Researchers use motion capture technology to install 3D printed sensors directly on organs

China3D printingNet, June 22, researchers at the University of Minnesota have devised a novel method that can deploy Hollywood-style motion capture technology to help integrate sensors3D printingTo the organs that can expand and contract.
Printing directly on moving soft tissues is challenging because the sensor needs to be able to adapt to the changing parameters of the organ.On the other hand, the research team’s new technology uses dual cameras to create real-time3D printingThe tool path to overcome this obstacle. Not only can this method be used by medical staff to monitor infectious patients from a safe distance, but according to the research team, it can even be used to diagnose and monitor the lungs of COVID-19 patients.

The lead researcher of the project, Professor Michael McAlpine, said: “We are breaking through in a way that we never thought of 3 years ago.3D printingThe boundaries. future,3D printingIt will be more than just printing, but part of a larger autonomous robotic system. This can be very important for diseases like COVID-19, because in this case, healthcare providers are at risk when treating patients.


3D bioprintingAnd sensor

The latest developments in additive manufacturing have been compatible with3D printingThe range of materials has expanded from traditional plastics and metals to conductors and biological materials. Despite this progress, AM still finds it difficult to influence bioprinting applications that target living biological surfaces because they are usually soft and constantly move and deform.In addition, due to3D printingThe machine relies on the prescribed design of off-line manufacturing, and then transfers to the biological surface, which will cause mismatches, and the system is essentially making “blind holes”.

Manual adjustment of the printing process is also not a viable option, as the transfer process can destroy fragile 3D structures (such as hydrogel materials) and cause costly human error. The researchers identified another solution: in-situ printing seamlessly integrates the sensor on the target surface in an autonomous manner. In order to achieve this, a new closed-loop artificial intelligence (AI) is needed to adjust the manufacturing process in real time by dynamically sensing the geometric state of the biological substrate.

The research team in Minnesota has developed a closed-loop system that can track the movement of “undeformed hands,” but this method will require the development of a sophisticated algorithm that can accurately track high-dimensional deformation data.Instead, the researchers suggested3D scanningThe data set “learns” the deformation space of the target surface.This will allow a set of fiducial markers tracked by a stereo camera system to recover accurate surface geometry in 3D and be used to dynamically adapt in real time3D printingTool path.

In order for this method to be successful, the strain sensor used to calculate the deformation measure needs to be in contact with the surface of the lung tissue and in situ3D printingProcess compatible. Conventional sensor designs are based on densely packed miniature sensor arrays, electrodes and interconnections to improve resolution, but this approach is not compatible with the uncertainty that may arise during printing. Or, due to its high transparency and stretchability, the research team has developed a new sensor that combines an ionic hydrogel and an electrical impedance tomography (EIT) sensor.In order to demonstrate their novel sensor and motion capture technology, the researchers will be based on3DThe EIT strain sensor of the hydrogel is directly printed on the breathing lung to monitor its deformation.




The research team used a dual-camera motion tracking method to accurately place the sensor on the pig’s lungs. The picture is from “Science Progress”.

The Minnesota State Team’s Motion Capture Method

Although researchers could use a stereo camera system to restore the changing 3D geometry of the organ surface in real time, the reconstruction algorithm is not accurate enough because it requires sub-millimeter accuracy to avoid tissue damage. In addition, the research team designed a two-stage program through which the system learned the parametric model of the surface geometry and used a set of fiducial markers to estimate the lung parameters, thereby supplementing the model.

In the first step, the surface deformation is modeled using the movement of 12 fiducial markers and the movement of 3968 waypoints, and they are extracted to create a printing tool path. Before projecting the planar tool path, some corrections have been made to reflect the physical growth of the sensor size when the surface expands and the shrinkage when the surface shrinks.In the second stage, two machine vision cameras are installed on3D printingOn the extrusion head of the machine to monitor the sensing process in real time. The time series of parameters reflects the sweat of the lungs and is used to estimate the conformal tool path that changes with the rigid body motion of the tissue.

Transfer adaptive toolpaths to3D printingAfter the machine, the squeeze nozzle will follow the user-specified printing speed profile and resampled waypoints. The combination of deformation estimation reduces the average error of shape modeling from 0.3 mm to less than 0.02 mm, and the total error is within a tolerance of ±0.8 mm. In addition, when the shear rate of the ionized hydrogel ink is above 0.1 s-1, it shows shear thinning behavior as the viscosity decreases. This lower viscosity allows the ink to be smoothly squeezed out of the printing nozzle under air pressure and improves its controllability. After the use of UV light crosslinking, according to the uniaxial tensile test results, the hydrogel ink also showed tissue-like stretchability.

The EIT sensor consists of a continuous thin layer of hydrogel and copper electrodes, embedded in a soft silicone ring, which can form a chemical bond with the hydrogel to maintain a stable hydrogel-electrode interface. During the test, 40 measurements were made from adjacent electrode pairs, and a complete estimation was made with the corresponding map. The maximum average error recorded was 5.25%, which is similar to CT scan reconstruction.To prove its new in-situ AI drive system3D printingAbility, the researchers installed the EIT strain sensor directly on the pig’s lungs.

The tracking results produced by the AI-driven system are comparable to those obtained using the CT scanning method. The picture is from the University of Minnesota.
Experiments and future applications

In order to simulate the deformation of the organ, the lung trachea was connected to a digital pneumatic regulator, and then the surface geometry in each deformed state was sampled by a structured light scanner.Then use customized3D printingIn the gantry system, the electrodes embedded in the silicone ring are connected to the printed layer and exposed to ultraviolet light for cross-linking. Estimated and displayed the spatial map of the lungs in real time, successfully capturing the periodic contractions of the organs. The sensor can not only adhere to the lung surface under repeated deformation, but also can be removed immediately without damaging the tissue or leaving it behind.

As a result, the research team successfully demonstrated the in-situ monitoring of organ deformation, and its soft sensor was3D printingOn the breathing lungs. The new method integrates offline machine learning with online tracking based on computer vision and conductive hydrogel ink with EIT sensing configuration to achieve this result. Although the research team admits that the biocompatibility of the sensor can be improved, the accuracy of the method has improved, but they believe that this may open a new field of surgery for bioprinting.For example, in clinical situations where the injection of biological materials (such as surgical glue and skin grafts) is required, autonomous in situ3D printingIt can replace manual operation to achieve precise space control in a longer period of time.
3D bioprintingSensor in

In recent years,3D printingIt has been used to create a series of different biosensors for medical monitoring.For example, in May 2020, researchers at Sungkyunkwan University used commercially available inkjet print heads to make wearable medical biosensors for personalized health monitoring. The team used soft, soft silicone elastomers and sugar scaffolds to print high-resolution images in a high-resolution environment. Lightweight conductive packaging.

Scientists at Georgia Institute of Technology and Hanyang University developed the first aerosol jet printing (AJP) biosensor for wireless monitoring of blood flow in August 2019.research team3D printingAn implantable and stretchable electronic system that can monitor the blood flow of aneurysms in the brain. treat.

In December 2018, scientists at Washington State University (WSU) used3D printingTechnology has created more effective glucose screening equipment for diabetics. The WSU biosensor is attached to the skin and can monitor the glucose content of a person’s sweat instead of relying on their blood.

China3D printingNet compile article!

(Editor in charge: admin)