Electrodiagnostic accuracy in polyneuropathies: supervised learning algorithms as a tool for practitioners.
Diagnostic accuracy
Electrodiagnosis
Polyneuropathies
Supervised learning algorithms
Journal
Neurological sciences : official journal of the Italian Neurological Society and of the Italian Society of Clinical Neurophysiology
ISSN: 1590-3478
Titre abrégé: Neurol Sci
Pays: Italy
ID NLM: 100959175
Informations de publication
Date de publication:
Dec 2020
Dec 2020
Historique:
received:
16
03
2020
accepted:
30
05
2020
pubmed:
11
6
2020
medline:
15
5
2021
entrez:
11
6
2020
Statut:
ppublish
Résumé
The interpretation of electrophysiological findings may lead to misdiagnosis in polyneuropathies. We investigated the electrodiagnostic accuracy of three supervised learning algorithms (SLAs): shrinkage discriminant analysis, multinomial logistic regression, and support vector machine (SVM), and three expert and three trainee neurophysiologists. We enrolled 434 subjects with the following diagnoses: chronic inflammatory demyelinating polyneuropathy (99), Charcot-Marie-Tooth disease type 1A (124), hereditary neuropathy with liability to pressure palsy (46), diabetic polyneuropathy (67), and controls (98). In each diagnostic class, 90% of subjects were used as training set for SLAs to establish the best performing SLA by tenfold cross validation procedure and 10% of subjects were employed as test set. Performance indicators were accuracy, precision, sensitivity, and specificity. SVM showed the highest overall diagnostic accuracy both in training and test sets (90.5 and 93.2%) and ranked first in a multidimensional comparison analysis. Overall accuracy of neurophysiologists ranged from 54.5 to 81.8%. This proof of principle study shows that SVM provides a high electrodiagnostic accuracy in polyneuropathies. We suggest that the use of SLAs in electrodiagnosis should be exploited to possibly provide a diagnostic support system especially helpful for the less experienced practitioners.
Identifiants
pubmed: 32518996
doi: 10.1007/s10072-020-04499-y
pii: 10.1007/s10072-020-04499-y
doi:
Types de publication
Journal Article
Langues
eng
Sous-ensembles de citation
IM
Pagination
3719-3727Références
Martyn CN, Hughes RA (1997) Epidemiology of peripheral neuropathy. J Neurol Neurosurg Psychiatry 62:310–318
doi: 10.1136/jnnp.62.4.310
Dyck PJ, Dyck PJ, Grant IA, Fealey RD (1996) Ten steps in characterizing and diagnosing patients with peripheral neuropathy. Neurology 47:10–17
doi: 10.1212/WNL.47.1.10
England JD, Asbury AK (2004) Peripheral neuropathy. Lancet 363:2151–2161
doi: 10.1016/S0140-6736(04)16508-2
England JD, Gronseth GS, Franklin G, Miller RG, Asbury AK, Carter GT, Cohen JA, Fisher MA, Howard JF, Kinsella LJ, Latov N, Lewis RA, Low PA, Sumner AJ (2005) Distal symmetric polyneuropathy: a definition for clinical research, report of the AAN, AAEM, and AAPM&R. Neurology 64:199–207
doi: 10.1212/01.WNL.0000149522.32823.EA
Watson JC, Dyck PJ (2015) Peripheral neuropathy: a practical approach to diagnosis and symptom management. Mayo Clin Proc 90:940–945
doi: 10.1016/j.mayocp.2015.05.004
Allen JA, Lewis RA (2015) CIDP diagnostic pitfalls and perception of treatment benefit. Neurology 85:498–504
doi: 10.1212/WNL.0000000000001833
Allen JA, Lewis RA (2018) Electrodiagnostic errors contribute to chronic inflammatory demyelinating polyneuropathy misdiagnosis. Muscle Nerve 57:542–549
doi: 10.1002/mus.25997
Gelinas D, Katz J, Nisbet B, England JD (2018) Current practice pattern in CIDP: a cross-sectional survey of neurologists in the United States. J Neurol Sci 397:84–91
doi: 10.1016/j.jns.2018.11.031
Nicolas G, Maisonobe T, Le Forestier N, Léger JM, Bouche P (2002) Proposed revised electrophysiological criteria for chronic inflammatory demyelinating polyradiculoneuropathy. Muscle Nerve 25:26–30
doi: 10.1002/mus.1214
