Artificial intelligence utilising corneal confocal microscopy for the diagnosis of peripheral neuropathy in diabetes mellitus and prediabetes

AIMS/HYPOTHESIS

We aimed to develop an artificial intelligence (AI)-based deep learning algorithm (DLA) applying attribution methods without image segmentation to corneal confocal microscopy images and to accurately classify peripheral neuropathy (or lack of).

METHODS

The AI-based DLA utilised convolutional neural networks with data augmentation to increase the algorithm's generalisability. The algorithm was trained using a high-end graphics processor for 300 epochs on 329 corneal nerve images and tested on 40 images (1 image/participant). Participants consisted of healthy volunteer (HV) participants (n = 90) and participants with type 1 diabetes (n = 88), type 2 diabetes (n = 141) and prediabetes (n = 50) (defined as impaired fasting glucose, impaired glucose tolerance or a combination of both), and were classified into HV, those without neuropathy (PN-) (n = 149) and those with neuropathy (PN+) (n = 130). For the AI-based DLA, a modified residual neural network called ResNet-50 was developed and used to extract features from images and perform classification. The algorithm was tested on 40 participants (15 HV, 13 PN-, 12 PN+). Attribution methods gradient-weighted class activation mapping (Grad-CAM), Guided Grad-CAM and occlusion sensitivity displayed the areas within the image that had the greatest impact on the decision of the algorithm.

RESULTS

The results were as follows: HV: recall of 1.0 (95% CI 1.0, 1.0), precision of 0.83 (95% CI 0.65, 1.0), F₁-score of 0.91 (95% CI 0.79, 1.0); PN-: recall of 0.85 (95% CI 0.62, 1.0), precision of 0.92 (95% CI 0.73, 1.0), F₁-score of 0.88 (95% CI 0.71, 1.0); PN+: recall of 0.83 (95% CI 0.58, 1.0), precision of 1.0 (95% CI 1.0, 1.0), F₁-score of 0.91 (95% CI 0.74, 1.0). The features displayed by the attribution methods demonstrated more corneal nerves in HV, a reduction in corneal nerves for PN- and an absence of corneal nerves for PN+ images.

CONCLUSIONS/INTERPRETATION

We demonstrate promising results in the rapid classification of peripheral neuropathy using a single corneal image. A large-scale multicentre validation study is required to assess the utility of AI-based DLA in screening and diagnostic programmes for diabetic neuropathy.

Corneal nerve loss in patients with TIA and acute ischemic stroke in relation to circulating markers of inflammation and vascular integrity

Vascular and inflammatory mechanisms are implicated in the development of cerebrovascular disease and corneal nerve loss occurs in patients with transient ischemic attack (TIA) and acute ischemic stroke (AIS). We have assessed whether serum markers of inflammation and vascular integrity are associated with the severity of corneal nerve loss in patients with TIA and AIS. Corneal confocal microscopy (CCM) was performed to quantify corneal nerve fiber density (CNFD), corneal nerve branch density (CNBD) and corneal nerve fiber length (CNFL) in 105 patients with TIA (n = 24) or AIS (n = 81) and age matched control subjects (n = 56). Circulating levels of IL-6, MMP-2, MMP-9, E-Selectin, P-Selectin and VEGF were quantified in patients within 48 h of presentation with a TIA or AIS. CNFL (P = 0.000, P = 0.000), CNFD (P = 0.122, P = 0.000) and CNBD (P = 0.002, P = 0.000) were reduced in patients with TIA and AIS compared to controls, respectively with no difference between patients with AIS and TIA. The NIHSS Score (P = 0.000), IL-6 (P = 0.011) and E-Selectin (P = 0.032) were higher in patients with AIS compared to TIA with no difference in MMP-2 (P = 0.636), MMP-9 (P = 0.098), P-Selectin (P = 0.395) and VEGF (P = 0.831). CNFL (r = 0.218, P = 0.026) and CNFD (r = 0.230, P = 0.019) correlated with IL-6 and multiple regression analysis showed a positive association of CNFL and CNFD with IL-6 (P = 0.041, P = 0.043). Patients with TIA and AIS have evidence of corneal nerve loss and elevated IL6 and E-selectin levels. Larger longitudinal studies are required to determine the association between inflammatory and vascular markers and corneal nerve fiber loss in patients with cerebrovascular disease.

