[in Japanese]

View full abstract

Number of patients of the soft tissue tumor is relatively very small to the number of well-known cancers. The pathological diagnosis is essential for the treatment of malignant tumors, but there are few pathologists familiar with malignant soft tissue tumors or medical institutions specializing in their diagnosis. Pathological images are used for postoperative treatment planning and to determine whether chemical treatment should be administered. However, this is too difficult to decide for physician, so it is burden on. Therefore, an objective diagnostic system based on pathological images is required. This paper proposes survival time estimation methods using classification and label distribution learning for soft tissue tumor patients. We believe that this will realize treatment tailored to the patient's medical condition. In this paper, we propose a method that uses the label distribution learning in survival estimation can improve accuracy. The minimum MAE for soft label was 7.69 months and compared to hard labels, the maximum improvement was about one month.

View full abstract

In recent years, medical tubing has been used more frequently, and there is concerned about risk of self-extubation, in which patients pull out their medical tubing unplanned. It may lead to serious incidents such as massive bleeding or cardiac arrest. Frequent nurses visit and restrain patients being carried out, but there is a desire for new system due to the heavy burden. In this study, sheet-type self-extubation detection device is proposed. This sensor has a structure with capacitive touch sensor matrix mounted on top of wound dressing, and it can detect the tube's position and tube detachment. An initial stages of sensor design, a prototype was created. This prototype was attached to a dummy tube, and evaluation was conducted by applying tensile force to the tube. It was observed that tube movement due to and detachment could be detected.

View full abstract

Osteochondritis dissecans (OCD) is one of the most common sports disorders in elementary and junior high school baseball players. OCD is often detected after it has progressed to a painful condition. Therefore, early detection in the early stages is considered to be important. In this study, we propose an OCD detection model using deep learning from ultrasound images to assist diagnosis. As a result, the OCD classification model using VGG16 achieved 0.845 in accuracy and 0.983 in AUC.

View full abstract

The objective of this study is to elucidate the issue of dropping during the flushing of closed catheters and to improve awareness of dropping prevention. Simulating a subject equipped with a ventilator, a comparative study on the causes of dropping during flushing was conducted using three different closed catheters widely used in the medical field. The results indicated significant differences in adult catheters based on ventilation volume and mode, with some cases showing significant differences according to PEEP (positive end-expiratory pressure) and angle. Additionally, pediatric and adult + pediatric catheters exhibited significant differences in dropping based on ventilation volume and type, revealing structural issues with the catheters. Specifically, it was suggested that flow from the ventilator directly affects the catheter tip and that saline buildup in the anti-drop valve can cause dropping. Furthermore, dropping issues were observed depending on the technique of the individual performing the suction. Based on the novel insights gained from this study, there is consideration to form a research team for the development of an improved device in the future.

View full abstract

 Since Peripheral Capillary Oxygen Saturation(SpO₂) allows non-invasive measurement of oxygen saturation in arterial blood, it is used not only in clinical settings but also in home healthcare and, more recently, as an indicator of respiratory failure in Covid-19. SpO₂ can be measured at the fingertip, earlobe, nose, and forehead, but peripheral circulatory failure in the fingertip, commonly used for measurement, may prevent accurate SpO₂ readings. Undesirable conditions for measurement include manicured nails, ambient light interference, and body movement.
 However, the recently introduced Philips Nasal Alar SpO₂ sensor can measure SpO₂ even in cases of peripheral circulatory failure and is also utilized during surgeries. In this study, we report on basic experiments conducted to investigate the extent of discrepancies in measuring normal values in simulated conditions of peripheral circulatory failure, colorimetric testing, and manicured nails using the Philips Nasal Alar SpO₂ sensor and a conventional SpO₂ monitor.

