[in Japanese]

View full abstract

A fuzzy theory is applied to thinking the facial differences between the two racial group-types, the Jomon (more than 2000 years ago, RI=1.000) and the Kofun (about 1500 years ago, RI=2.000). Many modern Kumamoto men show the racial index (RI) of the Kofun type as well as the modern Kanto men and Tokugawa Ienobu. The Yayoi men show the middle value, but RI of the Medieval men is close to that of the Kofun. One of the Kumamoto Yayoi men and two modern Kumamoto men are very close to the Jomon type.

View full abstract

Objective of this paper is to find capabilities to realize a face animation system, combining the emotion processing system developed in our laboratory in which the investigated useful results concerning the mixed emotions by the well-known psychologist, R. Plutchik are taken into account. The face expression is done by the combination of the movements of eyes, eye-brows and a mouth. And each of the movements is drawn by Bezier function whose control points can be extracted from the resultant image codes after the emotional processing.

View full abstract

This paper describes identification a lanclmark;the point which becomes charactrisitic of human's face, from the contour of the skull using neural networks which have four layers in automation of cephalometric analysis.

View full abstract

At first, we introduce Frame Oriented Theory, and explain the basic concept of the theory. We can classify a certain thing or concept into the constructing attributes (items) in general, it can be represented by the matrix whose values on each item is 1 (if the thing reside on the item) and 0 (if it doesn't reside the item). Then the matrix represented in this, way is regarded as a kind of image patterns meaning of the thing or concept. Thus the original matrix consisted of only the items is called by "Frame". Next, we describe the vectors derived from a certain Frame, giving some examples to help easier understanding. Then, we show an application of the theory to the processing for the sentences on medical diagnosis, and show some interesting results of the experiment. This research has just been started, then it remains a lot of thing unstudied. Furthermore the theory is applicable to other scientific fields besides this kind of medical ones.

View full abstract

The purpose of this paper is to introduce a basic concept of the Emotional Interactive System. We discuss here the dialogue between two simulated persons who knows only the emotional words. At first, we show an example of Image Code Table (ICT) concerning Mixed Emotions which are regarded as general emotions in usual life, introduced by R.Plutchik, and also show that the values of ICT are produced based on 8 pure emotions. Second, we refer to the theory for Subjective Observation Model (SOM), which have high capability to several kinds of fields. And last, we propose the basic concept of Emotional Interactive System based on the theory for SOM. We consider a number of SOM as a kind of cognition (or interpretation). The system has two local systems (conditional Euclidean Space) corresponding to the two persons on dialogue. The emotional words on each local system dispatched to the partners are mapped on the particular observation space on which the origin of mapping and the scaling are taken into special consideration.

View full abstract

Facial esthetic improvement is one of the goals of orthognathic surgery. When facial esthetics as a result of surgery is appraised, facial profile is especially paid attention. One of the appraisal method of profile esthetics is a surgeon's subjective estimation, the other is soft tissue profile analysis in which several linear and angular values measured on a lateral cephalometric radiogram are simply compared with the standard values. However, neither ones have an objectivity for the esthetic estimation of an individual profile, We are developing a profile esthetics appraisal system which takes account of fuzziness in evaluating facial esthetics using an expert system.

View full abstract

This paper presents an identification result of a pulse rate prediction model for activity sensing pacer via SOFIA (Self-Organizing Fuzzy identification Algorithm), which is a modeling method for nonlinear complex systems. Input variable and output variable of the pulse rate prediction model are workload and pulse rate, respectively. It is difficult to predict pulse rate because of nonlinearity between the input variable and the output variable. As a result of cornpairing identification result with those of other pacers, it is shown that SOFIA is useful for predicting pulse rate.

View full abstract

This study was designed to introduce fuzzy set type data-base for health balancing. In this study work load and nutritional balances were dealt with in a process of constructing a program, which served to calculate nutritional value of food intake and energic output by work load. We applied multidimentionally weighted selection of appropriate combination of foodstuffs defuzzyfied from fuzzy set type data-base. Optimal nutritional volume against work load was also determined by fuzzy reasoning. This method will be extended further to program to find a way balancing motor activity, care of health or medical choice.

View full abstract

Stroboscopy is a major means of making diagnosis in phonetic surgery. It requires a certain experience to diagnoze each case, not to mention the tecnique. Determining the course of treatment, for white lesion in particular depends highly on the observation procedure. Some hospitals employ detailed analytical chart for stroboscopic judgement. However a visual observation of the movement with naked eye is sufficient. Although confined to a prognosis of white lesion, we were able to put our experience into an application soft for personal computers using fuzzy logics. The results were classified into four levels and the validity of the soft was judged on its results 'coincident rate with those of experienced doctors'. The results of 44 cases were obtained and the coincident rate was 73%.

