No abstract

In the cost sensitive healthcare industry, an unplanned downtime of diagnostic and therapy imaging devices can be a burden on the financials of both the hospitals as well as the original equipment manufacturers (OEMs). In the current era of connectivity, it is easier to get these devices connected to a standard monitoring station. Once the system is connected, OEMs can monitor the health of these devices remotely and take corrective actions by providing preventive maintenance thereby avoiding major unplanned downtime. In this article, we present an overall methodology of predicting failure of these devices well before customer experiences it. We use data-driven approach based on machine learning to predict failures in turn resulting in reduced machine downtime, improved customer satisfaction and cost savings for the OEMs. One of the use-case of predicting component failure of PHILIPS iXR system is explained in this article.

In this paper we introduce a coarse-to-fine arrhythmia classification technique that can be used for efficient processing of large Electrocardiogram (ECG) records. This technique reduces time-complexity of arrhythmia classification by reducing size of the beats as well as by quantizing the number of beats using Multi-Section Vector Quantization (MSVQ) without compromising on the accuracy of the classification. The proposed solution is tested on MIT-BIH arrhythmia database. This work achieves a highest computational speed-up factor of 2.2:1 in comparison with standard arrhythmia classification technique with marginal loss (<;1%) in classification accuracy. The clinical application of this technique enhances physician's throughput by factor of 2x while processing large ECG records from Holter system.

In this paper, we introduce a Concept Graph Engine (CG-Engine) that generates patient specific personalized disease ranking based on the laboratory test data. CG-Engine uses the Unified Medical Language System database as medical knowledge base. The CG-Engine consists of two concepts namely, a concept graph and its attributes. The concept graph is a two level tree that starts at a laboratory test root node and ends at a disease node. The attributes of concept graph are: Relation types, Semantic types, Number of Sources and Symmetric Information between nodes. These attributes are used to compute the weight between laboratory tests and diseases. The personalized disease ranking is created by aggregating the weights of all the paths connecting between a particular disease and contributing abnormal laboratory tests. The clinical application of CG-Engine improves physician's throughput as it provides the snapshot view of abnormal laboratory tests as well as a personalized disease ranking.

In clinical environment, Interventional X-Ray (IXR) system is used on various anatomies and for various types of the procedures. It is important to classify correctly each exam of IXR system into respective procedures and/or assign to correct anatomy. This classification enhances productivity of the system in terms of better scheduling of the Cath lab, also provides means to perform device usage/revenue forecast of the system by hospital management and focus on targeted treatment planning for a disease/anatomy. Although it may appear classification of each exam into respective procedure/anatomy a simple task. However, in real-life hospital settings, it is well-known that same system settings are used to perform different types of procedures. Though, such usage leads to under-utilization of the system. In this work, a method is developed to classify exams into respective anatomical type by applying machine-learning techniques (SVM, KNN and decision trees) on log information of the systems. The classification result is promising with accuracy of greater than 90%.