EXPERIMENT_INNITIUS

Context:

Support for real-time guidance of the Innitius transvaginal probe for torsional wave-based measurement, aiming to improve inter- and intra-observer agreement. R&D in the fields of computer vision and artificial intelligence, along with augmented reality, to restore the image from the endoscopic camera, detect the cervix, indicate the appropriate location, and adapt it to the embedded device.

Project objective:

The goal of this project is to automate the guidance process of the Innitius transvaginal probe to enable accurate measurement of cervical stiffness. The device includes a torsional wave emitter and sensor, along with an embedded endoscopic camera and screen. For hygiene purposes, the device is covered with a sheath, which degrades image quality and complicates navigation.

The innovative scientific-technological objective involves research and development in artificial vision, artificial intelligence, and augmented reality to restore image quality, automatically identify the cervix location, and guide the operator via the device’s screen to the correct position for cervical stiffness measurement using torsional waves.

Proposed solution:

Augmented reality (AR)-based software that displays the cervix location on the screen when visible and guides the user to place the probe precisely in the correct position. In addition to AR-based guidance, the screen shows an enhanced version of the image, mitigating issues caused by the sheath and lubricant.

Key features:

• Image transformation algorithm based on deep learning (e.g., context adaptation, noise reconstruction, transformation) to enhance the image.
• Cervix detection algorithm.
• AR-based positioning algorithm.

The industrial mission of the project is the implementation of the artificial intelligence algorithm in an embedded system such as the Innitius device, addressing technological challenges related to computing resources, memory, cost, size, etc.

The Innitius device aims to diagnose preterm birth threats and avoid unnecessary hospitalizations and medication for both mother and baby. The company’s social impact is clear, as it will increase diagnostic effectiveness, thereby reducing the risk of adverse events and significantly improving and reducing problems related to childbirth.

Results:

• Image reconstruction algorithm for enhancement.
• Cervix detection algorithm.
• Visual guidance application based on augmented reality combining: a) image quality improvement, b) cervix detection, c) alignment with cervix detection.