Robot-Assisted US-Based Vertebral Segmentation for Pedicle Screw Trajectory Identification

Robot-Assisted US-Based Vertebral Segmentation for Pedicle Screw Trajectory Identification: A Comparison With Handheld US-Reconstructions and Ground Truth CT-Data

This other clinical trial compares robot-assisted US scanning with handheld US scanning and ground-truth CT data of the lumbar spine in healthy, young volunteers. The main questions it aims to answer are:

• Is a 3D reconstruction of a lumbar spine from robot-assisted US scanning equivalent to or better quality than a 3D reconstruction from handheld US scanning?
• Can a machine learning algorithm automatically segment the bone anatomy from robot-assisted and handheld US scanning to generate 3D lumbar spine reconstructions?
• Can pedicle screw trajectories be identified based on posterior vertebral landmarks of 3D reconstructions of lumbar spines from both robot-assisted and handheld US scanning?

Participants will:

• fill out a medical history questionnaire
• get clinically examined
• have an ultra-low-dose (ULD) CT Scan of the vertebra L1 to S1
• have a handheld US scan of the vertebra L1 to S1
• have a robot-assisted US Scan of the vertebra L1 to S1
• fill out a post-study questionnaire

Study Overview

• Status: Completed
• Conditions: Healthy
• Intervention / Treatment: Other: Robot-assisted US Scan of the vertebra L1 to S1; Other: Handheld US scan of the vertebra L1 to S1; Radiation: ULD CT scan of the vertebra L1 to S1

Detailed Description

The following hypotheses are tested:

1. A 3D reconstruction of a lumbar spine from robot-assisted US scanning is equivalent to or of better quality than a 3D reconstruction handheld US scanning.
2. A machine learning algorithm can automatically segment the bone anatomy from robot-assisted and handheld US scanning to generate said 3D lumbar spine reconstructions.
3. Pedicle screw trajectories can be identified based on posterior vertebral landmarks of 3D reconstructions of lumbar spines from robot-assisted and handheld US scanning.

The project consists of three pillars as objectives to help solidify the US reconstruction of the lumbar spine as a novel navigational method in interventional spine applications.

• 1st Pillar: A first-of-a-kind in-vivo robot-assisted and handheld US reconstruction dataset of the lumbar spine in healthy subjects is acquired. The collected dataset is compared to ground truth CT data to assess quality.
• 2nd Pillar: A novel machine learning algorithm is trained to segment the US reconstructions of all the collected lumbar spine data into each identified vertebra.
• 3rd Pillar: A novel measurement method to identify pedicle screw trajectories based on posterior vertebral landmarks is applied to the segmented US reconstructions. This research further promotes US for future use in robot-assisted interventions.

This project consists of two phases. First, a preliminary pilot study is planned to assess the project's feasibility and improve the planned workflow and safety measures. For this pilot, the investigators will mouth-to-mouth recruit two volunteers. After completing and thoroughly evaluating the pilot, the investigators will conduct the actual study.

The volunteers for the actual study are selected through public calls for participation. Possible volunteers are young, healthy, and not affected by illness or deformation of the lumbar spine. The selected volunteers are screened by asking about their medical history. If included and willing to participate, the volunteers are invited to the study at Balgrist Campus and will be clinically examined regarding the lumbar spine. Furthermore, a low-dose CT scan, a handheld US scan, and a robot-assisted US scan are held.

The CT scans are manually segmented into 3D surface models to obtain a "segmentation ground truth". A novel machine learning algorithm automatically performs 3D reconstruction and segments the robot-assisted and handheld US scans.

The 3D US reconstructions are then utilized to identify pedicle screw trajectories through a novel method based on the posterior anatomical landmarks of lumbar vertebrae.

This single-center study combines the clinical and computer-science knowledge from the Research in Orthopedic Computer Science (ROCS) team of the University of Zurich, Switzerland, with the robotics and US application knowledge from the Faculty of Engineering of the University of Leuven, Belgium. The data collection is performed at Balgrist Campus.

• Study Type: Interventional
• Enrollment (Actual): 63
• Phase: Not Applicable

Contacts and Locations

Study Locations

• Switzerland
  • Zurich, Switzerland, 8008
    • University Hospital Balgrist, Balgrist Campus

Participation Criteria

Eligibility Criteria

• Ages Eligible for Study: Adult
• Accepts Healthy Volunteers: Yes

Description

Inclusion Criteria:

• Oral and written informed consent from the volunteer
• Volunteers aged ≥18 y/o and ≤35 y/o
• BMI above or equal to 19 kg/m2 or below or equal to 25 kg/m2
• Proficient in German or English language

Exclusion Criteria:

• Documented volunteer refusal
• Volunteers in whom CT cannot be performed
• Positive pregnancy test prior to radiology (contraindication to CT)
• Pregnancy
• Chronic pain in the lumbar spine
• Moderate or severe deformity of the lumbar spine

• Any prior intervention to the lumbar spine:

  • Chiropractic adjustment therapy
  • Injections such as local anesthetics and corticosteroids
  • Surgery
• Fracture of the lumbar spine
• BMI below 19 kg/m2 or above 25 kg/m2
• Anatomies, such as subcutaneous fat or tendon, occlude the bony surface or do not allow a clear image in the US scan

Study Plan

How is the study designed?

