Please use this identifier to cite or link to this item:
Title: Prognosis of negative symptoms in Schizophrenic patients using kinect recordings
Authors: Ranganaj, Nishaanthi
Keywords: DRNTU::Engineering::Electrical and electronic engineering
Issue Date: 2018
Abstract: This report summarizes the progress of the dissertation called “Prognosis of Negative Symptoms in Schizophrenic Patients using Kinect Recordings” under the guidance of Professor. Justin Dauwels. The prime objective of this project is to identify the presence of Schizophrenia in the patients based on their body movements. The patients affected by schizophrenia have reduced body movements when compared to a healthy individual. This project includes the extraction of the distance moved by each body joint of the patients and healthy individuals. Several machine learning classifiers are used to classify the test data based on the movement of the test subjects and the accuracy of each classification is determined to determine a better classifier. The interview with all the patients and healthy individuals are recorded using the Kinect sensor. A large database of Kinect recording is collected. Then, these data are stored as csv files consisting of x,y,z co-ordinate of 20 joints. Further processing will be performed on the acquired data to detect the distance moved by each joint (20 joint co-ordinate) using the Euclidean distance. In this project the machine learning classifiers such as Support Vector Machines (SVM), K-Nearest Neighbors and Naive Bayes classifier are used for classification of patients and healthy individuals. Then, the classifier with highest accuracy that classifies the test subjects based on their movement is identified. During classifier analysis the 20 joint co-ordinates are considered as features for each patient/healthy individual.
Fulltext Permission: restricted
Fulltext Availability: With Fulltext
Appears in Collections:EEE Theses

Files in This Item:
File Description SizeFormat 
  Restricted Access
Main article1.71 MBAdobe PDFView/Open

Google ScholarTM


Items in DR-NTU are protected by copyright, with all rights reserved, unless otherwise indicated.