- Email: [email protected]
An European database for neuro-muscular diseases
SIOO [Z]
[3]
Pobter Sessiolzs
Haas W.F., Meyer M., “An automatic EMG decomposition system for routine clinical examinations and clinical research”, Computer-Aided Electromyography and Expert Systems, Elsevier, Amsterdam, 1989. Gut R., Moschytz G.S., “High-precision EMG signal decomposition using a modified Viterbi algorithm”, Abstracts of International Congress of Electromyography and Clinical Neurophysiology, Israel, 1992.
140-04
1 Usefulness of Myopathy Index in Quantitative Electromyography
Kazuhiko Hirose *, Toshiyuki Ohtake **, Tetsuo Komori **. * Tokyo Metropolitan Fuchu General Hospital; ** Tokyo Metropolitan Neurological Hospital Myopathy index (Ml) in quantitative electromyography (QEMG) was studied. The electromyogram for one second at maximal effort was analyzed by computer with the noise level of 50 fiV Ml consisted of the following 12 items: (1) the total number of spikes (TNS), (2) their mean amplitude, (3) maximum frequency in amplitude histogram, (4) its ratio to TNS, (5) increase of TNS with the change of noise level from 100 to 50 @Lv.(6) its ratio to TNS, (7) number of spikes of 50-100 FV in amplitude, (8) its ratio to TNS, (9) number of spikes of 50-200 @V in amplitude, (IO) its ratio to TNS, (11) number of spikes of 50-100 &V in amplitude and of l-2 msec in duration, (12) its ratio to TNS. MI was calculated bygiving 0.5 or 1 to 12 items according to the significant level of each item, and full Ml score was 10. Ml was applied to the biceps brachii muscle in several myopathies and usefulness of Ml was investigated. In QEMG. myopathic changes was found to be expressed initially as the increase of TNS and then as the increase of number of small and short spikes, Thereafter, TNS tended to decrease while the items on ratio to TNS inclined to increase. On Ml itself, high value was supposed to show the degree of severity of myopathy.
140-05
/ An European
Database for Neuro-Muscular
Diseases
B. Johnsen, A. Fuglsang-Frederiksen. S. Vingtoft, P Fawcett *, I. Schofield *, R. Liguori **, W. Nix ***, G. Otte #, G. Sieben #, M. Veloso ##, A. Vila ###. Dept. of Clinical NeurophysiologK University of Copenhagen, Gentofte Hospital, Copenhagen; * Dept of Clinical NeurophysiologL: Newcastle; **Dept. of Clinical Neurophysiology, Bologna; ***Dept. of Clinical NeurophysiologL: Mainz; # Dept. of Clinical Neurophysiology, Gent ## Dept. of Clinical Neurophysiologx Lisbon, ### Dept. of Clinical Neurophysiologx Grenoble The routine use of knowledge-based system (KBS) for EMG (Eiectromyography) is very sparse despite much research effort and the possibilities to improve the quality of the EMG examination. The development of more generally usefui KBS requires 1) a formalised evaluation procedure, 2) the consideration of interlaboratory differences and 3) the technical integration of these systems towards other equipment. These objectives are addressed in an European multicentre project ESTEEM (European Standardised Telematic tool to Evaluate EMG knowledge-based systems and Methods). Patient cases are collected in a common European database and a procedure to seek consensus on diagnoses and intermediate pathophysiological states is carried out among nine neurophysiologists. Differences in the diagnostic process are revealed by an analysis of used examination techniques, planning strategies, normal material and diagnostic criteria. The database will serve as a reference for the evaluation of diagnostic performance of future KBS. A general form for the representation of EMG data has been made. This form allows for the storage of data from all examination techniques used in routine at seven European EMG laboratories including diagnoses and intermediate pathophysiological states of anatomical structures.
