What is the idea behind the book?

The idea is that this book will serve as an important reference for researchers and practitioners in FMI for years to come. Think, if you like, of the function of the book "Rule Based Expert Systems: The Mycin Experiments of the Stanford Heuristic Programming Project" by Bruce G. Buchanan, Edward H. Shortliffe (Editors) from 1984. Although now outdated, it has consolidated the research in Rule-Based Expert Systems and has served as a monumental reference for many years.

What is the coverage of the book?

The book is not meant to be complete. It is also not meant to include (only) the essentials in FMI, so it would not be entitled "Essentials in FMI". It is more in line with "Elements of FMI" in that it selects fundamental research issues that are important for FMI.

What are the themes of the book?

This will depend on the workshop and on the teams that will work on its chapters. Just to give you an idea of the themes, consider the following categories:

Inductive-logic programming in BioMedicine;
Using logical techniques e.g. in designing terminological systems, diagnostic systems, and for checking correctness of guidelines;
Supervised learning of models for diagnosis and prognosis prediction in medicine;
Unsupervised learning in biomedicine;
Modelling of biomedical knowledge using Bayesian networks, logic or both;
Modelling, learning and reasoning with causal and temporal patterns;
Mathematical modeling of biological function and structure

These categories could be further clustered. For example, if the emphasis is put on logical representation, then it might be under the heading of "Logic Engineering in Biomedicine". Many of the topics belong to machine-learning research. In some topics, the emphasis is on clinical medicine, whereas in others the research is more basic biologically oriented. Understanding the best way to categorise this research is likely to develop during the workshop.

Who will write the chapters?

The most ambitious goal is to have different contributors work on the same chapter together, even though a chapter is likely to be based on different papers of the contributors. A chapter could correspond to a category (e.g. "Inductive logic programming in biomedicine") or a cluster (e.g. "Logic engineering in Biomedicine"). At the workshop we will need to assess the feasibility of this approach. From the abstract(s) that we will receive from you [as described below] we will come up with a preliminary suggestion of the chapters and co-authors. One or more co-authors will function as co-editor(s). Needless to say, the success of this approach depends on you as co-author/editor of a chapter.

Do I need to eventually contribute completely new material for a chapter?

No. At least not necessarily. The idea is to consolidate research in some theme. This may concern work that you have been involved with recently (e.g. last year) but also work that is perhaps more than 7 years old. The "consolidation" of a theme, which is the contribution of a chapter, consists of 3 elements:

the updating of older material (e.g. by including more recent developments);
putting your work in a standardized form (e.g. that helps the reader see the problems, challenges, solution approaches, applicability, outlook); and
team-work on a chapter by different authors to complement each other (the authors can define a conceptual structure under which they can "plug-in" their contributions).

What do you want from me NOW?

As preparation measures for the workshop and as possible source of material for the chapters we would like to have 1 (or possibly 2 papers) from you. Again, you do not have to submit a new paper specially for the workshop (papers submitted are not considered new publications). In addition, the papers do not have to be recent. It is important that the paper (or 2 papers) are relevant for the workshop themes and include the theme you would like to work on. The papers will be collected by us, assembled, copied and sent to all participants. Each participant can skim through the relevant papers in her or his area before the workshop. The bundle can be studied in more detail later once teams and a working method have been finalized. In order to facilitate communication we ask you however to accompany each paper with one A4-size abstract in which:

The content is 'updated' (by putting the published paper in perspective of more recent work by you or by others).
The dominant theme of the paper is clearly described (you can use for example categories and clusters appearing above, or add others) along with its relevancy for the workshop.
The problem, challenge, solution approach and application are briefly mentioned.
A concrete short statement is mentioned about your 'vision' concerning the theme and its relevance to the workshop. For example "This approach [formal verification of guidelines/computing probabilities in a BBN/sound modelling of pathofysiological phenomena/...] ... [will not work/is especially beneficial in/is not yet understood/...] ... [therapy planning/domains where causality is known/because of...].

How do you want me to send the paper?

Electronically please! otherwise please send the paper(s) asap per post to Ameen Abu-Hanna. If you can use an A4-wide Latex style (this is included in the standard LaTeX distribution; insert

\usepackage{a4wide}

after the \documentclass{article} declaration) with point size 10 then pls do so. Try not to choose papers of length > 50 pages.

When is the deadline?

11 October 2002. This would allow us to bundle and send the collection of papers + updated abstracts (with normal mail) just on time.

Where do I need to send my contribution to?

E-mail: a.abu-hanna@amc.uva.nl

Address:
Ameen Abu-Hanna
Department of Medical Informatics
AMC-University of Amsterdam
Meibergdreef 15
1105 AZ Amsterdam
The Netherlands