A microcomputer system for clinical signal analysis of vestibular laboratory

Using decision tree induction to model oculomotor data

Decision tree induction is a machine learning method used to generate classification models from data sets. Numerous decision trees were constructed to examine relationships between oculomotor test parameters and lesion sites in a data set containing cases with operated cerebello-pontine angle tumour, operated hemangioblastoma, infarction of cerebello-brainstem and Ménière's disease, and control subjects. The aim was to find useful parameter combinations with discriminatory power. Decision trees constructed using both pursuit eye movements and saccadic eye movements yielded the best classification results. This is reasonable: oculomotor test results vary according to the site of the lesion and so the performance ability of subjects has to be taken into account in the classification. The decision tree program was able to generate classification models from the oculomotor data set. Generated decision trees were intelligible and can be utilized in physicians' research work.

Lyme borreliosis--an unusual cause of vertigo

A total of 2055 consecutive vertigo patients were examined in a prospective study in an area endemic for Lyme borreliosis for clinical signs of Lyme borreliosis or serum antibodies against Borrelia burgdorferi. Of these, 41 patients (2%) had positive levels of serum antibodies against B. burgdorferi. The incidence of seropositivity against B. burgdorferi among the vertigo patients did not differ from the incidence of the normal Finnish population. In addition to seropositivity the criteria used for Lyme borreliosis included previous erythema migrans, a positive polymerase chain reaction (PCR) or positive serum immunoblot. Eight patients were diagnosed as having Lyme borreliosis. This disease is a rare but possible cause of vertigo. Seropositivity alone is an insufficient finding for the diagnosis of Lyme borreliosis and should be supported by the clinical findings, the patient's history and other laboratory findings, such as immunoblotting or PCR. Although Lyme borreliosis seems to be a rare cause of vertigo, it must be kept in mind in the differential diagnosis of vertigo.

Hemangioblastoma and oculomotor pathology

We examined voluntary eye movements of 20 ophtalmologically screened patients (mean age of 51 years) with operated hemangioblastoma (HAB) in the cerebellum. Constant and pseudo-random saccades and pseudo-random smooth pursuit eye movement (PRPEM) test (with 4 frequency combinations) were evaluated. As controls 38 healthy subjects were examined. In the logistic regression analysis latency and accuracy of constant saccades were the best predictors for operated HAB with correct overall classification rate of 79.3%. Accuracy was worse and latency longer in operated HAB group than in control group. In pseudo-random saccades correct classification between the groups was achieved in 82.8% of all cases with latency and accuracy as predictors. In PRPEM test the best frequency combinations in differentiating operated HAB patients from controls were 0.25 and 0.425 Hz for gain, and 0.3 and 0.7 Hz for phase and gain, the correct overall classification rate being 73.3% in both cases. The characteristic changes in voluntary eye movements after removal of cerebellar hemangioblastoma seem to be insufficient timing of initiating the eye movement.

Oculomotor pathology in Menière's disease

In this study data of voluntary eye movements 62 patients (mean age was 49.6 years) with Meniere's disease and 38 healthy control subjects were examined. Pseudo-random smooth pursuit (PRPEM) tests were conducted with frequency combination of 0.25 and 0.425 Hz. Saccades at constant targets and pseudo-randomly shifting targets were evaluated. In logistic regression analysis constant saccades correctly classified the cases in 57% as saccadic latency discriminating the groups best, latency being longer in Meniere group. In pseudo-random saccades, correct classification was achieved in 75.4% of all cases as latency discriminating the groups best, and latency was longer in Meniere group. In PRPEM, correct classification was achieved in 65.3% of all cases as gain by amplitude (GA), discriminating the groups best, and GA was smaller in Meniere group. Pseudo-random saccade test is a demanding task and that may explain why a peripheral vestibular lesion may interfere visual tracking and scanning performance. Latency is the most vulnerable of parameters to be lesioned. Results indicate that a peripheral vestibular lesion influences the control of voluntary eye movements and may explain the complaints of visual targeting on objects that patients with Meniere's disease have.

Intratympanic gentamicin in bilateral Menière's disease

Treatment of bilateral Menière's disease is a delicate task. Streptomycin has demonstrated its ability to successfully control the vertigo attacks, but in some cases it results in oscillopsia and transforms the periodic attacks to permanent instability. If the site of the active labyrinth can be determined, a more specific treatment can be tried. We have treated 14 patients with intractable bilateral Menière's disease with intratympanic gentamicin (Garamycin, 40 mg/ml) administered initially in either 1, 2, 3, or 4 injections. The patients were tested at frequent intervals and followed up for 2 years. The work capacity, severity of vertigo, and gait difficulties were scored before treatment and during each test occasion. The postural stability was evaluated on a force platform and sway velocity was analyzed. Before treatment, moderate or severe reduction of work capacity was experienced by all of the subjects, which correlated to severity of vertigo and gait problems. Two years after treatment, the vertigo attacks were eliminated in 11 subjects and controlled in three subjects. The work capacity of three subjects was still moderately or severely reduced. This reduction depended on gait disturbance and Tumarkin attacks. The average postural stability returned to pretreatment level 2 years after the treatment was begun. The outcome of the caloric responses did not correlate with the outcome of the treatment. Hearing was not significantly affected by the treatment. Intratympanic gentamicin treatment is a relatively safe and effective way to treat bilateral Meniere's disease when the symptoms can be localized to one ear.

A syntactic analysis method for sinusoidal tracking eye movements

A syntactic method for the recognition of sinusoidal eye movements is presented. The method is based on a piecewise transformation of an eye movement signal to strings of a formal language and use of a formal grammar which generates the formal language. Gain and phase parameters of sinusoidal eye movements between a stimulus and its response are computed after the recognition of eye movements and after removing corrective saccades from the transformed signal. The method is used at an otoneurological laboratory.