Distinguishing normal, neuropathic and myopathic EMG with an automated machine learning approach

OBJECTIVE

Distinguishing normal, neuropathic and myopathic electromyography (EMG) traces can be challenging. We aimed to create an automated time series classification algorithm.

METHODS

EMGs of healthy controls (HC, n = 25), patients with amyotrophic lateral sclerosis (ALS, n = 20) and inclusion body myositis (IBM, n = 20), were retrospectively selected based on longitudinal clinical follow-up data (ALS and HC) or muscle biopsy (IBM). A machine learning pipeline was applied based on 5-second EMG fragments of each muscle. Diagnostic yield expressed as area under the curve (AUC) of a receiver-operator characteristics curve, accuracy, sensitivity, and specificity were determined per muscle (muscle-level) and per patient (patient-level).

RESULTS

Diagnostic yield of the classification ALS vs. HC was: AUC 0.834 ± 0.014 at muscle-level and 0.856 ± 0.009 at patient-level. For the classification HC vs. IBM, AUC was 0.744 ± 0.043 at muscle-level and 0.735 ± 0.029 at patient-level. For the classification ALS vs. IBM, AUC was 0.569 ± 0.024 at muscle-level and 0.689 ± 0.035 at patient-level.

CONCLUSIONS

An automated time series classification algorithm can distinguish EMGs from healthy individuals from those of patients with ALS with a high diagnostic yield. Using longer EMG fragments with different levels of muscle activation may improve performance.

SIGNIFICANCE

In the future, machine learning algorithms may help improve the diagnostic accuracy of EMG examinations.

Deriving reference values for nerve conduction studies from existing data using mixture model clustering

OBJECTIVE

to obtain locally valid reference values (RVs) from existing nerve conduction study (NCS) data.

METHODS

we used age, sex, height and limb temperature-based mixture model clustering (MMC) to identify normal and abnormal measurements on NCS data from two university hospitals. We compared MMC-derived RVs to published data; examined the effect of using different variables; validated MMC-derived RVs using independent data from 26 healthy control subjects and investigated their clinical applicability for the diagnosis of polyneuropathy.

RESULTS

MMC-derived RVs were similar to published RVs. Clustering can be achieved using only sex and age as variables. MMC is likely to yield reliable results with fewer abnormal than normal measurements and when the total number of measurements is at least 300. Measurements from healthy controls fell within the 95% MMC-derived prediction interval in 97.4% of cases.

CONCLUSIONS

MMC can be used to obtain RVs from existing data, providing a locally valid, accurate reflection of the (ab)normality of an NCS result.

SIGNIFICANCE

MMC can be used to generate locally valid RVs for any test for which sufficient data are available.¹.

Cold extremities in migraine: a marker for vascular dysfunction in women

Background and purpose

Migraine is recognized as a vascular risk factor, especially in women. Presumably, migraine, stroke and cardiovascular events share pathophysiological mechanisms. Self‐reported cold extremities were investigated as a marker for vascular dysfunction in migraine. Secondly, it was hypothesized that suffering from cold extremities affects sleep quality, possibly exacerbating migraine attack frequency.

Methods

In this case–control study, a random sample of 1084 migraine patients and 348 controls (aged 22–65 years) from the LUMINA migraine cohort were asked to complete questionnaires concerning cold extremities, sleep quality and migraine.

Results

A total of 594 migraine patients and 199 controls completed the questionnaires. In women, thermal discomfort and cold extremities (TDCE) were more often reported by migraineurs versus controls (odds ratio 2.3, 95% confidence interval 1.4–3.7; P < 0.001), but not significantly so in men (odds ratio 2.5, 95% confidence interval 0.9–6.9; P = 0.09). There was no difference in TDCE comparing migraine with or without aura. Female migraineurs who reported TDCE had higher attack frequencies compared to female migraineurs without TDCE (4 vs. 3 attacks per month; P = 0.003). The association between TDCE and attack frequency was mediated by the presence of difficulty initiating sleep (P = 0.02).

