Objectives

A source of high-quality data of both morphological and kinematics models of human joints will be created. The data will be collected by novel techniques developed by the proposers. These techniques will allow data to be obtained that is of potential interest in related fields across industry, medical education and research. At present, little accurate or reliable data concerning joint modelling is available, which often leads to simplification. Once created, the data will be available via the Internet to allow users to gain new knowledge on functional anatomy. Raw data will be available for downloading to allow industrial users to improve their own products and competitiveness. The interface will employ state-of-the-art multimedia techniques and 3D simulations in a virtual reality environment to demonstrate joint behaviour. Electronic tutorials will be supplied to help give medical staff and students to a better understanding of functional anatomy.

Description of the work

The project is divided into several components; these are overlapping and consist of several general tasks described below. The Project Co-ordinator will supervise all aspects of the project and will monitor the circulation of the results and data between the partners. Discussion will take place with potential users throughout the development. 1. COLLECTING THE DATA. (a) Morphological data will be acquired for bones (including joints), and also for muscles and ligaments. Analysis will be made of a large number of musculo-skeletal structures and a variety of individual morphologies, including female and male subjects of several different ages. (b) Kinematics data will be collected experimentally in parallel with the morphological data using novel methods. 2. BUILDING THE COMPUTER MODELS. (a) Morphological models will be created using modern 3D reconstruction techniques including geometrical surface modeling and finite element methods. (b) These models will be animated using the kinematics data (see 1b). 3. DEVELOPING THE SERVICE TOOLS. Two forms of service will be offered to the end user. Industrial users will want access to the raw data (morphological, finite elements, kinematics) - this will be of a quality currently not available. Educational users will have the opportunity to interact with 3D joint models; they will be able to select the level of educational support supplied with these interactively. They will also be able to view tutorials presented by eminent physicians discussing and demonstrating interesting cases and to test their knowledge using on-line "questionnaires". 4. DISSEMINATION. Demonstrators will be presented to potential users who have expressed interest. They will be invited to a meeting organised as part of the Accompanying Measures Programme. 5. EXPLOITATION. Exploitation of the data will be performed by Engineering Systems International France s.a. (Partner 7) for its specific needs and by academic partners (University of Brussels, Istituti Ortopedici Rizzoli, De Montfort University, University of Sassari) for educational and research purposes.

Milestones and expected results

Accurate modelling of joint kinematics will allow us to gain new knowledge and a better understanding of joint behaviour. The quality of the expected results should provide an opportunity to develop unique services (described above) for the potential user. This project will focus on the development, dissemination and the initial exploitation of the complete method and will apply it on the lower limb. Further developments of the database are planned for the future.

Accurate knowledge of human functional anatomy is of importance in several fields, including medical education for undergraduate and postgraduate students, retraining of medical professionals, clinical research, biomedical engineering and other industrial areas, sport and entertainment. The three-dimensional aspects of functional anatomy are particularly important, but there has been a tendency either to neglect or to simplify them because reliable data about both normal and pathological anatomy has not been available. There has been a similar lack of tools suitable for such demonstration.

Approaches to anatomical data have been re-appraised recently as a result of the Visible Human Project of the National Institute of Health (NIH, USA). This rich source has created a considerable "industry" of projects that utilise the data in widely-varying ways. However, the data relates only to one male and one female subject, neither of which contains specific pathologies; no kinematic data was stored.

An alternative data source is that of a large European consortium IAEVA, funded by the EC Telematics Applications Programme. Here, a large set of three-dimensional models of pathological organs, together with clinically-important information such as radiological slices, commentaries from clinicians, and animated "fly-throughs", are stored in a distributed database to support medical education. A commercial service will be widely available at the completion of the project. While this project has demonstrated that modern tools can provide new insights into three-dimensional anatomy, it has considered only static situations.

Objectives

Thus, despite their value in other areas, both NIH and IAEVA projects have a limited use for those with an interest, whether educational, clinical or industrial, in functional anatomy. The objective of the current proposal is to find a solution to these limitations:

• No information about kinematics is available in either project. Joint behaviour and pathology are widely acknowledged to be strongly related to joint kinematics – greater understanding of this area will lead to improvements in treatment of joint disease and in related industrial developments (e.g. joint prostheses).
• Neither project includes finite-element models, which are extremely useful for bioengineering simulation.
• Some projects, e.g. at Purdue University or the CHARM project (ESPRIT 9036), have used the NIH data with simplified or artifically-created kinematics data; little attention was paid to combined motions though these are fundamental to joint physiology, so the results produced are indicative rather than functionally accurate.
• The NIH project raised serious ethical issues by the way in which the male cadaver was obtained; a number of scientists refuse to use this data because of their ethical objections.
• The medical imaging datasets available from the NIH project are not always of the best quality, especially for the limbs. In the recent past, two of the proposers of this project have made available, through the Internet and free of charge, high-quality data for several bones. Feedback received suggests that these are widely used. As this was done on a voluntary basis, these services are limited to a few datasets.
• The NIH project made available only raw data, which must then be processed to obtain a usable product. While industrial companies or universities have the facilities to process the data to create "products" suitable for their requirements, most individuals do not have the resources to do this, so the data provides relatively little information for them.

