Artificial neural network: a new diagnostic posturographic tool for disorders of stance

Automated classification of neurological disorders of gait using spatio-temporal gait parameters

OBJECTIVE

Automated pattern recognition systems have been used for accurate identification of neurological conditions as well as the evaluation of the treatment outcomes. This study aims to determine the accuracy of diagnoses of (oto-)neurological gait disorders using different types of automated pattern recognition techniques.

METHODS

Clinically confirmed cases of phobic postural vertigo (N = 30), cerebellar ataxia (N = 30), progressive supranuclear palsy (N = 30), bilateral vestibulopathy (N = 30), as well as healthy subjects (N = 30) were recruited for the study. 8 measurements with 136 variables using a GAITRite(®) sensor carpet were obtained from each subject. Subjects were randomly divided into two groups (training cases and validation cases). Sensitivity and specificity of k-nearest neighbor (KNN), naive-bayes classifier (NB), artificial neural network (ANN), and support vector machine (SVM) in classifying the validation cases were calculated.

RESULTS

ANN and SVM had the highest overall sensitivity with 90.6% and 92.0% respectively, followed by NB (76.0%) and KNN (73.3%). SVM and ANN showed high false negative rates for bilateral vestibulopathy cases (20.0% and 26.0%); while KNN and NB had high false negative rates for progressive supranuclear palsy cases (76.7% and 40.0%).

CONCLUSIONS

Automated pattern recognition systems are able to identify pathological gait patterns and establish clinical diagnosis with good accuracy. SVM and ANN in particular differentiate gait patterns of several distinct oto-neurological disorders of gait with high sensitivity and specificity compared to KNN and NB. Both SVM and ANN appear to be a reliable diagnostic and management tool for disorders of gait.

Vestibular migraine in children and adolescents: clinical findings and laboratory tests

INTRODUCTION

Vestibular migraine (VM) is the most common cause of episodic vertigo in children. We summarize the clinical findings and laboratory test results in a cohort of children and adolescents with VM. We discuss the limitations of current classification criteria for dizzy children.

METHODS

A retrospective chart analysis was performed on 118 children with migraine related vertigo at a tertiary care center. Patients were grouped in the following categories: (1) definite vestibular migraine (dVM); (2) probable vestibular migraine (pVM); (3) suspected vestibular migraine (sVM); (4) benign paroxysmal vertigo (BPV); and (5) migraine with/without aura (oM) plus vertigo/dizziness according to the International Classification of Headache Disorders, 3rd edition (beta version).

RESULTS

The mean age of all patients was 12 ± 3 years (range 3-18 years, 70 females). 36 patients (30%) fulfilled criteria for dVM, 33 (28%) for pVM, 34 (29%) for sVM, 7 (6%) for BPV, and 8 (7%) for oM. Somatoform vertigo (SV) co-occurred in 27% of patients. Episodic syndromes were reported in 8%; the family history of migraine was positive in 65%. Mild central ocular motor signs were found in 24% (most frequently horizontal saccadic pursuit). Laboratory tests showed that about 20% had pathological function of the horizontal vestibulo-ocular reflex, and almost 50% had abnormal postural sway patterns.

CONCLUSION

Patients with definite, probable, and suspected VM do not differ in the frequency of ocular motor, vestibular, or postural abnormalities. VM is the best explanation for their symptoms. It is essential to establish diagnostic criteria in clinical studies. In clinical practice, however, the most reasonable diagnosis should be made in order to begin treatment. Such a procedure also minimizes the fear of the parents and children, reduces the need to interrupt leisure time and school activities, and prevents the development of SV.

A randomised double-blind, cross-over trial of 4-aminopyridine for downbeat nystagmus--effects on slowphase eye velocity, postural stability, locomotion and symptoms

OBJECTIVE

The effects of 4-aminopyridine (4-AP) on downbeat nystagmus (DBN) were analysed in terms of slow-phase velocity (SPV), stance, locomotion, visual acuity (VA), patient satisfaction and side effects using standardised questionnaires.

METHODS

Twenty-seven patients with DBN received 5 mg 4-AP four times a day or placebo for 3 days and 10 mg 4-AP four times a day or placebo for 4 days. Recordings were done before the first, 60 min after the first and 60 min after the last drug administration.

RESULTS

SPV decreased from 2.42 deg/s at baseline to 1.38 deg/s with 5 mg 4-AP and to 2.03 deg/s with 10 mg 4-AP (p<0.05; post hoc: 5 mg 4-AP: p=0.04). The rate of responders was 57%. Increasing age correlated with a 4-AP-related decrease in SPV (p<0.05). Patients improved in the 'get-up-and-go test' with 4-AP (p<0.001; post hoc: 5 mg: p=0.025; 10 mg: p<0.001). Tandem-walk time (both p<0.01) and tandem-walk error (4-AP: p=0.054; placebo: p=0.059) improved under 4-AP and placebo. Posturography showed that some patients improved with the 5 mg 4-AP dose, particularly older patients. Near VA increased from 0.59 at baseline to 0.66 with 5 mg 4-AP (p<0.05). Patients with idiopathic DBN had the greatest benefit from 4-AP. There were no differences between 4-AP and placebo regarding patient satisfaction and side effects.