40-06
Unbiased Acquisition
of Motor Unit Action Potentials
C. Krarup. K. Dahl, L.S. Jakobsen. A. Talbot, J.A. Sorensen. Dept of Clinical Neurophysiologfi Rigshospitalet, 2100 Copenhagen, Denmark Motor unit action potentials (MUPs) are sampled during weak effort in order to determine whether their duration, amplitude, and shape show abnormalities consistent with a neurogenic lesion or with myopathy. A trigger and delay function are used to record the 20-40 MUPs necessary for the analysis. This technique does, however, not ensure that a representative sample is
examined since large MUPs may be preferentially recorded. Analysis of ail MUPs independent of amplitude would reduce this error. A computer based analysis program has been developed to sort into separate classes the digitized MUPs (sampled at 25 kHz, A/D converted into 14 bits) recorded over a time period of IO set during weak effort The signal is divided into segments separated by silent periods which are then compared in pairs to compute the squared difference (@V2) between them. A minimum spanning tree (MST) is constructed using the squared difference between the segments. Classes of segments are determined by cutting at the edges of the MST. If a class contains more than a given number of segments it is used to define a MUP Those segments not fulfilling the criterion of a MUP are further analyzed by decomposition using combinations of the MUP templates obtained from the MST To evaluate the method, MUPs were recorded using conventional methods in parallel with the clustering technique. The main difference between the methods was acquisition of a larger number of small MUPs using the clustering technique. The technique allowed definition of even extremely complex polyphasic MUPs.
140~07
1 Body Composition Analysis with Bioelectric Impedance Measurement in Neuromuscular
Diseases
K.-H. Krause, C.-D. Reimers, N. Goebels, P Berlit. Friedrich-Baur-lnstitut, University of Munich, FIR G. Most of generalized neuromuscular diseases show atrophy of muscles and mesenchymal alteration processes. Bioelectric impedance analysis (BIA) is a fast and well tolerated procedure for measurement of body composition (fat, water) (Baumgartner, R. N. et al., J. Clin. Nutr. 48: 16, 1988). BIA was assessed in 207 adult patients (110 males, 97 females) with neuromuscular diseases, confirmed by clinical symptoms, electromyography, and muscle biopsy. The results were compared with those of 118 healthy controls (65 males, 53 females). Using multiple regression and correlation analysis for sex, age, height, and weight normal values with standard deviations (SD) were established for the control group. The median value of relative body fat [% of body weight] in male patients lay 3.8 SD (p < 0.001) (Mann-Whitney-U-test), in female patients 1 .O SD (p < 0.001) above that of controls, the median value of relative body water 1% of body weight] in male patients 2.5 SD (p < 0.001). and in females 1.1 SD (p x 0.001) below that of the controls, 30.1% of the male and 21.4% of the female patients showed a minimal SD of 3.0 in percent body fat, compared with 7.7% of males and 1.9% of females in the control group. The deviations were highly correlated with the severity of muscle weakness. In conclusion, BIA seems to be a very simple and fast, but not very sensitive method for assessing the degree of muscle lipomatosis in generalized neuromuscular diseases.
/ 40 08
Correlations Between Surface and Near-Nerve Recorded Tactile Evoked Responses
J. Nilsson, M. Panizza, G. Grioni. Fondazione Goffredo (MN), Italy
Clinica Del Lavoro, IRCCS, Caste1
In electrical stimulation of the median nerve at the tip of digit Ill, the stimulus spreads and tends to depolarize the nerve at a point more proximal than the stimulating electrode. Tactile stimulation activates the most distal nerve segment, because adequate tapping of touch receptors generates an action potential that can be recorded along the nerve using near-nerve or surface electrodes. In ten healthy subjects the maximum sensory conduction velocity evoked by electrical stimulation to the tip of digit Ill was 58.1 m/s (SD 5.6 m/s) when recorded via near-nerve needles, and 59.1 m/s (SD 5.8 m/s) via surface electrodes at the wrist, By tapping the tip of digit Ill, the mean conduction velocity was 42.9 m/s (SD 5.4 m/s) for near-nerve recordings, and 44.9 m/s (SD 5.1 m/s) for surface recordings. The compound action potential obtained via needle electrodes had a mean peak-to-peak amplitude of 9.3 @V (SD 3.1 FV), and the mean amplitude of the negative peak for surface recorded potentials was 13.6 pV (SD 4.9 WV) with a mean duration of 1.2 ms (SD 0.2 ms). Our studies showed that surface recorded potentials elicited by tactile stimulation behave similarly to responses recorded via near-nerve needles. Thus, this technique could supplement routine nerve conduction studies and may prove valuable as a non-invasive tool for early detection of lesions involving the most distal branches of peripheral nerves.
