SuperTouch - Data-driven tactile sensing for robot-assisted cancer surgery

Mentored by Roberto Calandra & Stefanie Speidel
at Learning, Adaptive Systems and Robotics (LASR) Lab, Department of Computer Science, TU Dresden and National Center for Tumor Diseases (NCT), UKD Dresden

Sense of touch is a crucial component for surgery as it enables the surgeon to locate tumors or pulsating vessels inside the organ. In robot-assisted surgery, this sensory input is lost and currently has to be replaced by other means. To enable sense of touch in surgery, appropriate hardware for perception as well as real-time interpretation based on machine learning are required.

The PhD topic tackles tactile sensing for robot-assisted cancer surgery based on modern high-resolution touch sensors mounted on a robot arm within a robotic surgery setup. The goal is to enable classification and localization of tumors and pulsating structures within tissue based on tactile sensory input. Research challenges that are crucial in this context include:

• How can tactile data be analyzed to sense and classify such structures using machine learning?
• Can we learn policies to control a robot for performing automatic exploration and mapping of tissues with diverse mechanical properties, based on touch inputs?
• What are the right machine learning models to efficiently integrate touch data over the temporal and active nature of touch?
• How can evaluation scenarios in this context be generated that serve as ground truth?

Work environment

You will be working at the Learning, Adaptive Systems and Robotics (LASR) Lab at the department of computer science, TU Dresden and the National Center for Tumor Diseases (NCT) Dresden at the medical faculty.

LASR is a newly established lab at TU Dresden focused on research at the conjunction between artificial intelligence (AI) and robotics. One of the signature research in which LASR specializes is the vertical development of touch sensing and processing. This includes design of novel touch sensors, such as DIGIT (https://digit.ml/) which is the most widely used high-resolution touch sensor in robotics, as well as the development of novel machine learning-based algorithms for making sense of the rich sensorial information of touch.

The 2nd workplace will be NCT Dresden located at the medical campus combines patient therapy and research under one roof and offers a unique research platform including an experimental operating room and a novel simulation room for robot-assisted surgery.

Prerequisites

• Master’s Degree (or equivalent) in computer science, electrical engineering, biomedical engineering, applied mathematics or related fields of expertise
• Very good programming skills (e.g. C++, Python)
• Excellent skills and practical experience in one or more of the following research areas is beneficial:
• Machine Learning
• Computer vision
• Touch sensing
• Robotics
• Ability to collaborate well in an interdisciplinary environment