Marco Protti

Simulation of hepatological models: a study in visual interactive exploration of scientific problems

In many different fields of science and technology, visual expressions formed by diagrams, sketches, plots and even images are traditionally used to communicate not only data but also procedures. When these visual expressions are systematically used within a scientific community, bi-dimensional notations often develop which allow the construction of complex messages from sets of primitive icons. This paper discusses how these notations can be translated into visual languages and organized into an interactive environment designed to improve the users ability to explore scientific problems. To facilitate this translation, the use of Conditional Attributed Rewriting Systems has been extended to visual language definition. The case of a visual language in the programming of a simulation of populations of hepatic cells is studied. A discussion is given of how such a visual language allows the construction of programs through the combination of graphical symbols which are familiar to the physician or which schematize shapes familiar to him in that they resemble structures the observes in real experiments. It is also shown how such a visual approach allows the user to focus on the solution of his problems, avoiding any request for unnecessary precision and most requests for house-keeping data during the interaction.

Attributed conditional L-systems: a tool for image description

The problem addressed in this work is the construction of a description of a digital image using a kind of parallel rewriting system, the conditional attributed L-system. To this end, descriptions of digital images and digital structures are first defined. Then, attributed conditional L-systems capable of parallel rewriting are introduced. These systems are concise and readable, and their suitability for image description is shown by an example. Advantages and disadvantages of their use are discussed.<<ETX>>

Pattern-directed restoration and vectorization of digitized engineering drawings

Abstract An algorithm for the interpretation of raster data acquired by digitizing engineering drawings is described. The algorithm allows noise removal and the description of the original drawing in line format. The algorithm is based on the a priori study and exploration of the technical but not formalized language that engineers and draftsmen use to design and interpret the (analog) drawings. The derived information is stored in the form of rules in a modifiable data base and used as high-level knowledge by the restoration programs. The resultant vectorial description of the drawing may be used for an exact reproduction of the original drawings as well as operated graphically or archived and managed in a graphic data base.

Metareasoning in the determination of image interpretation strategies

Abstract Flexible determination of image interpretation strategies is made possible by reasoning on the state of the interpretation process. The state is here described as a string of attributed symbols denoting images, descriptions, actions and patterns of actions. Goals of the interpretation are specified by Conditional Attributed Rewriting Systems. An interpreter exploits these systems to interpret the image, i.e., to determine the evolution of the state from the initial state to the final interpretation. An example from liver biopsy interpretation is introduced, its results constrasted with those of human interpretations and the automatic interpreter efficacy commented on.

Knowledge-based contextual recognition and sieving of digital images

Abstract Sieves are operators which exploit disciplinary knowledge for a contextual elimination of structures hindering the interpretation of a digital image. A sieve is realized as a conditional attributed Lindenmayer-system with interaction, a generative tool defined in the paper.

An APL rule-based system architecture for image interpretation strategies

APL is both a language of choice for image processing and image description tasks, and a language for artificial intelligence applications, typically expert systems. The paper presents the architecture of an APL rule-based system for realizing description synthesis strategies i.e., sequences of actions which allow the evaluation and interpretation of the properties of structures detected in an image. Descriptions of images are stored in a data structure which is formally presented in APL2 syntax. Describing an image often requires that reflective actions be taken i.e., actions in which the system examines the state of the computation and its internal state to select the next action to be executed. It is shown how APL provides all the features needed to implement a reflective mechanism through the use of metarules of the same format as rules.

A Visual Environment for Liver Simulation Studies

The use of simulation tools is becoming more and more frequent in the everyday activity of the physician, at both researcher and practitioner levels.

A completely visual environment for agent-based computing

A visual environment for end-user visual programming and computing with populations of agents is presented.

Visual programming in a visual environment for liver simulation studies

The interactive system for hepatologist experimentation (ISHeE) is a visual explorative environment designed to allow a community of hepatologists to formalize, check, validate, and discuss their pathology and physiology liver models. In this environment the programming of simulation experiments is achieved by the use of visual languages. The use of conditional attributed rewriting systems is extended to the visual language definition, and the case of a visual language for programming the simulation of a population of hepatic cells is studied. How such a visual language allows the construction of programs which are expressed by the combination of graphical symbols which are familiar to the physician or which schematize shapes familiar to the physician and resemble the simulated system observed in the experiments is discussed. It is shown how a visual approach avoids the request of unnecessary precision during the interaction, leaving the user to focus on the solution of the problem.<<ETX>>

Metareasoning as a tool for pattern recognition

The paper introduces the concepts of image interpretation by form feature and interpretation strategy as a fall out of observing human interpreters at work. It discusses the need for flexible control in an automatic interpreter, outlining an automatic interpretation process and the required meta (or reflective) mechanisms.<<ETX>>