Health Telematics (AIM) Final Report


	Updated: Feb 1, 97
IMPACT

Integrating Microscopy for Pathology Activities and Computer Technology

Project Code: A2017

Project value: 7636 KECU

EC contribution: 3751 KECU

No of partners: 7

No of countries: 5

Duration: 36 months

Contact:
Prof. Gerard Brugal
Université Joseph Fourier
Institut Albert Bonniot
Faculté de Médecine
Domaine de la Merci
F-38 706 La Tronche Cedex, France
Tel.: +33-76-54.94.00
Fax: +33-76-54.95.49
E-mail: gerard.brugal@cyf-grenoble.fr

Overview

The ultimate goal is to provide the pathologists with the computer technology they repeatedly ask for with the purpose of : improving reliability of diagnostics and prognostics of cancers, sharing experience among the medical community so that a standard expertise will be available to all concerned (remote consultation), while increasing the quality (by built-in consistency controls).

Purpose and objectives

In modern hospitals, the pathology laboratory plays an essential role in the detection, diagnosis and prognosis of most serious diseases and especially cancers. Medical decision making is carried out through microscopic examination of biological specimens by a pathologist who reports his opinion back to the requesting therapist. Unfortunately, this essential opinion is always subjective and qualitative and, for the most difficult cases, the lack of reliability and reproducibility may result in somehow inadequate treatment of the patient. It was identified that improving the quality of screening, diagnosis and prognosis of cancers could only be expected through a full integration of microscopical observations and clinical management of patients. This required specific developments of (i) computer based image analysis of medical specimens to introduce objectivity and reproducibility in the assessment of biological lesions, and (ii) telematics technologies for remote consultation and sharing of expertise. The prime objective of the project was thus to build such interactive computer facilities and image processing tools in the microscope but without changing the way this instrument is used. This was the mandatory condition of acceptance by the medical community.

Results

In order to achieve the integration between microscopy and computer facilities, the technical approach has been multidisciplinary and combined:

basic research, carried out on university laboratory sites and putting the emphasis on the knowledge based approach to the medical behaviour and decision making at the microscope;
system developments carried out on industrial sites and putting the emphasis on the construction of the imaging microscope workstation;
medical evaluation carried out on pathology laboratory sites and putting the emphasis on the assessment of the software packages developed for (i) cancer screening, (ii) differential diagnosis based on morphometry, (iii) assistance to diagnosis reporting and (iv) education and self-training of medical actors.

The medical constraints were met by introducing a miniature computer monitor in the microscope optical pathways thus overlapping the optical conventional image with the computer generated images. The microscope was thus driven by a "mouse" which cursor was flying over the observed image and served for making zones of interest, for activating image analysis procedures on specific cells and tissue structures and for selecting standardised relevant diagnostic and prognostic clues by "clicking" on the corresponding items. Also, miniature transducers were placed on both the stage and the objective turret of the microscope to feed the computer at any time with all microscope settings. In order to bridge the gap between conventional practice in pathology and computer technology, the feasibility of direct interfacing between microscope and computer had to be analysed in depth with respect to (i) medical motivations and requirements (ii) eligibility of biological specimens measurements and (iii) current practice at both the laboratory and the medical personnel levels. Such studies provided the key for interface design.

The major achievement of the project was to turn the conventional microscope into a workstation which can be connected on a LAN so that any decision made at the microscope is integrated in real time with the other clinical information available in the hospital files. Consistency can then be checked on line as well as microscope observation operational features. Medical quality thus relies on clear figures such as the proportion of the specimen actually observed, the number and meaning of marks used by the observer, the intra- and inter-observer reproducibility and the overall consistency between the microscopical findings and the clinical data. Such a microscope computerised workstation can also be connected on a WAN so that the observation of a specimen can be carried out by a remote expert actually driving the microscope from his place. Also, medical information created at the microscope level can contribute to transnational data bases and image reference banks expected to make possible large scale clinical studies.

