Urgent news
DS04 - Vie, santé et bien-être

MicroRobot-assisted Cholesteatoma Surgery – muRoCS

MicroRobotic Assisted Cholesteatoma Surgery

Cholesteatoma, an epidermal mass in the middle ear, often results from chronic otitis or tympanic perforation. Affecting 1 in 10,000 people/year, it progressively invades, impacting facial nerves and internal structures. Surgical treatment involves the total removal of pathological tissue. Detection of residues during surgery is crucial to avoid frequent recurrence (10-50%). It is imperative to develop effective, minimally invasive treatments.

Exhaustive Cholesteatoma Surgery using a Mini-Invasive Approach

The µRoCS project proposes an innovative integrated robotic system comprising a main robot and a flexible surgical microrobot. This microrobot offers high precision to reach specific areas of the middle ear and effectively remove cholesteatoma, notably behind the mastoid bone and between the eardrum and the ossicular chain.<br />The microrobot is designed with a millimetric operating channel for the passage of investigative and interventional instruments. Its distal part is equipped with a multimodal fibered medical imaging system, combining hyperspectral fluorescence and optical coherence tomography (OCT). This combination, combined with machine-learning methods, enables real-time classification of residual pathological cells and healthy tissue. Cholesteatoma areas are treated by a fibered surgical laser, with precise control provided by the microrobot. In this way, the system offers an advanced solution for the detection and treatment of cholesteatoma, improving surgical efficiency while minimizing risks.

At the Centre Hospitalier Universitaire de Besançon, the «Concerto« study explored the optical characterization of cholesteatoma via fluorescence spectroscopy and optical coherence tomography (OCT). A commercial OCT system combined with hyperspectral fluorescence imaging enabled the simultaneous acquisition of tissue signatures. Data obtained from middle ear biopsies were processed using principal component analysis and machine learning classification to distinguish normal from cholesteatomatous tissue. The ultimate goal is to integrate this multimodal system into a millimetric micro-robot. The study also developed optical micro-probes combining OCT (900 nm) and fluorescence (405 nm) using commercial gradient-index lenses and custom-made micro-axicons fabricated by femtosecond laser-assisted etching.
A comparative study of robotic solutions (comanipulation and teleoperation) with the traditional manual method was carried out. Participants had to perform a target-pointing exercise using the endoscope tip. The campaign involved the creation of a mock-up simulating the inside of the middle ear, and the installation of a device to measure time and record position. The criteria evaluated were execution time, pointing accuracy, and collisions with the walls of the ear canal. Twelve subjects took part. In addition, a method of assisting initial positioning was evaluated via a trial campaign with the existing macro-micro system. This method combines traditional control (comanipulation or teleoperation) for linear movements and visual control for the orientation of the robot tip. Finally, the robotic system was used to evaluate the automation of laser scanning. A specific device was assembled, including an integrated laser, a mock-up of residual cholesteatoma, and an endoscopic view simulating a real operation.

The mechatronic design of deformable robots can also be applied to other anatomical areas. The geometric-static design of the concentric-tube robot and the tendon-actuated, asymmetrically-cut wrist make it possible to reach hard-to-reach areas of the middle ear. A compact, lightweight, piezoelectrically-actuated prototype ensures sufficient precision and travel for these tasks. Validation of the fiber ablation laser was carried out on a tabletop demonstrator, attesting to its integration into the hybrid robot with concentric tubes. Tests on the plug to hold the fiber and tendon proved the superiority of femtosecond laser-assisted wet etching (FLICE) on glass, which withstands high temperatures during laser ablation. Laser ablation trials, followed by OCT, enabled the prediction of the volume to be removed according to pulse power and duration.
In the comparative study between robotic approaches and manual otologic surgery, robotic solutions showed better precision despite a longer operating time. Teleoperation slightly improved operating comfort, with programming to avoid collisions. The hybrid method of assisting initial positioning, combining traditional control and visual control, was successful, surpassing in convergence and precision the comanipulated or teleoperated approaches alone. Evaluation of laser scanning automation was conclusive, achieving 96% coverage of residual cholesteatoma with high follow-up accuracy.
The clinical study collected 38 samples from 23 patients, generating 3787 fluorescence spectra. In-depth analysis achieved 93.9% accuracy, 96.0% sensitivity, and 90.9% specificity in the detection of cholesteatoma compared to other tissues. The axicon micro-lenses developed produce a Bessel beam of over 3.5 mm with a diameter of 5.5 µm. With a diameter of 500 µm, they can be integrated into a microrobot, ensuring OCT imaging without losing resolution over a wide area of investigation. The quality of the tips reaches a level unmatched to date, suggesting their applicability beyond the project.
The gerbil was selected for its potential to spontaneously develop a cholesteatoma similar to humans. An Apafis protocol was drawn up and validated by the Institut Pasteur's Ethics Committee to ensure that the experiments were carried out in compliance with regulations. Difficulties arose during the surgical procedures, notably the absence of a mouthpiece adapted to inhalation anesthesia for this species. Precise adjustments were undertaken to design, test, and validate a specific mouthpiece, overcoming anatomical differences with other rodents.