View full abstract

Treatment with a combination of home mechanical ventilation (HMV) and home oxygen therapy (HOT) is sometimes considered for severe patients with chronic obstructive pulmonary disease (COPD). The Japanese Respiratory Therapy Society recommends supplementing the fraction of inspired oxygen (FiO₂) by increasing ventilation frequency instead of raising the FiO₂. However, conventional HOT has been reported to fall short of achieving the target FiO₂ in a hospital setting. Furthermore, previous studies have reported a decrease in FiO₂ due to leaks during mask ventilation with non-invasive positive pressure ventilation (NPPV). In this study, we attempted to explore methods for increasing FiO₂ values in home medical care using a combination of HMV and HOT beyond what previous research has achieved. We conducted a study introducing liquid oxygen in addition to the traditional oxygen concentrator, aiming to generate new insights into the combined use of HMV and HOT. This report presents our findings.

View full abstract

Deep learning, which has been the focus of much attention in recent years, provides highly accurate results, but has the problem of unclear input-output relationships. On the other hand, fuzzy inference models that use If-Then rules can represent human knowledge and make the inference process easy to understand. One such model is the deep SIRMs coupled fuzzy inference model. This model is characterized by its single-input rules, which not only make the rules easy to understand, but also realize exclusive OR. However, rules that use the output of the previous layer as one of the input variables of the next layer are difficult to understand.
In this study, we propose a new interpretation of the output of each layer of the deep SIRMs coupled fuzzy inference model to discover new conditional attributes that are useful for inference. We also evaluate the feasibility of the method using medical diagnosis data.

View full abstract

In the wake of the COVID-19 pandemic, the inclination towards online shopping has grown, further highlighting the importance of recommendation systems in e-commerce. Traditional recommendation algorithms require extensive datasets to perform effectively, which poses a challenge for small to medium-sized enterprises with limited data. This study proposes a recommendation system that significantly reflects user preferences, even with constrained data volumes, to enhance user satisfaction. We conducted computational experiments simulating actual products and typical users to validate the effectiveness of our approach. Our results, presented through a series of rankings and user preference data, demonstrate the system's ability to adapt to user feedback and improve recommendation quality. Future work will focus on extracting specific reasons from product evaluations to refine the rating of similar items automatically and re-construct user ratings based on the similarity of preferences across a database.

View full abstract

After the subjects watched the drive recorder video, they were asked to draw a map from memory and answer the direction of the end point as seen from the start point, the driving distance, and the linear distance. The subjects were asked to watch two courses, "Course A with sound" and "Course B without sound," or "Course A without sound" and "Course B with sound. The data of "Course A with sound" and "Course A without sound" were compared, and no significant differences were found. The same was seen for "Course B with sound" and "Course B without sound.

View full abstract

Chronic lung disease (CLD) is the most common and serious lung lesion in extremely preterm infants. Early intervention in predicted CLD patients is effective in improving prognosis. However, methods for early detection of onset have yet to be established. Previous studies have utilized clinical data to predict the CLD, but none have employed chest X-ray images. This paper proposes a novel method for predicting CLD onset on chest X-rays using convolutional neural networks (CNNs). We also compare the performance of CLD prediction among 10 state-of-the-arts CNNs, and conducted a longitudinal study to assess change in prediction accuracy at each of 3, 7, 14, and 28 days of age. Validation was conducted with 115 cases (51 CLD and 64 normal). The proposed method achieved the best performance with an accuracy of 0.715, 0.652, 0.739, 0.765, and an AUC of 0.796, 0.673, 0.778, 0.828 at 3, 7, 14, 28 days, respectively. The CNN selected was NASNet_Large, DenseNet121, DenseNet169, DenseNet201, at 3, 7, 14, 28 days, respectively. Our findings conclude that the performance of image-based prediction is comparable to the previous study using clinical data, and further improvement can be achieved by incorporating both clinical data and images.

