve needle placement accuracy for image-guided prostate interventions, including biopsy and focal treatment. Building upon the success of the previous cycle, the project seeks to enhance the robustness and effectiveness of the technology in complex anatomical structures, thereby aiding clinical translation for prostate cancer care and broader applications. Percutaneous needle placement is a critical procedure in both the diagnosis and treatment of prostate cancer. Although these procedures are often assisted by an external needle-guiding device for accuracy, the unpredictability of needle deflection due to interactions with varying tissue densities frequently necessitates multiple placement attempts. This prolongs the procedure time and can lead to excessive tissue damage. The project’s previous cycle aimed to tackle this problem by developing two technologies: a fiber-Bragg-grating (FBG)-based shape-sensing needle (sensorized needle) for real-time feedback and a data-driven needle steering algorithm (COADAP) for active compensation of needle deflection. Together, these technologies created a closed-loop adaptive needle placement system, enhancing needle placement accuracy. However, the team identified these technologies’ limitations when the needle encountered complex, interconnected multistructural anatomy. In such situations, the needle’s interactions with different structures led to significant deflection, limiting the technology’s application in clinical settings. Therefore, this phase aims to address this challenge by extending the capabilities of the sensorized needle and COADAP algorithm. The research plan comprises three specific aims: (Aim 1) Develop a multi-core FBG sensorized needle for robust distributed shape sensing: We will enhance the design of sensorized needles using multi-core fiber (MCF) sensors for robust and distributed needle shape sensing. We will develop a machine-learning model to predict the needle trajectory using real-time shape information. (Aim 2) Extend the COADAP algorithm for interconnected multistructural anatomy: The objective is to compensate for needle deflection in interconnected multistructural anatomy. This involves the development of an extended COADAP algorithm, called Shape- Control COADAP (SC-COADAP), to account for the full needle shape in model predictive control. (Aim 3) Validate the sensorized needle with COADAP in interconnected multistructural anatomy: We will test the hypothesis that adaptive needle placement with the MCF sensorized needle and SC-COADAP meets the required accuracy. This will be done via ex vivo and in vivo validation, using a multistructural anatomy-mimicking phantom and swine models, respectively.