[3]
Pobter Sessiolzs
Haas W.F., Meyer M., “An automatic EMG decomposition system for routine clinical examinations and clinical research”, Computer-Aided Electromyography and Expert Systems, Elsevier, Amsterdam, 1989. Gut R., Moschytz G.S., “High-precision EMG signal decomposition using a modified Viterbi algorithm”, Abstracts of International Congress of Electromyography and Clinical Neurophysiology, Israel, 1992.
140-04
1 Usefulness of Myopathy Index in Quantitative Electromyography
Kazuhiko Hirose *, Toshiyuki Ohtake **, Tetsuo Komori **. * Tokyo Metropolitan Fuchu General Hospital; ** Tokyo Metropolitan Neurological Hospital Myopathy index (Ml) in quantitative electromyography (QEMG) was studied. The electromyogram for one second at maximal effort was analyzed by computer with the noise level of 50 fiV Ml consisted of the following 12 items: (1) the total number of spikes (TNS), (2) their mean amplitude, (3) maximum frequency in amplitude histogram, (4) its ratio to TNS, (5) increase of TNS with the change of noise level from 100 to 50 @Lv.(6) its ratio to TNS, (7) number of spikes of 50-100 FV in amplitude, (8) its ratio to TNS, (9) number of spikes of 50-200 @V in amplitude, (IO) its ratio to TNS, (11) number of spikes of 50-100 &V in amplitude and of l-2 msec in duration, (12) its ratio to TNS. MI was calculated bygiving 0.5 or 1 to 12 items according to the significant level of each item, and full Ml score was 10. Ml was applied to the biceps brachii muscle in several myopathies and usefulness of Ml was investigated. In QEMG. myopathic changes was found to be expressed initially as the increase of TNS and then as the increase of number of small and short spikes, Thereafter, TNS tended to decrease while the items on ratio to TNS inclined to increase. On Ml itself, high value was supposed to show the degree of severity of myopathy.
140-05
/ An European
Database for Neuro-Muscular
Diseases
B. Johnsen, A. Fuglsang-Frederiksen. S. Vingtoft, P Fawcett *, I. Schofield *, R. Liguori **, W. Nix ***, G. Otte #, G. Sieben #, M. Veloso ##, A. Vila ###. Dept. of Clinical NeurophysiologK University of Copenhagen, Gentofte Hospital, Copenhagen; * Dept of Clinical NeurophysiologL: Newcastle; **Dept. of Clinical Neurophysiology, Bologna; ***Dept. of Clinical NeurophysiologL: Mainz; # Dept. of Clinical Neurophysiology, Gent ## Dept. of Clinical Neurophysiologx Lisbon, ### Dept. of Clinical Neurophysiologx Grenoble The routine use of knowledge-based system (KBS) for EMG (Eiectromyography) is very sparse despite much research effort and the possibilities to improve the quality of the EMG examination. The development of more generally usefui KBS requires 1) a formalised evaluation procedure, 2) the consideration of interlaboratory differences and 3) the technical integration of these systems towards other equipment. These objectives are addressed in an European multicentre project ESTEEM (European Standardised Telematic tool to Evaluate EMG knowledge-based systems and Methods). Patient cases are collected in a common European database and a procedure to seek consensus on diagnoses and intermediate pathophysiological states is carried out among nine neurophysiologists. Differences in the diagnostic process are revealed by an analysis of used examination techniques, planning strategies, normal material and diagnostic criteria. The database will serve as a reference for the evaluation of diagnostic performance of future KBS. A general form for the representation of EMG data has been made. This form allows for the storage of data from all examination techniques used in routine at seven European EMG laboratories including diagnoses and intermediate pathophysiological states of anatomical structures.