View full abstract

Organic deterioration in the rotator cuff muscles often causes shoulder pain, limitation of motion and shoulder disabilities in patients. Fat content, as calculated by the Dixon method, is a quantitative assessment of rotator cuff deterioration and is relevant to the outcome of reconstructive surgery. However, Dixon method requires four different muscle regions of interest on shoulder MR images manually, which is a barrier to its clinical dissemination. In this study, rotator cuff muscles segmentation method is proposed to aid clinical quantitative assessment: using 28 MR images, Segmentation of the four rotator cuff muscles and supraspinatus fossa and from these the fat content and muscle atrophy rates of the Dixon method are estimated. As a result Dice coefficients of 0.93, 0.95, 0.88, 0.82, and 0.96 were shown for the subscapularis, supraspinatus, infraspinatus, teres minor, and supraspinatus fossa, and the atrophy rate was also highly accurate with RMSE = 4.29. In addition, it was confirmed that the accuracy of segmentation affects the prediction accuracy of fat content.

View full abstract

Studies have been conducted to predict outcomes of extracorporeal shock wave lithotripsy (ESWL) using clinical information from patients. However, non-stone areas in CT images were not included, as the main predictors were stones and patient body mass index. In this paper, two types of convolutional neural networks are applied to non-stone areas of CT images to predict ESWL outcomes. Resulting AUC of 0.94 and accuracy of 0.91 were achieved. Multivariate analysis was also performed and confirmed that the proposed predictors are independent of the conventional stone factors.

View full abstract

The authors have conducted experiments and examined the use of two features for continuous identity authentication for the timing of keystrokes and the sound of the Enter and Space keystrokes. However, these two features may not be sufficient for telework and online classes. To compensate for this problem, we conducted experiments and examined a new method of identity authentication using keypoints of the fingers obtained from OpenPose. The final goal of this study is to perform multimodal authentication using keystroke timing, the Enter and Space keystroke sounds, and hand keypoints. In this paper, we report on the analysis method and results of identity authentication using hand key points obtained from OpenPose.

View full abstract

The purpose of this study is to elucidate the neutrophil propulsion mechanism by cytokine. It is generally known that the interactions between cytokines and receptors on the neutrophil surface are related to signaling by calcium ions (Ca²⁺). In this study, to clarify the effect of calcium ion concentration on the neutrophil propulsion mechanism, observation experiments with a microscopy are conducted. The time history of calcium ion concentration is obtained by Fura-2. As a result, it was found that stimulus by cytokines reduces calcium ion concentrations on the neutrophil surface. It was also found that calcium ions on the surface are consumed in dead neutrophils, but it doesn’t move. It is concluded that calcium ions have an important role in signal transduction and are consumed by cytokine stimulus.

View full abstract

This paper proposed a new method of calculating the microscopic quantum system based on the classical mechanics. Quantum mechanics is essential in many microsystems, but it may be difficult to apply it to many complex phenomena with many degrees of freedom. Therefore, instead of the quantum mechanics, we apply the classical mechanics with stochastic process, namely the intranuclear cascade (INC) model. In this paper, we show that INC reproduces well the experimental data for nonelastic reaction cross sections of ¹²C, ²⁷Al, ⁵⁶Fe and ²⁰⁸Pb. The INC is a promising method to calculate complicated quantum systems involving many processes of nuclear reactions, and INC is superior in providing an intuitive picture for the phenomena.

View full abstract

The authors are conducting research on a programming learning support system to make it easier for beginners to learn C programming. We have designed a standard coding form that assists learners using a form mechanism, studied each function of the system, created a prototype, and conducted an interim evaluation of the system. In programming education, the hints and explanations given to learners do not always match what is expected of them. Therefore, it is desirable to estimate the cause of each learner's "stumbling block" and provide content based on that. In this paper, we examine the estimation of the cause of stumbling and the presentation of hints accordingly, based on the information that can be collected during coding, and report on the results of our experiments.