Van den Bergh PY, Pieret F (2004) Electrodiagnostic criteria for acute and chronic inflammatory demyelinating polyradiculoneuropathy. Muscle Nerve 29:565–574
doi: 10.1002/mus.20022
Wilson J, Chawla J, Fisher M (2005) Sensitivity and specificity of electrodiagnostic criteria for CIDP using ROC curves: comparison to patients with diabetic and MGUS associated neuropathies. J Neurol Sci 231:19–28
doi: 10.1016/j.jns.2004.12.004
Rajabally YA, Nicolas G, Pieret F, Bouche P, van den Bergh PY (2009) Validity of diagnostic criteria for chronic inflammatory demyelinating polyneuropathy: a multicentre European study. J Neurol Neurosurg Psychiatry 80:1364–1368
doi: 10.1136/jnnp.2009.179358
Rondinelli RD, Stolov WC, Fujimoto WY, Osberg JS (1988) Electrodiagnosis of diabetic peripheral polyneuropathy. A multivariate analytic approach. Am J Phys Med Rehabil 67:12–23
doi: 10.1097/00002060-198802000-00004
Rondinelli RD, Robinson LR, Hassanein KM, Stolov WC, Fujimoto WY, Rubner DE (1994) Further studies on the electrodiagnosis of diabetic peripheral polyneuropathy using discriminant function analysis. Am J Phys Med Rehabil 73:116–1123
doi: 10.1097/00002060-199404000-00008
Uncini A, Ippoliti L, Shahrizaila N, Sekiguchi Y, Kuwabara S (2017) Optimizing the electrodiagnostic accuracy in Guillain-Barré syndrome subtypes: criteria sets and sparse linear discriminant analysis. Clin Neurophysiol 128:1176–1183
doi: 10.1016/j.clinph.2017.03.048
Joint Task Force of the EFNS and the PNS (2010) European Federation of Neurological Societies/Peripheral Nerve Society guideline on management of chronic inflammatory demyelinating polyneuropathy: report of a joint task force of the European Federation of Neurological Societies and the Peripheral Nerve Society-First Revision. J Peripher Nerv Syst 2010(15):1–9
doi: 10.1111/j.1529-8027.2010.00245.x
Saperstein DS, Amato AA, Wolfe GI, Katz JS, Nations SP, Jackson CE et al (1999) Multifocal acquired demyelinative sensory and motor neuropathy: the Lewis-Sumner syndrome. Muscle Nerve 22:560–566
doi: 10.1002/(SICI)1097-4598(199905)22:5<560::AID-MUS2>3.0.CO;2-Q
Katz JS, Saperstein DS, Gronseth G, Amato AA, Barohn RJ (2000) Distal acquired demyelinating symmetric neuropathy. Neurology 54:615–620
doi: 10.1212/WNL.54.3.615
Dyck PJ, Albers JW, Andersen H, Arezzo JC, Biessels GJ, Bril V, Feldman EL, Litchy WJ, O'Brien PC, Russell JW, on behalf of The Toronto Expert Panel on Diabetic Neuropathy (2011) Toronto expert panel on diabetic neuropathy. Diabetic polyneuropathies: update on research definition, diagnostic criteria and estimation of severity. Diabetes Metab Res Rev 27:620–628
doi: 10.1002/dmrr.1226
Vapnik VN (2000) The nature of statistical learning theory. Springer-Verlag
Hastie T, Tibshirani R, Friedman J (2009) The elements of statistical learning: data mining; inference and prediction, 2nd edn. Springer-Verlag, pp 241–251
Flach P (2012) Machine learning. The art and science of algorithms that make sense of data. Cambridge University Press, pp 53–58
Edwards A (1948) Note on the ‘correction for continuity’ in testing the significance of the difference between correlated proportions. Psychometrika 13:185–187
doi: 10.1007/BF02289261
Enders CK (2003) Using the expectation maximization algorithm to estimate coefficient alpha for scales with item-level missing data. Psychol Methods 8:322–337
doi: 10.1037/1082-989X.8.3.322
Theodoridis S (2015) Machine Learning: A Bayesian and Optimization Perspective. Academic Press p 314
James G, Vitten D, Hastie T, Tibshirani R (2013) An introduction to statistical learning: with applications in R. Springer, pp 24–26
Bromberg M (2011) Review of the evolution of the electrodiagnostic criteria for chronic inflammatory demyelinating neuropathy. Muscle Nerve 43:780–794
doi: 10.1002/mus.22038
Rajabally YA, Fowle AJ, Van den Bergh PY (2015) Which criteria for research in chronic inflammatory demyelinating polyneuropathy? An analysis of current practice. Muscle Nerve 51:932–933
doi: 10.1002/mus.24496