A new robot should significantly increase the accuracy of biopsies. The robot combines MRI images with ultrasound into a 3D model, which, among other things, helps doctors determine the correct placement and route of the biopsy needle. In addition, because fewer MRI images are needed, biopsies can be performed more cheaply.
This is reported by Stefano Stramigioli (affiliated with UT's Department of Robotics and Machetronics), Françoise Siepel (UT postdoc), Jeroen Veltman (radiologist at ZGT) and Wouter Nijhof (Siemens Healthineers) in an interview with HealthMatters Magazine from Siemens Healthineers. The robot is being developed by within the MRI and Ultrasound Robotic Assisted Biopsy (MUR&B) project. This project aims to dramatically improve the accuracy and effectiveness of biopsies for making cancer diagnoses. Within the MUR&B project, a new method of taking biopsies is being developed that minimises the use of expensive MRI equipment.
Determine placement and route of needle
The biopsy robot uses ultrasound to scan a patient's body part. This scan is combined with previously taken MRI images to create a 3D model. Based on this model, the robot helps the doctor determine the optimal placement and route of the biopsy needle.
Using the robot offers several advantages, according to the parties involved. For example, the robot allows:
- perform accurate ultrasound-guided biopsies without using a more expensive and time-consuming MRi scan;
- perform ultrasound-guided biopsies for abnormalities that cannot normally be seen on ultrasound images;
- perform biopsies faster and more accurately;
- perform biopsies with less patient discomfort;
- reduce the number of over-diagnoses and thus the number of over-treatments.
Making diagnoses with almost 100% certainty
The project was set up by Stramigioli in 2016. "The biopsy robot combines the advantages of the systems we already have," says Stramigioli. "An MRI is expensive but very precise, allowing a doctor to puncture exactly the tissue he or she needs. Echo equipment gives real-time images, allowing a radiologist to see exactly what is happening while taking a biopsy. However, abnormalities are less easy to distinguish on this and are therefore not always recognised. Based on the 3D model, a radiologist can take a biopsy very accurately and make a diagnosis with almost 100% certainty."
Veltman: "Because an MRI does not give a real-time image, you have to make several scans. Beforehand, to select your target, after you insert the needle and afterwards, to see if you have actually passed through the desired tissue. For a patient, this is a lengthy procedure that requires lying uncomfortably with the breast in a breast coil. Add to this the fact that the doctor's freedom of movement is restricted by the breast coil and MURAB means double gain. The breast hangs freely and can be prodded from all sides, even if the abnormality is near the armpit. Because MRI images are loaded into the robotic arm's software, a doctor can still distinguish a suspicious spot on ultrasound. If you know where a spot is, you recognise it faster on the ultrasound."
Universities, hospitals and industry partners
Several Dutch universities and hospitals are involved in the project, including the University of Twente, Radboud UMC and ZGT. In addition, the University of Verona and Medical University of Vienna have also joined the MUR&B project as partners. The robotic arm on which the biopsy robot is based was developed by KUKA, while Siemens supplies the imaging systems. The project was previously awarded a European Horizon 2020 grant.
"By getting into a project early, we can facilitate what researchers need. We provide MRI and ultrasound and make sure those systems are optimally tuned to work together with the robotic arm," says Nijhof. Stramigioli adds: "Siemens Healthineers has the beautiful eyes. We are now developing the effective hands, so that a doctor can achieve exactly what he or she wants, with the best possible accuracy."
Test setup at UT
The biopsy robot is currently being tested in a test rig at UT. The robot is placed under a treatment table, in which there is a recess at breast level. A phantom breast hangs through this hole in the test setup. The robotic arm scans the breast and combines the data collected with data previously obtained using MRI and ultrasound. Based on this, a 3D model of the inside of the breast is created.
The 3D model is shown on a screen to the doctor, who can analyse the images here. To perform the biopsy, the doctor only needs to select a location on the screen. The robotic arm then moves the biopsy needle to the designated position and suggests a puncture route. This last part in particular is very challenging, according to the researchers, partly because the flexibility of the breast and the deformation that occurs when the needle enters the breast must be taken into account. The robot's proposal is always assessed by the doctor, who then inserts the needle himself. During the biopsy, the doctor can follow the needle in real-time on a monitor.
Applications
The project is initially aimed at diagnosing breast cancer and muscle diseases. In time, the parties involved want to explore other applications. Stramigioli: "Due to its flexibility and wide variety of tissue, the breast is very complicated to take a biopsy from. If the breast examination succeeds, the way is open to other applications. Moreover, the discomfort experienced by these patients during a biopsy is relatively high and so is the benefit we can achieve for them."
The project is now entering the clinical testing phase. "We now first need to make sure everything works, so to speak, at the touch of a button. Then we can scan and diagnose phantom breasts. Then we will submit an application to the medical ethics committee. Only when they have confirmed that our setup is safe will the tests with volunteers start," Siepel explains. The project will run for four years and should be completed by 2020.
Author: Wouter Hoeffnagel
Source: Siemens Healthineers
Source: MUR&B project
Source photo: Pixabay / jarmoluk