While the NIH project provides only the raw data, the IAEVA project presents a broader service. The user can access directly a variety of forms of information concerning the cases presented, but not complete sets of data.

We intend to combine the philosophies of the two projects by providing a system to support education at different levels ranging from undergraduate to experienced practitioner, while make original datasets available for industrial users. Our general objectives will focus on each of the following points:

• The database will contain the raw data, a range of models created from the data, and tutorials on functional anatomy that will be developed for medical education and training by partners with recognised experience in that field. Improved learning will be made possible by use of a rich multimedia environment. Users will be able to select, interactively, the type of information and level of educational support they require.
• The raw morphological and kinematics data available from the database will support the innovation of improved anatomical or ergonomic products (e.g. implantable medical devices, car crash simulation, body contouring products, etc.). Thus, European companies will be able to buy high-quality data directly from European providers.
• 3D models of joints will be available, together with kinematics data. Experimental collection of the data will employ an accurate and robust methodology to be developed by the consortium. Data concerning muscles and ligaments will be collected for future use, though it will not be employed during this proposal.
• 3D models will be available in two formats – geometrical and finite element.
• All data collection will respect current European and national ethics. Data collection on cadavers will take place at an institution which is empowered to do it, under control of an ethical committee.
• Medical imaging will take place in a state-of-the-art installation by highly-trained technicians. This will ensure that the collected data is of the highest quality.
• A multilingual interface will be provided to enable users to access the service in their national language.

This proposal will focus on the Research and Development aspects of the above objectives and will develop a model for the lower limb, chosen because of it is frequently involved in clinical and industrial research.

Our quantifiable objectives are the following:

• Kinematics data will be collected from two sources: 3D-electrogoniometry for accurate measurements of individual joints, and a marker-tracking video system for full-limb segment kinematics. This latter data will be used to animate morphological models of the major lower-limb segments by matching a number of common anatomical landmarks. The combination of joint and segment kinematics will also be investigated to allow possible simultaneous and accurate modelling of the limb – the challenging problem to be solved is the registration of the recorded motions to allow this combination. The movements to be analysed initially will be common motions performed in every day activity: walking, running, sitting, etc. The data-collection process will be fully documented and each method involved will be quantified in terms of accuracy and repeatability. Thus, at the end it will be possible to declare the class of spatial or spatio-temporal accuracy of each data set – this information will be critical in many applications that will use the data.
• Initially, we will make available 3D models of all major bone components of the lower limb from a male and a female of each of the following ages: young adult, elderly adult and adolescent. Finite-element elements of the same bones and the major muscles of the lower limb will be processed to support future development. We will need multiple anatomical instances of at least one of these to validate the kinematics data collection and registration – we aim to perform this validation on at least 10 specimens.
• A database will be created to store the data. It will allow easy retrieval of the data using standard database search features. An interface will be create to allow downloading of both the 3D models (kinematics and morphological data) and original medical imaging slices.
• Tutorials on functional anatomy will be developed to explain the kinematics using the knowledge brought by the data-collection method. Highly-realistic 3D models of the joint motion will be integrated into a multimedia environment. The first set of tutorials will be in English, French and Italian.
• Feedback from all types of prospective user is considered critical to the development and subsequent user-acceptance of the system. Indications of support have already been received from professional groups, possible industrial users, relevant charities, firms in the computing industry and publishers (on-line and off-line). It is recognised that some of these user groups may find difficulty in fully defining their requirements until a prototype system has been seen. Informal feedback will be sought from the medical profession at all stages of development, and a full demonstration of the system is planned (using IST Accompanying Measures) to introduce the system to end-users and to prepare exploitation of the results. Two demonstrators will be available at that time: an "educational" tutorial on the lower limb including interactive simulation of joint kinematics and on-line assessment; and an "industrial" demonstration using kinematics data and finite element models to visualise tissue deformation, stresses, etc., during dynamic activity.

We plan to achieve these goals in a 24-month period. It is anticipated that a follow-on proposal will be introduced to continue the exploitation of the current proposal, to broaden the range of data available within the database, and to extend the modelling to the muscular and ligament system (the data for this will be collected during the first period to avoid duplication of effort).

List of Participants

Last update: 31 Nov 1999