In this study, we focus on a diagnostic system for the field of orthopedic surgery that classifies patients' pain with high accuracy. The proposed system uses machine learning on questionnaire response data to determine pain and incorporates questionnaire items in addition to the traditional ones. Traditionally, the screening of neuropathic pain using existing methods has yielded diagnostic accuracy rates of 10 - 20%. However, our proposed approach achieves an 80% accuracy rate based on empirical data.

The self-healing network is particularly interesting with regard to ATM networks, because the restoration time can be shortened by using the advantages of the ATM network. This paper studies a self-healing ATM network based on virtual path (VP) protection switching. First, a novel self-healing algorithm/spl minus/the double-search self-healing algorithm/spl minus/is proposed. It is shown that this algorithm can restore failed bidirectional VPs faster and find alternate VPs more effectively than existing self-healing algorithms. Second, it is shown that the restoration information for self-healing control (SHC) messages must be transferred by specific cells carrying the control and OAM information (Ic&o). Message parameters and a cell format are proposed. Third, evaluation of the restoration characteristics using the proposed self-healing algorithm by computer simulation indicates that good performance against a transmission link failure is obtained even in a large-scale network model with 110 nodes. The results also indicate that the VP group (VPG) method can improve the restoration time without reducing the restoration ratio.>

A restorable network comprised of a mesh of ATM virtual paths (AVPs) is expected to be superior to distributed SONET digital cross-connect self-healing networks in terms of restoration time. The paper presents a reliable protocol and evaluation of its performance for control information transfer in distributed AVP self-healing networks. Two error-recovery technologies for the data-link level (selective retransmission and go-back-N) in AVP networks are reviewed, and a model for evaluating the delay of the go-back-N method that has been used in existing terrestrial systems is derived analytically. The proposed analytic model is combined with a self-healing simulator for AVP mesh networks to evaluate the restoration time characteristics. Case study results for the worst-case scenario show that the total restoration time, including link-by-link error control for self-healing control messages, is expected to be less than 2 s, which is the time objective for complete restoration for a metropolitan LATA network. The results also show that the estimated average delay penalty for adding the link-by-link error recovery process is 7.5 ms for complete service restoration.>

Automatically tracing a network connection and gathering network configuration data in an SDH network are necessary for future advanced network operation. This paper proposes two methods for these. They are achieved by using the OH bytes that are provided in an SDH frame for OAM applications. It first explains the mechanism of the methods and how to solve practical implementation problems. Then, quantitative analyses show that these methods take less than ten seconds to gather the network configuration data of a path going through one thousand modules in a network. It is also shown that the equipment can be realized with practical amount of hardware based on the functional block diagrams. Finally, an appropriate application area for each method is discussed from the viewpoint of required processing in an operations system and fault localization capability.>

The power spectrum of the DMI coded signal containing random errors is investigated. Results of experiments on DMI code transmission are given, confirming that low-frequency components increase according to line errors. These experimental results agree with the theoretical values for a wide error rate range. The investigation method presented in this paper can be applied to the error estimation of other line codes, such as CMI code.

This paper describes a new line code which is suitable for a high-speed optical digital transmission system of more than 400 Mbits/s. This line code, namedmB1Ccode, is created by means of a complemental pulse which is inserted everymbits. This complemental pulse prevents long series of "l"'s or "0"'s. Power spectrum, jitter, and optical transmission characteristics as well as error monitoring performance are described on theoretical analyses and experimental results.