LCSR Seminar: Junichi Tokuda

Description

Junichi Tokuda, an associate professor of radiology at Brigham and Women's Hospital and Harvard Medical School, will give a talk titled "Bridging Robotics and Clinical Practice Through the Iterative Development of Robotic Needle Placement" for the Laboratory for Computational Sensing + Robotics.

Abstract:

Image guidance and robotics are transforming clinical practice, enabling precise interventions that improve patient outcomes. In this talk, I will use my research on percutaneous needle placement—an essential procedure for cancer diagnosis and treatment—as an example of how we integrate robotics research into clinical practice through an iterative process of clinical testing and fundamental research.

Percutaneous needle placement is critical for both diagnosing and treating cancer. Needle biopsies are widely performed to collect tissue samples from suspected lesions while confirmed lesions may be treated with brachytherapy or thermal ablation using applicator needles. Accurate needle placement is essential to avoid false-negative diagnoses and ensure optimal dose distribution. As diagnostic imaging advances, clinicians can now pinpoint subregions within heterogeneous tumors with greater accuracy, increasing the demand for precision in needle placement.

Despite efforts to enhance prostate needle placement using image guidance and robotic assistance, in vivo accuracy remains a challenge due to uncertainties in needle-tissue interactions, particularly needle deviation in heterogeneous tissue. Addressing this issue requires a continuous cycle of in vivo validation and development of new technology, refining both the conceptual models and technological solutions.

In this talk, I will present our approach, which moves fluidly between in vivo testing and foundational research to develop and validate new technologies. These efforts include collaborations with Johns Hopkins University and other institutions on smart shape-sensing needles, needle-guiding robots, and data-driven steering control. Additionally, I will demonstrate how modern tools such as 3D printing, open-source research software, and large language models accelerate this iterative process, allowing us to prototype, test, and refine new technologies before translating them into clinical practice.