Auxiliary Diagnosis Of Cervical Spondylotic Myelopathy Based On Gait Analysis And Deterministic Learning

Cervical spondylotic myelopathy(CSM)is the main cause of cervical spinal cord dysfunction in adults.The early symptoms of CSM are mild gait disturbances.The progression of CSM will lead to severe spinal cord compression and paralysis.Currently,the diagnosis of CSM relies heavily on imaging tools,but static imaging methods are difficult to describe the motion function of the cervical spine.In addition,the assessment of motion function in CSM patients relies on subjective rating scales,which has limitations.In this context,a method of gait analysis based on deterministic learning theory is proposed to detect CSM by constructing a gait dataset of CSM patients and age-matched healthy controls(HCs).The main works of this paper are summarized as follows:1.A gait dataset containing 45 CSM patients and 45 HCs is constructed to compensate for the absence of a gait dataset of CSM patients and age-matched HCs.Firstly,a gait acquisition experimental environment based on optical motion capture system is constructed and calibrated.Secondly,the static acquisition experiment and the motion acquisition experiment are carried out by pasting the reflection marker on the experimental target subjects.Finally,the dataset is preprocessed by gait cycle extraction,annotation,interpolation,and filtering.In addition,this paper also completes the gait acquisition of 33 CSM patients and 31 HCs using an inertial-based motion capture system,further enriching the dataset.2.The gait data are expanded and gait features are extracted,which to a certain extent solves the current problem of insufficient data acquisition in the field of gait analysis due to economic and site constraints.Firstly,a stitching method of C3D data is proposed to stitch the C3D data from the level of data structure to increase the number of consecutive gait cycles.Secondly,Visual3D software is used to build the human skeleton model according to the anatomical bone markers.Finally,the C3D data are substituted into the human skeleton model to complete the extraction of kinematic and kinetic features of lower limbs.3.A new model with the combination of deterministic learning theory and extracted features is used to train and classify nonlinear dynamics of gait patterns between patients with CSM and HCs with high accuracy.Firstly,discriminant kinematic gait features,including angles of hip and knee joints in the sagittal and coronal planes,are extracted based on statistical analysis and clinicians’ empirical investigation.Second,deterministic learning theory is used to model and identify nonlinear gait system dynamics of HCs and patients with CSM,which are approximated and stored in constant Radial Basis Function(RBF)neural networks(NN).The disparity of gait system dynamics between the two groups of participants is used for classification and detection of the presence of CSM by constructing a bank of dynamic estimators with constant RBF NN.Finally,experiments are carried out on the self-constructed CSM gait database to evaluate the performance of the proposed method,in which gait data from 45 CSM patients and 45 age-matched HCs are involved.By using 2-fold and leave-one-out cross-validation styles,the achieved average classification accuracy is reported to be 94.44%and 95.56%,respectively.To summarize,this paper proposes a new auxiliary diagnosis method of CSM based on gait analysis and deterministic learning theory by performing a more comprehensive and in-depth gait analysis of CSM patients and HCs in the 3D plane with a self-constructed dataset.The method uses fewer input features and achieves better classification results,which has some application value.