Conclusion

Women with migraine more often reported cold extremities compared with controls, possibly indicating a sex‐specific vascular vulnerability. Female migraineurs with cold extremities had higher attack frequencies, partly resulting from sleep disturbances. Future studies need to demonstrate whether cold extremities in female migraineurs are a predictor for cardiovascular and cerebrovascular events.

Sensitivity of MG-ADL for generalized weakness in myasthenia gravis

Background and purpose

Myasthenia gravis activities of daily living (MG‐ADL) is a commonly used questionnaire in MG trials. To investigate whether MG‐ADL is equally sensitive to oculobulbar and generalized weakness, its correlation with the oculobulbar and generalized domain of the quantitative myasthenia gravis (QMG) score was analyzed (QMGob and QMGgen, respectively). To test whether the sensitivity of MG‐ADL for generalized weakness could be improved, the additional value of ACTIVLIM on top of MG‐ADL in the prediction QMGgen in was investigated.

Methods

MG‐ADL, QMG and ACTIVLIM, an ADL questionnaire focusing on generalized weakness, were analyzed in a prospective cohort of 112 MG patients. A generalized linear model was used to calculate the correlation of MG‐ADL with QMGob and QMGgen and to assess the additional value of ACTIVLIM on top of MG‐ADL for its correlation with QMGgen.

Results

MG‐ADL had a higher correlation with QMGob than with QMGgen (B = 0.68, P < 0.001, and B = 0.38, P < 0.001, respectively). A similar trend was found for changes in the scores (B = 0.68, P = 0.132, and B = 0.39, P = 0.492, respectively). ACTIVLIM had a significant additional value on top of MG‐ADL in the prediction of QMGgen, both cross‐sectionally (B = −0.61, P < 0.001) and for changes within individual patients (B = −0.93, P = 0.041).

Conclusion

MG‐ADL has a lower sensitivity for generalized weakness than for oculobulbar weakness. Adding questions on generalized weakness would improve the sensitivity of the MG‐ADL for generalized weakness.

Nerve surgery and gene therapy: a neurobiological and clinical perspective

Despite major microsurgical improvements the clinical outcome of peripheral nerve surgery is still regarded as suboptimal. Over the past decade several innovative techniques have been developed to extend the armamentarium of the nerve surgeon. This review evaluates the potential of gene therapy in the context of peripheral nerve repair. First the main challenges impeding peripheral nerve regeneration are presented. This is followed by a short introduction to gene therapy and an overview of its most important advantages over the classical delivery of therapeutic proteins. Next, this review focuses on the most promising viral vectors capable of targeting the peripheral nervous system and their first application in animal models. In addition, the challenges of translating these experimental results to the clinic, the limitations of current vectors and the further developments needed, are discussed. Finally, four strategies are presented on how gene therapy could help patients that have to undergo reconstructive nerve surgery in the future.

Differential activation of intestinal gene promoters: functional interactions between GATA-5 and HNF-1 alpha

The effects of GATA-4, -5, and -6, hepatocyte nuclear factor-1 alpha (HNF-1 alpha) and -beta, and Cdx-2 on the rat and human lactase-phlorizin hydrolase (LPH) and human sucrase-isomaltase (SI) promoters were studied using transient cotransfection assays in Caco-2 cells. GATA factors and HNF-1 alpha were strong activators of the LPH promoters, whereas HNF-1 alpha and Cdx-2 were strong activators of the SI promoter, although GATA factors were also necessary for maximal activation of the SI gene. Cotransfection of GATA-5 and HNF-1 alpha together resulted in a higher activation of all three promoters than the sum of the activation by either factor alone, demonstrating functional cooperativity. In the human LPH promoter, an intact HNF-1 binding site was required for functional synergy. This study is the first to demonstrate 1) differential activation of the LPH and SI promoters by multiple transcription factors cotransfected singly and in combination and 2) that GATA and HNF-1 transcription factors cooperatively activate intestinal gene promoters. Synergistic activation is a mechanism by which higher levels of tissue-specific expression might be attained by overlapping expression of specific transcription factors.