Research Abridgment

If you are not too familiar with MRI imaging modalities and photon-based radiation therapy for cancer treatment, please skip to the motivation section before continuing.


  • Non-magnetic patient motion stabilizing mechanism compatible with magnetic resonance imaging (MRI) machines
  • Radio-transparent patient motion stabilizer for real-time and precise cancer radiation therapy modalities
  • Verify that a 6-DOF target motion of a patient is <= 0.5 mm and <= 0.5 deg for greater than 95% of the treatment time.


  • Simulation of a 6-DOF motion compensation soft robot for MRI-LINACs

  • Design and construction of an MRI-LINAC soft robotic motion correction mechanism

  • Phantom-based and healthy human volunteer trials

These exploratory lines of research inquiry are relevant to public health and have transformational clinical potential because they may provide:

  • Proof-of-concept evidence that soft robots are compatible with standalone MRI imaging modalities;

  • Evidence of precise and automatic motion management with non-magnetic and radiation-transparent soft robots in emerging hybrid MRI-linear accelerators;

  • An emergence of a better brain and head and neck (H&N) cancer management technology that can be adapted to confined spaces under MRI coils.


This technology shall be applicable to the following treatment modalities:

  • Standalone MRIs

  • Emerging MRI-LINAC technologies

  • Head and Neck Cancer Radiation Therapy

  • Brain Cancer Radiation Therapy


This proposed technology shall have the following advantage over rigid immobilization systems:

  • Negate the deleterious effects of interfractional setup variation on patients;

  • Correct the complex intrafractional geometric uncertainties such as posture changes, and body deformation with minimal invasiveness;

  • Eliminate radiation attenuation associated with the metallic components of frames and rigid robotic patient motion compensation systems;

  • Correct the flex drifting errors associated with thermoplastic face masks;

  • Do not interfere with the MRI’s magnetic field.

Background and Motivation

  • Current radiation therapy (RT) treatment modalities use computed tomography scans of body tissues to segment organs before treatment.

  • Accurate radiation dose targeting requires subdegree and submillimeter patient motion correction.

  • These CT images lack fine contrast that distinguish bony-tissues from non-bony and soft tissues.

  • High contrast delineation of cancerous tissues from healthy surrounding tissues can further improve dose escalation to the tumor while simultaneously sparing surrounding healthy tissues; this is especially true for brain or head and neck (H&N) cancers.

  • Magnetic resonance imaging (MRI) is an advanced imaging modality for internal body organs. Combined with RT, MRI-based RT is becoming an emerging technology with the potential for improving target and organs-at-risk (OAR) contrast for brain and H&N cancers.

  • Recent research directions have demonstrated the compatibility of MRI with linear accelerator (LINAC)-based photon treatment of cancers. _Unfortunately, the quality of MRI imaging is limited by the artifacts caused by patient head motion._

  • What if we could automatically correct patient motion during MRI imaging/RT/stereotactic radiosurgery so that we can eliminate the deleterious effects of patient motion uncertainities in MRIs, and photon/proton-based therapies?

Existing Technologies and Limitations

  • Currently in clinics, we use a frame or an immobilization mask to render the patient static while they lie supine on a couch
    • ☒ This is incapable of real-time closed-loop feedback head motion corrections when the treatment beam is on (See Fig. 1).

    • ☒ The invasiveness, inconvenience and discomfort associated with the frame are a principal cause of poor patient compliance and poor clinical efficacy.

    • ☒ For some patients, frame placement is not possible due to extreme cranial anatomy or prior surgical bone flaps. In addition, the frame prohibits cases when multiple RT deliveries are needed as patients cannot be subjected to daily attachment and removal of the frame.
    • ☒ The limitations of frames have spurred clinics using thermoplastic face masks. These result in decreased accuracy arising from mask flex (drift of up to 2-6mm), and changes in the mask from repeated application and shrinking
    • ☒ Such inaccuracies are not suitable for deep tumors located near critical structures such as the brain stem or for newer treatment modalities such as single isocenter multiple-target stereotactic radiosurgery (SRS), which are highly sensitive to rotational head motions.

Fig 1. L-R (a) The Brown-Robert-Wells SRS frame; (b) A thermoplastic face mask in RT (c) A thermoplastic facemask with add-on MRI coils in MRI imaging (d) The Wiersma Stewart-Gough platform (e) The Ostyn platform

Recent Research Directions

  • ☒ Explorative robotic positioning research studies have demonstrated the feasibility of maintaining stable patient cranial motion consistent with treatment plans. For example, the Wiersma Lab’s Stewart-Gough platform, illustrated in Fig 1d, achieves <= 0.5mm and <= 0.5 deg positioning accuracy 90% of the time, while the Ostyn six degrees-of-freedom (DOF) plastic Stewart-Gough platform (Fig. 1e) uses stepper motors to actuate the legs of its parallel plastic platform.

  • ☒ These systems, while aiding better clinical accuracy, utilize rigid metallic components, electric motors and linear actuators which are not suitable for MRI imaging: they interfere with the magnets of the MRI machine and can lead to patient fatality or significant damage to the MRI machine.

Fig 2. MRI Treatment Setup. © Kevin Teo/Rodney Wiersma, UPenn Radiation Oncology. L-R (a) Patient immobilization with thermoplastic masks under the MRI tube (b) the MRI coils are typically overlaid on the mask above the patient’s face (c) owing to the large magnetic fields of the MRI machine, metallic objects are not admissible. Hence, parallel rigid mechanisms such as the Wiersma or Ostyn robot would not be feasible. These lack soft compliance necessary in such advanced imaging modalities.

Other Technologies

  • The CyberKnife Synchrony, while capable of precise, non-surgical tumor and lesions treatment in SRS and stereotactic body radiotherapy (SBRT), only executes a-priori trajectories (see Fig 3).

  • Real-time closed-loop head motion compensation for the CyberKnife system is inhibited by its high load-to-weight ratio which indirectly affects its repeatability

    • Given its stiffness (it weighs 160kg), it has a complicated actuation system so that its passive bending stiffness overwhelms the degree of deformation for rapid patient repositioning.

Fig 3. The Cyberknife system ©Cyberknife.