Artificial Neural Networks (ANN) are extremely useful to relate the nonlinearly depending outputs with the inputs. Various architectures are available for the ANNs to speedup the training period and reduce the square error. In this paper, new classes of neural networks with differential feedback are presented. The different orders of differential feed back form a manifold of hyperplanes. Interesting properties of this differentially fed ANN (DANN) are derived through these hyperplanes.

Abstract-Radiotherapy plays an important role in the treatment of cancer patients. As part of clinical workflow, patient has to undergo through diagnostic imaging procedures, which are used to identify the tumor location and size. Enormous amounts of data are generated during this procedure. The volume of medical information is so large and complex that it becomes difficult to mine for relevant information. The Digital Imaging and Communications in Medicine (DICOM) standard is widely used in medicine for storing and transmitting medical information. The DICOM-RT is the extension to DICOM standard, and dedicated to radiotherapy. In this paper, we propose a technique to store clinical relevant features from DICOM files using semantic concepts. The proposed technique defines a novel method to delayer the hierarchy of DICOM-RT for storing the clinical relevant information into triples in Resource Description Framework (RDF) repository. The methodology also proposes different combinations for storing data such as DICOM-RT with tumor information, DICOM-RT with pathology details. The proposed method uses the Semantic Web Technology to store and represent the information from DICOM-RT files along with into RDF graph and a data mining approach. Natural Language processing technique is used for the retrieval of data. We have evaluated our methodology qualitatively for 20 patients including combinations such as RTSTRUCT, tumor size data along with CT data, pathology information, by producing 25 varieties of different queries. We have analyzed quantitatively with accuracy of 90% for different hypothetical conditions using our proposed methodology.Keywords-DICOM-RT, Semantic Web, RDF, SPARQL, Natural Language Processing. IntroductionRadiation therapy is one method of the cancer treatment, and plays an important role for patient during the course of the disease [1]. In this process, patients have to undergo diagnostic imaging procedures, which are performed to identify the tumor location and size. Data generated during this procedure contains large volume of information as well as complex structures, which makes it a challenging task for clinicians to query and retrieve relevant data The DICOM-RT objects provides information about patient related structures identified from diagnostic data known as radiotherapy structure set (RTSTRUCT), contains radiotherapy treatment plan information (RTPLAN) and also provides total dose distributions from the planning systemdose information (RTDOSE) [5]. The DICOM-RT objects are stored in hierarchal manner; this restricts the search path while traversing the DICOM modalities and the DICOM query model itself and current DICOM tools do not support this required traversing well [6]. The literature shows planning for dose to be received by a certain region of interest in a radiotherapy treatment is to find the region of interest (ROI) name and contour points in the RTSTRUCT object. The coordinates and slices are defined in the CT objects, the treatment information stored in RTPLAN object and dos...

Performance modeling of a network is challenging especially when it involves multimedia traffic. The present day home network makes uses of internet for multimedia content transfer on to the devices in real time. In such a communication system, reliability and in time data transfer is critical. The system has to support the streaming of multimedia and entertainment data from mobile to infrastructure and vice versa. In this paper, a novel modeling method for the network and its traffic shaping is introduced and simulation model is provided. The performance with this model is analyzed.