View full abstract

This research proposes a simulation of human behavior selection using Three-seconds rule intelligence. Three-Second Rule Intelligence is an artificial intelligence that makes behavior decisions based on predictions of the situation up to Three seconds in the future and is composed of an action search unit and an action selection unit from the experience database. In this study, we created the web-based trial system for determining the activation of patient support actions in radiation therapy. As a result of the trial, we will report on the possibility of action selection based on Three-second rule intelligence.

View full abstract

With the progress of information technologies such as AI (Artificial Intelligence) and IoT (Internet of Things), research on machine learning using big data has been conducted. In some cases, complex and large-scale computations are required to realize machine learning, and in such cases, cloud and edge systems are widely used. On the other hand, in these systems, there is a risk of data loss or leakage by depositing data in an external server. For this reason, research on learning methods that maintain data security and are easy to use is widely conducted. As for secure computation methods using distributed processing, learning methods using subsets of data or decomposed data such as FL are well known. These methods are characterized by the fact that they achieve high confidentiality by repeating distributed processing, in which data and parameters in machine learning are distributed or decomposed and computed, and integrated at a central server. On the other hand, the role of the central server is important for the realization of machine learning by data distribution and integration. Therefore, it is desirable to realize machine learning by a decentralized autonomous method instead of such a centralized method.
In this paper, we propose a BP learning method for decentralized distributed systems using decomposed data and demonstrate its effectiveness.

View full abstract

Rapid changes in society require the development of human resources who can create new value. Therefore, the enrichment of entrepreneurship education from elementary school to university is expected and promoted. We will discuss the awareness gained after using entrepreneurship education materials for students who want to get a job in companies or organizations, rather than as entrepreneurs or startups.

View full abstract

This paper evaluates whether Wi-Fi sensing using CSI can detect NC milling machine tools' machining status and other operations without contact sensors. Since the transmission path of Wi-Fi communication changes slightly due to equipment vibrations, the CSI data changes depending on the NC machine tool's operation. We evaluate whether the NC machine tool's operation status can be estimated from these changes. Experiments conducted by changing the operating conditions and cutting materials showed that NC milling machines' operating and machining conditions could be estimated using CSI.

View full abstract

The frequency of occurrence of the symptom in twelve meridian patterns was investigated to construct a clinically relevant Meridian patterns. In this paper, the more frequently appearing symptoms in the survey were identified as the adopted symptoms and their sources were examined. The result shows that the sources of the adopted symptoms were the ancient medical books and "Acupuncture and Moxibustion " at just under 40% each, and others at over 20%. The adopted symptoms in this survey were compared to the symptoms of ICD-11 Meridian patterns to determine the concordance rate. The concordance rate for these compared symptoms was 54.5%, which was low.

View full abstract

Acupuncture/Moxibustion treatment was adopted to multiple system atrophy (MSA), which includes cerebellar ataxia, parkinsonian syndrome and autonomic disorder. MSA is one of the designated intractable diseases. The pathogenesis and the treatment are not clear today. If nothing is done, the condition will get worse and worse over time. Therefore, it is necessary to find a way to suppress it. Acupuncture/Moxibustion treatment may be one of the candidates.

View full abstract

In order to examine the growth status of cherry blossom in the western part of Tottori Prefecture, we investigated the causes (their attributes, components) that hinder the growth of cherry trees for each tree species. Furthermore, their numerical data were analyzed by using the significance analysis of the SOM method. And the growth status of each cherry blossom spot was examined.

View full abstract

The principle of the significance method is explained using iris data, and this method is explained. We will present the results of applying it to the benchmark data and others.

View full abstract

Traveling Salesman Problem (TSP) is a problem of finding the shortest route back to the starting point after visiting each city only once. It is called an NP-complete problem, and many computational methods have been proposed. Referring previous Self-Organizing Maps (SOM) research, we have studied how to shorten the computation time for applying SOM to the TSP while keeping the original form. In this study, we applied and examined the problem of traveling cities on the earth.