Design Details

• Primary Purpose: Other
• Allocation: N/A
• Interventional Model: Single Group Assignment
• Masking: None (Open Label)

Number of Arms

1

Arms and Interventions

Participant Group / Arm

Intervention / Treatment

Other: Young, healthy volunteers; Medical History (20 minutes); Clinical examination (20 minutes); Clothes changing time (15 minutes); ULD CT scan of the vertebra L1 to S1 (30 minutes); Clothes changing time (15 minutes); Handheld US scan of the vertebra L1 to S1 (20 minutes); Platform changing time (15 minutes); Path planning and robot-assisted US Scan of the vertebra L1 to S1 (40 minutes); Clothes changing time (15 minutes); Volunteer Questionnaire (10 minutes) Total study time per volunteer: 3 Hours 20 minutes

Other: Robot-assisted US Scan of the vertebra L1 to S1; A standardized robot-assisted US-Scan of the vertebra L1 to the S1 is executed. The two probe trajectories are generated automatically by an optical camera or RedGreenBlue-Depth (RGB-D) camera. Then the robot will scan at two mm/s, automatically following the trajectory. A 6 degrees of freedom (DoF) force sensor (Nano25, Ati Industrial Automation Inc.) was assembled at the US probe. During the scanning, it measures the interaction force and torque between the US probe and volunteers' skin to ensure volunteer safety. The US images and corresponding robot poses are recorded for the reconstruction. A previously developed algorithm is used to obtain a preliminary 3D reconstruction that will be used to adapt the scanning pattern in case of incomplete scanning.; Other: Handheld US scan of the vertebra L1 to S1; A handheld US-Scan of the vertebra L1 to S1 is executed. An optical marker for an Atracsys Camera is mounted to the US probe to track its position. Simultaneously, the optical tracking allows for standardization of the US-Scan through navigation. The acquired US images and corresponding poses are used to reconstruct the lumbar spines and compared to the ground truth (CT scan). Handheld US scanning will be performed at around two mm/s in two scanning trajectories. The US probe orientation will be adjusted during the scanning by comparing the optical marker orientation with the initial probe orientation. This combination leads to different scanning patterns. For each pattern, the scanning procedure will be repeated three times. At least once per volunteer, each handheld scanning pattern will be performed with a higher manual speed to assess the reconstruction quality with faster scanning. A graphical user interface (GUI) will synchronize and store the data during the scanning.; Radiation: ULD CT scan of the vertebra L1 to S1; A ULD CT scan (CT, Siemens Naeotom Alpha: A-207883-62) of the lumbar spine (L1-S1) is performed at the Swiss Center for Musculoskeletal Imaging (SCMI) at Balgrist Campus, Zurich. The total duration estimate for the CT examination is 30 minutes, whereas the scan takes 15 minutes. The volunteers lie on their abdomen during this procedure to simulate the spine's position during the subsequent US examination. The vertebrae (L1-S1) in the CT Scans are manually segmented with global thresholding and the region growing tool in a standard segmentation software (Materialise Mimics, Leuven, Belgium). The 3D surface models are then exported as Standard Triangle Language (STL) files.

What is the study measuring?

Primary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Target registration errors between US reconstructions and ground truth CT data	Evaluating the accuracy of the US reconstructions	Up to 1 year
Pedicle screw placement - Trajectory errors in terms of position	Evaluating the positional accuracy of the Pedicle screw placement	Up to 1 year
Pedicle screw placement - Trajectory errors in terms of direction	Evaluating the directional accuracy of the Pedicle screw placement	Up to 1 year

Secondary Outcome Measures

Outcome Measure	Measure Description	Time Frame
Gender	female, male, non-binary, do not know, other	Up to 4 weeks
Age	in years	Up to 4 weeks
Smoking status	yes/no; if yes, pack years	Up to 4 weeks
BMI	weight and height	Up to 4 weeks
Tegner activity score	Subjective activity score of the volunteers	Up to 4 weeks
ODI Oswestry Low Back Pain Disability Index	Standardized Low Back Pain Disability Index	Up to 4 weeks

Collaborators and Investigators

• Sponsor: Philipp Fürnstahl
• Collaborators: KU Leuven; Swiss Center for Musculoskeletal Imaging
• Investigators: Principal Investigator: Philipp Fürnstahl, PhD, Balgrist University Hospital

Study record dates

Study Major Dates

• Study Start (Actual): May 15, 2023
• Primary Completion (Actual): June 9, 2023
• Study Completion (Actual): June 9, 2023

Study Registration Dates

• First Submitted: May 3, 2023
• First Submitted That Met QC Criteria: June 6, 2023
• First Posted (Actual): June 15, 2023

Study Record Updates

• Last Update Posted (Actual): June 22, 2023
• Last Update Submitted That Met QC Criteria: June 21, 2023
• Last Verified: June 1, 2023

More Information

Terms related to this study

• Keywords: lumbar spine; CT scan; healthy, young volunteers; handheld ultrasound scan; robot-assisted ultrasound scan
• Other Study ID Numbers: W1001

Plan for Individual participant data (IPD)

• Plan to Share Individual Participant Data (IPD)?: YES
• IPD Plan Description: After completing our analysis, we plan to make the dataset of robot-assisted US Reconstructions available to other research groups. The individual participant data (IPD) will be stored in a data repository.
• IPD Sharing Time Frame: We plan to make the IPD available within 24 months of data acquisition or after peer-reviewed publication.
• IPD Sharing Access Criteria: First we evaluate the access by request. After intellectual property (IP) and publications are completed we will consider open-access for other research group.
• IPD Sharing Supporting Information Type: STUDY_PROTOCOL

Drug and device information, study documents

• Studies a U.S. FDA-regulated drug product: No
• Studies a U.S. FDA-regulated device product: No
• product manufactured in and exported from the U.S.: No