The application software developments resulted in the specifications and software for 4 graphics and 4 image medical applications running on the IMPACT workstations and which are: Cervical cancer screening-reviewing, Morphometry, Diagnostic assistance and Self-training (for GRAPHICS systems); and Histometry, Ploïdy, Immunomarkers and Cytogenetics (for IMAGE systems).

Medical evaluation

Although the medical evaluation is still ongoing, some objective figures already obtained will not significantly change. For example, a remarkable 17 % increase in quality has been demonstrated for cervical cancer screening; a decrease in processing time by factor 6 was measured for morphometry of cells and tissues and a decrease in analysis time of ploidy by factor 2.5 was obtained for breast cancers.

Exploitation of Results

The project has directly and entirely put out three product lines which are now commercially available from Carl ZEISS (Germany) and ALCATEL (France) : (1) Networked Interactive computerised microscope; (2) Networked computerised microscope for image processing and (3) Integrated microscope "telepathology" workstation. These systems have received an enthusiastic feedback from the medical community. About 50 systems have been distributed in European and North and South American Pathology laboratories. Because of its man-machine interfacing which perfectly meets the medical requirements and constraints, this new generation of instruments is easily assimilated by the end-users and makes a great step forward in the contribution of computer and telematics to quality control and quality assurance in medicine. Exploitation plans are presently implemented by international companies whose marketing strategies are well targeted.

The main obstacles

Obstacles for a fast take take-off are sometimes the marketing strategy of an industrial company and also the shift in the scope of the main use of the products by the user community during the time of the project.

List of Deliverables

Year 1

IMPACT Graphics specifications (4/S/R)
IMPACT Image specifications (8/S/R)
IMPACT Graphics language (9/R/P)
IMPACT Networked graphics workstation (12/P/P)

Year 2

IMPACT Graphics evaluation (6/R/I)
IMPACT Image language (8/R/P)
IMPACT Graphics for daily routine (12/P/P)

Year 3

IMPACT image workstation (1/P/P)
IMPACT Image evaluation (6/R/I)
Integrated Pathology Laboratory (9/P/P)
Computer Assisted Pathology Today. Review Project Report (9/R/P)
Computer Assisted Pathology Today. Final Project Report (12/R/P)
Demonstration of Efficiency in Medical Operational Services (12/R/P)

List of Participants

Dr. James H. Tucker Department of Pathology The University of Edinburgh Medical School Tevlot Place Edinburgh EH8 9AG, U.K. Tel.: +44-131-650.28.76 Fax: +44-131-650.65.28 E-mail: jimt@hgu.mrc.ac.uk	Dr. Bernd Faltermeier Carl Zeiss Jena GmbH Microscopy Division Tatzendpromenade 1a D-07745 Jena, Germany Tel.: +49-3641-64.20.62 Fax: +49-3641-64.33.47 E-mail: faltermeier@zeiss.de
Mr. Didier Adelh Alcatel-TITN-Answare Samba Division Immeuble Epilobe 35 Chemin du Vieux Chêne F-38240 Meylan, France Tel.: +33-76-41.71.72 Fax: +33-76-41.06.31	Prof. Peter G. Vooijs Cytology Research Group Toernooiveld 138 NL-6525 EC Nijmegen, The Netherlands Tel.: +31-24-352.88.46 Fax: +31-24-354.00.90 E-mail: g.vooijs@fmwbes.kun.nl
Mr. Gerard Slavin Sultan Qaboos University Actg. Head of Department Dept of Pathology PO Box 35 SQU Al Khoud PC 123, Sultanate of Oman Tel.: +968-514529 Fax: +968-513419	Prof. Claes Lundsteen Rigshospitalet Genetics Section Blegdamsvej 9 DK-2100 Copenhagen, Denmark Tel.: +45-35-45.48.85 Fax: +45-35-45.48.92



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