Video analysis of Hammersmith lateral tilting examination using Kalman filter guided multi-path tracking

Video object tracking plays an important role in many computer vision-aided applications. This paper presents a novel multi-path analysis-based video object tracking algorithm. Trajectory of the moving object is refined using a Kalman filter-based prediction method. The proposed algorithm has been used successfully to analyze one of the complex infant neurological examinations often referred to as Hammersmith lateral tilting test. This is an important test of the infant neurological assessment process, and this test is difficult to grade by visual observation. It has been shown in this paper that the proposed video object tracking algorithm can be used to analyze the videos of fast moving objects by incorporating application-specific information. For example, the proposed tracking algorithm can be used to assess lateral tilting test of the Hammersmith infant neurological examinations. The algorithm has been tested with several video recordings of this test which were captured at the neurodevelopment clinic of the SSKM Hospital, Kolkata, India during the period of the study. It is found that the proposed algorithm is capable of estimating the score for the test with high values of sensitivity and specificity.