View full abstract

The Traveling Salesman Problem (TSP) is a problem of finding the shortest route back to the starting point after visiting each city only once. Referring previous Self-Organizing Maps (SOM) research, we have studied how to shorten the computation time for applying SOM to the TSP while keeping the original form. In this study, we applied SOM-TSP on the optimization of component placement in a chip mounter.

View full abstract

Recently, AI technology called deep learning has been attracting attention. Deep learning is a kind of neural network. It is a method of machine learning using a hierarchical neural network with a multi-layer structure. Deep learning has shown great results in image recognition. In this study, we consider the classification of unearthed coins using deep learning. There are many types of unearthed coins to be excavated from a ruin. One of the most famous coins is Kanei-tsuho coins. There are many subdivisions of Kanei-tsuho coins. However, the classification of unearthed coins is even difficult task for numismatics experts.
In this research, we use transfer learning to classify unearthed coins. Transfer learning is a type of machine learning. This method acquires knowledge through separate training data. Use the acquired knowledge to classify other data. We will verify whether unearthed coins can be classified using a model learned from large-scale image data.

View full abstract

Accidents involving large vehicles on highways at night are caused by reduced visibility and other factors. To prevent this, it is effective to increase the visibility of the vehicle, especially by contour affixing reflective materials to the rear of the vehicle. In this study, retroreflective material was affixed to the rear of a large vehicle and its effectiveness was tested on a highway at night. In the experiment, the distance between vehicles from behind and the behavior of lane changes were analyzed by comparing the vehicles without reflective material, with reflective material affixed in a horizontal line, and with reflective material affixed in a contour. As a result, it was observed that behind the contoured vehicles, the distance between the vehicles was extended and the position where the lane change was initiated was also farther away.

View full abstract

To investigate the effects of direction of sight on cognitive maps and scene memory, we conducted a psychological experiment using drive recorder images attached to the front (forward sightline) or left (leftward sightline) side of a car and analyzed the results. Here, we discuss the field of view of the drive recorder. Considering the central portion of the forward sightline as the central vision, part of the left side of the peripheral vision of the forward sightline coincides with part of the right side of the left side of the forward sightline. Landmarks in the distance of the car's travel direction appear small in the central vision, but appear larger in the peripheral vision range as the car approaches. We suppose that the central and peripheral visions each play a useful role in the cognitive map.

View full abstract

Various analyses using contingency tables are very useful for finding interactions between two variables. The authors have extended contingency tables by applying fuzzy set theory and defined fuzzy contingency tables (type-1/type-2 fuzzy contingency tables). This enables type-1/type-2 fuzzy partitioning of similarity indices in typical set theory. In addition, it is now possible to calculate vectors from the similarity using MDS. In this paper, we propose a similarity calculation method using contingency tables and a similarity evaluation method using MDS, and evaluate the similarity of simulated data as a simulation.

View full abstract

We have studied the structures and applications of fuzzy graphs. Related this study, we introduced one methodology, Fuzzy Core Index Analysis. By virtue of this method, we can transform the nodes of a fuzzy graph with a semi-ordered structure into a linear-ordered structure. It should be noted that this method decides a unique ordering (with respect to the importance) of the nodes of a fuzzy graph. The procedure for this method, however, was somewhat complicated in instruction structure analysis. In this paper, we then define fuzzy core indices directly from a contingency table and investigate the validity of the redefinition.

View full abstract

AHP (Analytic Hierarchy Process) has been widely used in decision making. However, it is very difficult to keep their independency among criteria. Therefore, there are two popular extension method, inner dependence AHP and fuzzy measure AHP. In this paper, we compare these two methods from a view point of their sensitivity analysis.

View full abstract

Membership function can be constructed by confidence intervals of a probability distribution. From the relation we consider an equivalence between fuzzy numbers and random variables and its application to analysis of fuzzy systems.

View full abstract