View full abstract

In clinical cytology, cytotechnologists and cytopatholegists make their diagnosis by comparing the features of observed cells with the criteria of malignant cells. In this cell interpretation process, observed cell features do not always coincide with the criteria by the term only "+" or "-". It also contains "±". Thus, this process is considered as "fuzzy" sample. In this study, we intend to produce the automated program for cytodiagnosis of uterine endometrial abnormalities using 17 criteria which we found in the retrospective study. Unfortunately, we do not succeed in findig the membership function for this program, so far.

View full abstract

Patients of HTV (Human Immunodeficiency virus) are increasing more than 2 times in recent Japan. HIV causes acquired immunodeficiency syndrome (AIDS), and also is causing "National tragedy" by its incurable death dependent disease. We predicted the Japanese AIDS disaster using fuzzy theory. Two major membership factors such as effect of education and effectiveness of propagation were input. The computer program using fuzzy theory predicted long period of AIDS disaster as if the "National tragedy" would be the truly occurring event.

View full abstract

The corrosion cast of the lung vascular system is reported. Chaos related with the changing from the vascular branching tree to the capillary network in the lung is discussed.

View full abstract

This paper discusses a topic whether max and rain operations used in fuzzy reasoning methods such as min-max-gravity method are suitable operations. It is shown through several examples that max operation is not suitable when it is used in the aggregation of fuzzy reasoning results. Moreover, alternative operations are recommended to use in fuzzy reasoning methods.

View full abstract

ECC is extremely important to cardiac surgery. We previously attempted the development of a microcomputer-based auto-regulation system for ECC, and fuzzy theory is applied to implement the system. Recently, we are developing a clinical model for automatic regulation of ECC which automatically controls blood flow volume. In this paper, we describe the control theory of the artificial heart-lung circuit, and the operating environment of the regulatory apparatus.

View full abstract

We have applied the diagnosis method proposed by D.Norris, B.W.Pilsworth and J.F.Baldwin to the diagnosis of valvular heart diseases. This method uses concepts from fuzzy set theory and consists of two independent methods: discrimination analysis and connectivity analysis. We performed the experiments in order to evaluate effectiveness of the proposed diagnosis methods. Also, we extended the original method to handle partial manifestation of symptoms and severity of diseases by using fuzzy sets. In Addition, we introduced the concept of prototypicalness of patients with a particular disease to improve the performance of the diagnosis. The results of the experiments are very promising for the discrimination analysis. In the best case, we achieved a rate of true positive diagnosis of 81% while maintaining a rate of false positive diagnosis at the low level of 10%. The connectivity diagnosis method proposed by D.Norris, et al. produced disappointing results in our experiments. In exploring the reasons for this, we designed a new algorithm which is both more intuitive and more effective. We report the quantitative results of the experiments.

View full abstract

Correct assessment of cervical lymph node metastasis is very important in the management of head and neck cancer. Although CT scan is beneficial to screening the metastatic cervical lymph nodes, CT scan diagnosis is usually based on a subjective assessment of the findings by the examiner. Subjective assessment differ from examiner to examiner/and is affected by first impression of examiner. It is possible that such subjective assessment lead to different results in CT scan diagnosis, even if the same diagnostic criteria are used. We have developed a computer diagnostic supporting system using fuzzy theory, which allows objective assessment of cervical lymph nodes metastasis in head and neck cancer. We used six items as the degree of variation in CT findings of lymph nodes which were included ring enhancement, low density area, internal density, size, shape, and margin of lymph nodes. Each item was measured using a "fuzzy scale" for estimating the degree of variation of features analogically. Fuzzy theory was adopted for the diagnostic logic. The input and output membership functions were prepared for each item and the final assessment was made by the center of gravity of the max-sets of the output membership function. Our diagnostic supporting system appears to be beneficial for medical educational system as well as an adjunct diagnosis for metastatic cervical lymph nodes.

View full abstract

We discuss a method of fuzzy clustering which is recently created by us. The effectivity of the method is verified not only in theoretical way but also by the computer simulations using the test samples in two dimensional case and also the sample derived from eastan medical diagnosis in five dimension. It is enphasized that the resultant dendgrams by this method can represent well how the samples closely distributes rather than the several traditional methods based on Euclidian distance.

View full abstract

A diagnosis assisting system of bone tumors was constructed using a relational data base management system (RDBMS). Radiographic findings of various tumors were categorized at every item and stored as degrees of possibility. Possible diagnoses of an unknown lesion can be listed by max-min calculation between the data base and findings of the lesion. Integration of items narrow the possible diagnosis down. Expansion of the system is expected as addition or renewal of data is easy on RDBMS.

View full abstract