In colloid liver scintigraphy, diagnosis of diffuse parenchymal liver disease such as chronic hepatitis or cirrhosis is evaluated by the size and distortion of the liver, distribution of tracer in the liver, size and activities of tracer in the spleen, visualization of the bone marrow and so on. It is not difficult to read a scintigram which shows a typical pattern ofnormal, chronic hepatitis and cirrhosis. But in some cases it is difficult to read normal or chronic hepatitis and chronic hepatitis or cirrhosis in visual diagnosis, so we tried to use fuzzy reasoning to perform differential diagnosis in chronic hepatitis (CH), bridging fibrosis (BF) and liver cirrhosis (LC). First, five features in colloid liver scintigrams were evaluated visually. These features were left lobe/right lobe, splenomegaly, visualization of the bone mar-row, liver deformity and distribution of tracer in the liver. Having fuzziness in these data, the tendencies of these features, ″small″, ″medium″ and ″large″ for instance were thought fuzzy sets and displayed in membership functions. Fuzzy reasoning was carried out using these data and fuzzy rules. Using fuzzy reasoning, differential diagnosis in LC could be performed by 100%, but CH and BF could not be differentiated completely. So using neural network CH, BF and LC could be differentiated reasonably well, and therefore combining fuzzy reasoning and neural network better differential diagnosis could be expected.

View full abstract

This study investigated the validity of the possibility theory used to minimize risk in making a diagnosis and in differential diagnosis. We constructed a computer-assisted diagnostic system for long bone lesions using a relational database management system, in which the radiographic findings of all parameters of various lesions were categorized and stored as degrees of possibility. With this system, possible diagnoses were listed by max-min calculations between the possibility values in the database and findings of the lesions. Integration of parameters by products of their values for each disease provided more accurate prediction of diagnosis than min operation. This computer-assisted diagnostic system based on the possibility theory proved useful for suggesting differential diagnoses for long bone lesions on plain radiographs.

View full abstract

Human beings are more subjective than objective. In this paper, we introduce Frame Oriented Theory, and explain the basic concept of the theory at first. The original matrix consisted of only items is called a ″Frame, ″and an arbitrary matrix whose values are real number (or fuzzy number), based on the Frame is regarded as a kind of image patterns of a thing or concept. Next, we show an application of the theory to decision processing and evaluation of psychiatric diagnosis. We constructed a psychiatric diagnostic frame using 10×6 matrix, and consider how a patient may be diagnosed by several psychiatrists who have individual subjectivities. Lastly, we suggest that it remains many things unstudied not only in theoretical consideration but also in the applicability to several kinds of scientific fields.

View full abstract

In this paper we propose a Fuzzy Artificial Intelligent System with learning function, which is called ″FAIS″short for. This system has functions which acquire new fuzzy knowledge from experiences under an unknown situation and solve some problems by deducing a new state through fuzzy reasoning. The purpose of constructing this system ″FAIS″is also to make a system that can solve many kinds of problems in various fields. As an example, the FAIS is applied to a medical diagnosis and treatment advice system, and then the usefulness of the FAIS is discussed from simulation results.

View full abstract

The goal of research is to realize natural human-machine communication system by givig a facial expression to computer. To realize this environment, it′s necessary for the machine to recognize human′s emotion condition appearing in the human′ face, and synthesize the reasonable facial image of machine. For this purpose, the machine should have emotion model based on parameterized faces which can express an emotion conditon quantitatively. Especially, both mapping and inverse mapping from face to the model have to be achieve. Facial expression of parameterized with Facial Action Coding System (FACS) which is translated to the grids′ motion of face model. An emotion condition is described compactly by the pint in a 3D space generated by 5-layered neural network and its evaluation result shows the high performance of this model.

View full abstract

The sociometry which has been proposed by Moreno [1] is one of the measrement and evaluation methods of the group structure by using crisp graph. But it would be more effective to represent human relations by applying the fuzzy graph [3]. According to the data obtained by a simple questinnaire, we could analyze the connective structure among the members and obtain the approximate sociogram. In the paper, the authors would not only discuss the analysis method of their fuzzy sosiogram, but also presents its practical effectivness with a case study in elementary school.

View full abstract

Diagnosis with medical imaging such as CT scan contains fuzziness because of subjective assessment of the findings by examiner. It is possible that such subjective assessment leads 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 inference, 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 inference was adopted for the diagnostic logic. The input and output membership functions were prepared for each it em and the final assessment was made by the center of gravity of the max-sets of the output membership function. The diagnostic accuracy of this system for 52 cases was above 90%, and there was not a false-negative case. This diagnostic supporting system appears to be beneficial for diagnosis of cervical lymph node metastasis by CT scan.

View full abstract

For a wheelchair user, the first concern is his safety in moving. In the case of attendance of acarer, this factor is fulfilled. But in that case, another problem occurs ; the problem of privacy. A human sometimes wants to be alone, being free from other s interference. ″Carer-free″activity also leads to reduction of human power. And this factor is crucial in running wheelcair guidance system in a public space or in a factory. In the present paper, human monitoring via video cameras is employed and offers unmanned and safe operation of automated wheelchairs. At the same time, human ideas, i. e., human characteristics and expertise, are fully introduced to opeating a wheelchair by utilizing human operators way of guiding a wheelchair. The present paper described this visually supervised automated wheelchair system by utilizing human ideas as operation principles extracted from the viewpoint of soft computing.

View full abstract

Hough transform is an effective method for line detection from noisy digital images. However, the coventioal Hough transform is not necessarily good enough for images under heavy noise. This paper describes a combinatorial fuzzy Hough transform, which enables the detection of a blurred line and also its segment (the start and the end of a line) from very noisy image. It is then extended in order that it may further detect a noisy circle. The effectiveness of the present method has been confirmed by applying it to the real microphotograph for cytodiagnosis.

View full abstract

The incidence of patients infected with the Human Immunodeficiency Virus (HIV) has increased almost 1.4 times in Japan. HIV infection may lead to full-blown acquired immunodeficiency syndrome (AIDS), and a ″national tragedy″, because it is as yet incurable. Currently there is no effective management of AIDS patients either environmentally or politically. We attempted to predict the future of Japanese AIDS disaster using microscopic simulation. The age and sex of the population, the risk factor based on sexual atttitude (highly dangerous, moderately dangerous, no danger), and the incidence of infection transmission from mothers to infants or children were used as factors for the calculation. The numerical calculations using these factors gave a long-term prediction (100 years) of the impact of the AIDS disaster on Japanese society. The numerical results were compared with those obtained by fuzzy simulation and an almost exact agreement was obtained.

View full abstract

Human growth of the stature and the body weight is analysedusing logistic curve. The human growth is the increase of the cell. In the body the cell division and the death of the cells occuralternatively. The growth curves of the stature and the body weightin Japanese males show S-shapes and two peaks. These peaks are locatedabout at 4 and 17 years of age. These curves run almost horizontallyalong the asymptote after the full growth. The logistic curves N(t)=CKexp(rt)/[Cexp(rt)+1] in the stature and the weight show almost thesame patterns like the growth curves obtained by measurements.

View full abstract