CONCLUSIONS

4-AP reduced SPV of DBN, improved near VA and some locomotor parameters. 4-AP is a useful medication for DBN syndrome, older patients in particular benefit from the effects of 5 mg 4-AP on nystagmus and postural stability.

Nonlinear variability of body sway in patients with phobic postural vertigo

Background: Subjective postural imbalance is a key symptom in the somatoform phobic postural vertigo (PPV). It has been assumed that more attentional control of body posture and / or co-contraction of leg muscles during standing is used to minimize the physiological body sway in PPV. Here we analyze nonlinear variability of body sway in patients with PPV in order to disclose changes in postural control strategy associated with PPV.

Methods: Twenty patients with PPV and 20 age-matched healthy subjects (HS) were recorded on a stabilometer platform with eyes open (EO), eyes closed (EC), and while standing on a foam rubber with eyes closed (ECF). Spatio-temporal changes of the center of pressure (CoP) displacement were analyzed to assess the structure of postural variability by computing the scaling exponent α and the sample entropy (SEn) of the time series.

Results: With EO on firm ground α and SEn of CoP displacement were significantly lower in patients (p < 0.001). For more difficult conditions (EC, ECF) postural variability in PPV assimilated to that of HS.

Conclusion: Postural control in PPV patients differs from HS under normal stance condition. It is characterized by a reduced scaling behavior and higher regularity. These changes in the structure of postural variability might suggest an inappropriate attentional involvement with stabilizing strategies, which are used by HS only for more demanding balance tasks.

Real versus imagined locomotion: a [18F]-FDG PET-fMRI comparison

The cortical, cerebellar and brainstem BOLD-signal changes have been identified with fMRI in humans during mental imagery of walking. In this study the whole brain activation and deactivation pattern during real locomotion was investigated by [(18)F]-FDG-PET and compared to BOLD-signal changes during imagined locomotion in the same subjects using fMRI. Sixteen healthy subjects were scanned at locomotion and rest with [(18)F]-FDG-PET. In the locomotion paradigm subjects walked at constant velocity for 10 min. Then [(18)F]-FDG was injected intravenously while subjects continued walking for another 10 min. For comparison fMRI was performed in the same subjects during imagined walking. During real and imagined locomotion a basic locomotion network including activations in the frontal cortex, cerebellum, pontomesencephalic tegmentum, parahippocampal, fusiform and occipital gyri, and deactivations in the multisensory vestibular cortices (esp. superior temporal gyrus, inferior parietal lobule) was shown. As a difference, the primary motor and somatosensory cortices were activated during real locomotion as distinct to the supplementary motor cortex and basal ganglia during imagined locomotion. Activations of the brainstem locomotor centers were more prominent in imagined locomotion. In conclusion, basic activation and deactivation patterns of real locomotion correspond to that of imagined locomotion. The differences may be due to distinct patterns of locomotion tested. Contrary to constant velocity real locomotion (10 min) in [(18)F]-FDG-PET, mental imagery of locomotion over repeated 20-s periods includes gait initiation and velocity changes. Real steady-state locomotion seems to use a direct pathway via the primary motor cortex, whereas imagined modulatory locomotion an indirect pathway via a supplementary motor cortex and basal ganglia loop.

Vestibular-neck interaction in cerebellar patients

Vestibulospinal reflexes are important for upright stance and locomotor control. Information from both the vestibular and the proprioceptive system must be combined centrally to guarantee appropriate compensation for a physical disturbance. Recent single-unit recordings from the monkey demonstrated vestibulo-proprioceptive interaction in the fastigial nucleus (deep cerebellar nucleus). The present study investigated whether integration of vestibular and proprioceptive signals is compromised in humans with cerebellar degeneration. Control subjects and patients were exposed to binaural, sinusoidal galvanic vestibular stimulation at 0.16 Hz, while their static head-on-trunk position was systematically altered in the head-horizontal plane from 60 degrees left to 60 degrees right. Controls responded to different head-on-trunk positions with fully compensatory changes in the direction of galvanically induced body sway, keeping it aligned with the head-frontal plane. In patients, this compensatory change was lacking. Findings support the assumption that the cerebellum plays a central role in the integration of vestibular and proprioceptive signals in humans. This form of impaired sensory interaction is probably a clinically important component of cerebellar stance and gait ataxia.