40-06
Unbiased Acquisition
of Motor Unit Action Potentials
C. Krarup. K. Dahl, L.S. Jakobsen. A. Talbot, J.A. Sorensen. Dept of Clinical Neurophysiologfi Rigshospitalet, 2100 Copenhagen, Denmark Motor unit action potentials (MUPs) are sampled during weak effort in order to determine whether their duration, amplitude, and shape show abnormalities consistent with a neurogenic lesion or with myopathy. A trigger and delay function are used to record the 20-40 MUPs necessary for the analysis. This technique does, however, not ensure that a representative sample is
examined since large MUPs may be preferentially recorded. Analysis of ail MUPs independent of amplitude would reduce this error. A computer based analysis program has been developed to sort into separate classes the digitized MUPs (sampled at 25 kHz, A/D converted into 14 bits) recorded over a time period of IO set during weak effort The signal is divided into segments separated by silent periods which are then compared in pairs to compute the squared difference (@V2) between them. A minimum spanning tree (MST) is constructed using the squared difference between the segments. Classes of segments are determined by cutting at the edges of the MST. If a class contains more than a given number of segments it is used to define a MUP Those segments not fulfilling the criterion of a MUP are further analyzed by decomposition using combinations of the MUP templates obtained from the MST To evaluate the method, MUPs were recorded using conventional methods in parallel with the clustering technique. The main difference between the methods was acquisition of a larger number of small MUPs using the clustering technique. The technique allowed definition of even extremely complex polyphasic MUPs.
140~07
1 Body Composition Analysis with Bioelectric Impedance Measurement in Neuromuscular
Diseases
K.-H. Krause, C.-D. Reimers, N. Goebels, P Berlit. Friedrich-Baur-lnstitut, University of Munich, FIR G. Most of generalized neuromuscular diseases show atrophy of muscles and mesenchymal alteration processes. Bioelectric impedance analysis (BIA) is a fast and well tolerated procedure for measurement of body composition (fat, water) (Baumgartner, R. N. et al., J. Clin. Nutr. 48: 16, 1988). BIA was assessed in 207 adult patients (110 males, 97 females) with neuromuscular diseases, confirmed by clinical symptoms, electromyography, and muscle biopsy. The results were compared with those of 118 healthy controls (65 males, 53 females). Using multiple regression and correlation analysis for sex, age, height, and weight normal values with standard deviations (SD) were established for the control group. The median value of relative body fat [% of body weight] in male patients lay 3.8 SD (p < 0.001) (Mann-Whitney-U-test), in female patients 1 .O SD (p < 0.001) above that of controls, the median value of relative body water 1% of body weight] in male patients 2.5 SD (p < 0.001). and in females 1.1 SD (p x 0.001) below that of the controls, 30.1% of the male and 21.4% of the female patients showed a minimal SD of 3.0 in percent body fat, compared with 7.7% of males and 1.9% of females in the control group. The deviations were highly correlated with the severity of muscle weakness. In conclusion, BIA seems to be a very simple and fast, but not very sensitive method for assessing the degree of muscle lipomatosis in generalized neuromuscular diseases.
/ 40 08
Correlations Between Surface and Near-Nerve Recorded Tactile Evoked Responses
J. Nilsson, M. Panizza, G. Grioni. Fondazione Goffredo (MN), Italy
Clinica Del Lavoro, IRCCS, Caste1
In electrical stimulation of the median nerve at the tip of digit Ill, the stimulus spreads and tends to depolarize the nerve at a point more proximal than the stimulating electrode. Tactile stimulation activates the most distal nerve segment, because adequate tapping of touch receptors generates an action potential that can be recorded along the nerve using near-nerve or surface electrodes. In ten healthy subjects the maximum sensory conduction velocity evoked by electrical stimulation to the tip of digit Ill was 58.1 m/s (SD 5.6 m/s) when recorded via near-nerve needles, and 59.1 m/s (SD 5.8 m/s) via surface electrodes at the wrist, By tapping the tip of digit Ill, the mean conduction velocity was 42.9 m/s (SD 5.4 m/s) for near-nerve recordings, and 44.9 m/s (SD 5.1 m/s) for surface recordings. The compound action potential obtained via needle electrodes had a mean peak-to-peak amplitude of 9.3 @V (SD 3.1 FV), and the mean amplitude of the negative peak for surface recorded potentials was 13.6 pV (SD 4.9 WV) with a mean duration of 1.2 ms (SD 0.2 ms). Our studies showed that surface recorded potentials elicited by tactile stimulation behave similarly to responses recorded via near-nerve needles. Thus, this technique could supplement routine nerve conduction studies and may prove valuable as a non-invasive tool for early detection of lesions involving the most distal branches of peripheral nerves.