Health Telematics (AIM) Final Report


	Updated: Feb 1, 97
CAMI

Computer Assisted Medical Interventions

Project Code: A2056

Project value: 550 KECU

EC contribution: 275 KECU

No of partners: 4

No of countries: 3

Duration: 18 months

Contact:
Prof. Philippe Cinquin
Université de Grenoble
Institut Albert Bonniot
Domaine de la Merci
F-38706 La Tronche-Grenoble, France
Tel.: +33-76-54.95.07/54.95.06
Fax: +33-76-54.95.55
E-mail: philippe.cinquin@imag.fr

Overview

CAMI has developed a pre-industrial demonstrator for computer assisted spinal surgery. The general scope is to perform more accurately and less invasively a previous defined operative strategy, thanks to guiding systems under the supervision of intra-operative sensors.

Purpose and objectives

CAMI uses a general methodology for Computer Assisted Medical Interventions in three steps: Perception - Decision - Action.

Perception is performed both pre-operatively and intra-operatively. The relevant information comes mostly from medical imaging devices: Computed Tomography (CT), Magnetic Resonance Imaging (MRI), Digital Radiology, Ultrasound Imaging (US) etc. Information coming from other types of sensors borrowed from Computer Vision, such as video cameras or range imaging sensors, is also available. Finally, since they contribute a priori statistical knowledge, geometrical and anatomical rules and models (such as Atlases) have to be taken into account.

Decision is an important step involving both construction of patient models and intervention planning. Construction of patient models means merging all the available information in order to build a "virtual patient". This implies accurate registering of all information sources to make optimal use of each one. Intervention planning is the modelling of an intervention and the simulation of its morphological and functional consequences. The result of this simulation is the selection of an optimal strategy.

Action consists in guiding the selected strategy while it is performed. When intra-operative sensors are combined with the already created "virtual patient", the operator is provided with "augmented reality" which is an effective combined mixture of the real and virtual worlds. Different levels of assistance exist to perform a selected strategy, including robots in some cases. In all cases, safety is essential and requires massive redundancy of sensors, processors and actuators.

Results

Clinical results have already been obtained on 5 patients in the department of Orthopaedic Surgery at Grenoble's University Hospital (Pr. P. Merloz). These validations were performed on patients in whom a pedicular screwing was required (this intervention consists in introducing a screw inside the pedicle of a vertebra, a narrow pass between the anterior body of the vertebra and its posterior arc). This intervention is required in various clinical situations including traumatisms, vertebral instability and scoliosis. CAMI methods allowed precise guidance of the screw.

Feasibility results on elastic registration methods, developed by University Joseph Fourier of Grenoble, between atlases obtained by Pr. D. von Keyserlingk and MR data of the brain were obtained. These results show the potential interest of these elastic registration algorithms.

The previously described methodology presents a very interesting feature for industrialisation. It is truly "generic", which means that the same "core" of tools may be applied in very different clinical situations. This has clearly been established in previous feasibility demonstrations. A start-up company, PRAXIM, has been created for the industrial exploitation of these results. Industrial agreements have been reached with major European industrial partners to develop industrial versions of the techniques that were validated in CAMI.

The results that were obtained provide a sound basis for further development, and we plan to be shortly in a position to perform interventions where direct vision will be replaced with X-ray and ultrasound images.

Finally, the same sort of methodology may be applied to many other orthopaedic and intra-radiation therapy situations, and we have reached industrial agreements with major European partners to do so.

List of Deliverables

Year 1

Intra-operative information acquisition specification
Real time pre-operative and intra-operative information matching specification
Specification for assistance to the performance of the strategy
Report about elastic matching
Prototype of elastic matching algorithm
Prototype of operating tools calibration system
Prototype of advanced user interface for guiding interventions
Prototype platform for CAMI

Year 2

Demonstrator for CAMI
Report about the validation of the demonstrator applied to spine surgery
Report about the pre-clinical validation of the demonstrator applied to electrophysiologically based neurosurgery
Final report

List of Participants

Prof. Philippe Cinquin Université de Grenoble Institut Albert Bonniot Domaine de la Merci F-38706 La Tronche-Grenoble, France Tel.: +33-76-54.95.07 / 54.95.06 Fax: +33-76-54.95.55 E-mail: philippe.cinquin@imag.fr	Dr. Frans A. Gerritsen Philips Medical Systems Nederland B.V. ICS Advanced Development Integrated Clinical Solutions Veenpluis 4-6 / Building QP-1105 NL-5680 DA Best, The Netherlands Tel.: +31-40-276.23.81 Fax: +31-40-276.26.73 / 276.33.93 E-mail: frans.gerritsen@ms.philips.nl
Dr. Diedrich Graf Von Keyserlingk Medizinsche Einrichtungen Institut für Anatomie Paulwelsstrasse 30 D-25057 Aachen, Germany Tel.: +49-241-808.91.00 Fax: +49-241-808.89.33	Mr. Philippe Boule Aleph Med Rue du Tour de l'Eau 16 F-38400 Saint Martin d'Heres, France Tel.: +33-76-42.29.99 Fax: +33-76-44.46.20



	Copyright 1997 © EHTO All rights reserved This server is the only official EHTO WWW knowledge repository. Mail suggestions to: webmaster@ehto.org