The µRoCS project has made significant advances in the detection and laser ablation of cholesteatoma using a continuous robot. The control interface for teleoperation and comanipulation of endoscopic and carrier robots was developed. The two-port minimally invasive surgical protocol was welcomed as a viable alternative.
The COnCERTO study confirmed the effectiveness of autofluorescence spectroscopy combined with machine learning to characterize cholesteatoma. Intraoperative use of a device based on this model could improve detection, facilitate resection, and potentially reduce recurrence. A self-contained, portable, real-time cholesteatoma detection system has been developed, using fluorescence spectra and discriminant analysis classification. This millimetric system can be integrated into the Hearing Institute's surgical robot for animal clinical trials.
The continuously deformable, hybrid-actuated robot is suitable for middle ear exploration. Although it is possible to integrate laser tools (for detection or ablation), a balance needs to be struck to avoid excessive rigidity, which would make handling tricky and could lead to tendon rupture through excessive force.
This project has demonstrated that the move towards otologic surgery using a microrobotic approach offers promising prospects for the future of middle ear surgery, and could even be extended to all interventions in the inner ear.

The µRoCS project has been a success, with six articles published in high-impact international journals such as Optics Express, IEEE Robotics and Automation Letters, and Frontiers in Robotics, three of which were multi-partner publications. In addition, the project has enabled four papers to be presented at international conferences. The teams are currently actively preparing three additional articles for peer-reviewed journals.

Cholesteatoma is a skin growth that occurs in an abnormal location in the middle ear. It is usually due to repeated infection. It was estimated that one new case per 10,000 citizens occurs each year. Over time, cholesteatoma expands in the middle ear, filling in the empty cavity around the ossicles and then eroding the bones themselves (ossicles, mastoid). Cholesteatoma is often infected and results in chronically draining ears. It also results in hearing losses and may even spread through the base of the skull into the brain. Nowadays, the most effective treatment of cholesteatoma is to surgically remove the infected tissues through a minimally invasive procedure. Therefore, there is a real need for a minimally invasive robotic system able to access the epitympanum cavity, with high accuracy and dexterity.

This project is part of the Challenge 4 – Life, Health and Well-being of the ANR call. It will focus in a surgical protocol breakthrough for the middle ear diseases through basic research in robotics, microrobotics, differential diagnosis methods, and image-guided interventions, following a cross-disciplinary approach.
The objectives of µRoCS are the drastic reduction of recurrence (50% to 10%) and of aggressive wall-down procedures (the most commonly used procedure). Therefore, this project proposes a novel integrated robotic system that will exhaustively and efficiently remove the cholesteatoma, especially in the hardly accessible area located behind the mastoid bone (epitympanic cavity). Ideally, the proposed system will travel through the ear canal, and enter the middle ear by a small incision below the eardrum and/or via a small access tunnel drilled through the mastoid bone.

The proposed surgical tool (used to remove the cholesteatoma) will consist of a flexible microrobot (based on a continuum kinematic and less than 2mm in diameter) able to reach any part of the middle ear thanks to a high bending radius of its distal end. The specificity of this novel system is a free channel through which the surgeon can pass either a fiber-based imaging system and a surgical laser. Furthermore, the microrobotic system and the multimodal diagnosis system will be integrated (macro-micro approach) into the otologic robotic system robOtol (marked CE in 2016) for preclinical experimentations.
Within µRoCS, we will also develop a multimodal in-situ tissue characterization technique which consists of fluorescence spectroscopy coupled with an optical coherence tomography. It will offer to the physicians the possibility to verify in real-time (during the intervention) if the tissue is keratin, normal or pathological one. The diagnosis system will allow acquiring at each position of the robotic arm, a 1-D signal that can be considered as an optical biopsy. The collected data will be used as inputs to real-time classification models in order to recognize the residual cholesteatoma cells. Eventually, again using the robotic arm, the surgeon control the laser probe at the detected pathological cells for vaporization, thanks to an original image-guided control laws.

Finally, through the cited medical goals, breaking with current practices in the cholesteatoma surgery, µRoCS will be a project with a high scientific potential, which can lead to significant breakthroughs (minimally invasive surgery of the cholesteatoma removal, zero risk of recurrence and reoperation). In addition, this may open new perspectives for the management of other tumors of the middle ear (eg. Paraganglioma) and even allow to approach differently the internal auditory canal and the cerebellopontine angle (auditory schwannomas, meningiomas). µRoCS will provide high-level scientific contributions: image guided surgery and surgeon-robot interface, micromechatronics design, biomedical images-based control schemes. Moreover, technological innovations will have the potential to be transferred to other industries.

Project coordination

Laurent Tavernier (CTRE HOSPITALIER UNIVERSITAIRE BESANCON- Service d'ORL)

The author of this summary is the project coordinator, who is responsible for the content of this summary. The ANR declines any responsibility as for its contents.

Partnership

CHRUB CTRE HOSPITALIER UNIVERSITAIRE BESANCON- Service d'ORL
INSERM REHABILITATION CHIRURGICALE MINI-INVASIVE ROBOTISEE DE L'AUDITION (INSERM)
FEMTO-ST Franche-Comté Electronique Mécanique Thermique et Optique- Sciences et Technologies
ISIR Institut des Systèmes Intelligents et Robotiques

Help of the ANR 686,810 euros
Beginning and duration of the scientific project: January 2018 - 48 Months

Useful links

Explorez notre base de projets financés

 

 

ANR makes available its datasets on funded projects, click here to find more.

Sign up for the latest news:
Subscribe to our newsletter