It cures the smart smarter, which works from artificial intelligence with 25 % faster

A device that can be worn called “A-Heal”, designed by engineers at the University of California, Santa Cruz, aims to improve each stage of the process. The system uses a small camera and Amnesty International to detect the healing stage and provide treatment in the form of a drug or an electric field. The system responds to the unique recovery process for the patient, providing personal treatment.

The mobile wireless system can make the wound treatment easier for patients in remote areas or with limited mobility. Prepo -clinical results, published in the magazine NPJ Biomedical InnovationsSuccessfully showing the device accelerates the healing process.

A-DELE design

A team of California University researchers in Santa Cruz and researchers at the University of California, Davis, sponsored the Darpa-Betr program, led by UC Santa Baskin Engineering Chair and Professor of Electrical Engineering and Computer importance (ECE) Marco Rolandi, designed a device that combines the camera, vital electron and AI for the healing. Including in one device makes a “closed loop system”-one of its first of its kind to heal wounds as much as researchers realize.

“Our system takes all the sermon from the body, and with external interventions, it improves the progress of healing,” said Rolandi.

The device uses a camera on the plane, developed by an associate professor of ECE MIRCEA TeODorescu and described it in the study of communications biology, to take pictures of the wound every two hours. The images are fed in the ML Learning model (ML), which was developed by Associate Professor of Applied Mathematics Marce Gomez, which researchers call the “artificial intelligence doctor” that works on a computer nearby.

“It is mainly a microscope in a bandage,” said Tudorisco. “Individual images say little, but over time, it allows the continuous photography of the factions of the sponing organization, the stages of wound healing, science issues, and the proposal of treatments.”

The artificial intelligence doctor uses the image to diagnose the wound stage and compares this to the place where the wound should be along a schedule for healing optimal wounds. If the image reveals delay, the ML model applies the treatment: either the drug, it is delivered via vital electronics; Or an electric field, which can enhance cell deportation towards the closure of the wound.

Treatment that is delivered topically across the device is Flucstin, and it is a selective serotonin absorption that controls serotonin levels in the wound and improves recovery by reducing inflammation and increasing the closure of the wound tissue. The dose, determined by pre -clinical studies conducted by the ISSEROFF group in the UC Davis group to improve healing, is given by vital electronic motors on the device, developed by Rolandi. An electric field is also delivered, improved to improve recovery and developed through the previous work of UC Davis’ Min Zhao and Roslyn Rivkah Isseroff, through the device.

The artificial intelligence doctor determines the optimum dose of medications that must be connected and the size of the applied electric field. After applying treatment for a certain period of time, the camera takes another image, and the process begins again.

While using it, the device transmits images and data such as the healing rate to a safe web interface, so that the human doctor can manually intervene and accurately treat as needed. The device directly suspends a bandage commercial for comfortable and safe use.

To assess the possibility of clinical use, the UC DAVIS team tested the device in pre -clinical wound models. In these studies, the wounds treated with A-Heal followed the recovery path around 25 % faster than the level of care. These results shed light on the promise of technology not only to accelerate the closure of sharp wounds, but also for the suspended recovery in chronic wounds.

You have to reinforce

The artificial intelligence model used for this system, led by Assistant Professor of Applied Mathematics Mariela Gomez, uses the reinforcement learning approach, described in a study in the Journal of Biological Engineering, to imitate the diagnostic approach used by doctors.

Reinforcement learning is a technique in which the model was designed to achieve a specific goal, learning through experience and error in the best way to achieve this goal. In this context, the model is granted a target to reduce time to reduce the closure, and is rewarded for progressing towards this goal. He learns constantly from the patient and adapts to the treatment approach.

The augmented learning model is guided by the algorithm that Gomez and its students were established called Deep Mapper, described in the Preprint study, which treats wounded images to measure the healing phase compared to natural progress, and draw them along the healing path. Over time with the device on the wound, a linear dynamic model for the past recovery learns and is used to predict how to continue to recover.

“It is not enough for you to have the picture only, but you need to process it and put it in a context. Then, you can apply control in the comments,” Gomez said.

This technique allows the actual time to learn the effect of drugs or the electric field on healing, and the repetitive decision of the learning model is directed on how to control the focus of the drug or the power of the electric field.

Now, the research team explores the possibility of this device to improve the healing of chronic and infected wounds.

Additional posts related to this work can be found here.

This research has been supported by the Advanced Defense Research Projects Agency and the Agency for Advanced Health Research Projects.