Stroke survivor’s functional mobility loss is expected to improve if the patient receives intensive therapy, and performs activities of daily living. Providing a quantitative assessment for patients’ recovery progression instead of the traditional ordinal scale has received significant attention. In this paper, an investigation was conducted using Mahalanobis-Taguchi-Adjoint, MTA method to develop an evaluation system for stroke patients with upper-limb impairments. Replicating upper-limb impairment, an experiment was conducted by restraining able-body subjects’ shoulder, elbow, and wrist joint motion under six restraint conditions. Fifteen able-body subjects’ data were collected and implemented as unit space/signal dataset, to detect difference of another fifteen ablebody subjects on their capability of performing under shoulder, elbow, wrist, and compound restraint conditions. From the results, MTA method was capable of detecting difference between subjects reaching movement ability between shoulder, elbow, and wrist joints. Measuring estimated value’s distribution size with coefficient of variation, CV, wrist restraint CV resulted in 0.575. For elbow and wrist restraint, CV resulted in 0.385, and shoulder restraint CV resulted in 0.354. Therefore, MTA method when using the proposed features, enable clinicians to detect changes on patients’ impairment upper-limb performance compared to able-body subjects.

抄録全体を表示

The conventional way to diagnose catalyst deterioration in internal combustion engines is to intentionally control the amount of fuel injected under specific operating conditions and determine the delay from the start of catalysis to the end of catalysis by observing the oxygen concentration signal. Problems with this conventional method include the large size of the catalytic converter, degradation of drivability, degradation of the exhaust gas, obtaining diagnoses with adequate frequency, lengthened development time, and a large computational load on the engine control computer. These raise a variety of issues in such areas as marketability, government regulations, and development costs. In this study we applied the MT method, focusing on the fact that the degree of deterioration can be determined by observing the oxygen concentration signal before and after catalysis during normal operation, without controlling the amount of fuel injected as in the conventional method. As a result, we were able to develop a method that solves all the above problems at once and can be implemented on board the vehicle to obtain a real-time diagnosis.

抄録全体を表示