Breath Analysis for the In Vivo Detection of Diabetic Ketoacidosis

Human breath analysis of volatile organic compounds has gained significant attention recently because of its rapid and noninvasive potential to detect various metabolic diseases. The detection of ketones in the breath and blood is key to diagnosing and managing diabetic ketoacidosis (DKA) in patients with type 1 diabetes. It may also be of increasing importance to detect euglycemic ketoacidosis in patients with type 1 or type 2 diabetes or heart failure, treated with sodium-glucose transporter-2 inhibitors (SGLT2-i). The present research evaluates the efficiency of colorimetry for detecting acetone and ethanol in exhaled human breath with the response time, pH effect, temperature effect, concentration effect, and selectivity of dyes. Using the proposed multidye system, we obtained a detection limit of 0.0217 ppm for acetone and 0.029 ppm for ethanol in the detection range of 0.05-50 ppm. A smartphone-assisted unit consisting of a portable colorimetric device was used to detect relative red/green/blue values within 60 s of the interface for practical and real-time application. The developed method could be used for rapid, low-cost detection of ketones in patients with type 1 diabetes and DKA and patients with type 1 or type 2 diabetes or heart failure treated with SGLT2-I and euglycemic ketoacidosis.

Abnormal corneal nerve morphology and brain volume in patients with schizophrenia

Neurodevelopmental and neurodegenerative pathology occur in Schizophrenia. This study compared the utility of corneal confocal microscopy (CCM), an ophthalmic imaging technique with MRI brain volumetry in quantifying neuronal pathology and its relationship to cognitive dysfunction and symptom severity in schizophrenia. Thirty-six subjects with schizophrenia and 26 controls underwent assessment of cognitive function, symptom severity, CCM and MRI brain volumetry. Subjects with schizophrenia had lower cognitive function (P ≤ 0.01), corneal nerve fiber density (CNFD), length (CNFL), branch density (CNBD), CNBD:CNFD ratio (P < 0.0001) and cingulate gyrus volume (P < 0.05) but comparable volume of whole brain (P = 0.61), cortical gray matter (P = 0.99), ventricle (P = 0.47), hippocampus (P = 0.10) and amygdala (P = 0.68). Corneal nerve measures and cingulate gyrus volume showed no association with symptom severity (P = 0.35–0.86 and P = 0.50) or cognitive function (P = 0.35–0.86 and P = 0.49). Corneal nerve measures were not associated with metabolic syndrome (P = 0.61–0.64) or diabetes (P = 0.057–0.54). The area under the ROC curve distinguishing subjects with schizophrenia from controls was 88% for CNFL, 84% for CNBD and CNBD:CNFD ratio, 79% for CNFD and 73% for the cingulate gyrus volume. This study has identified a reduction in corneal nerve fibers and cingulate gyrus volume in schizophrenia, but no association with symptom severity or cognitive dysfunction. Corneal nerve loss identified using CCM may act as a rapid non-invasive surrogate marker of neurodegeneration in patients with schizophrenia.

Corneal confocal microscopy for the diagnosis of diabetic peripheral neuropathy: A systematic review and meta-analysis

INTRODUCTION

Corneal confocal microscopy (CCM) is a rapid non-invasive ophthalmic imaging technique that identifies corneal nerve fiber damage. Small studies suggest that CCM could be used to assess patients with diabetic peripheral neuropathy (DPN).

AIM

To undertake a systematic review and meta-analysis assessing the diagnostic utility of CCM for sub-clinical DPN (DPN- ) and established DPN (DPN+ ).

DATA SOURCES

Databases (PubMed, Embase, Central, ProQuest) were searched for studies using CCM in patients with diabetes up to April 2020.

STUDY SELECTION

Studies were included if they reported on at least one CCM parameter in patients with diabetes.

DATA EXTRACTION

Corneal nerve fiber density (CNFD), corneal nerve branch density (CNBD), corneal nerve fiber length (CNFL), and inferior whorl length (IWL) were compared between patients with diabetes with and without DPN and controls. Meta-analysis was undertaken using RevMan V.5.3.

DATA SYNTHESIS

Thirty-eight studies including ~4,000 participants were included in this meta-analysis. There were significant reductions in CNFD, CNBD, CNFL, and IWL in DPN- vs controls (P < 0.00001), DPN+ vs controls (P < 0.00001), and DPN+ vs DPN- (P < 0.00001).

CONCLUSION

This systematic review and meta-analysis shows that CCM detects small nerve fiber loss in subclinical and clinical DPN and concludes that CCM has good diagnostic utility in DPN.

COVID-19 and the hidden threat of diabetic microvascular complications

The coronavirus disease 2019 (COVID-19) pandemic affected at least 200 million individuals worldwide and resulted in nearly 5 million deaths as of October 2021. According to the latest data from the International Diabetes Federation (IDF) in 2021, the diabetes pandemic has affected 537 million people and is associated with 6.7 million deaths. Given the high prevalence of both diabetes and COVID-19 and common pathological outcomes, a bidirectional relationship could have a catastrophic outcome. The increased risk of COVID-19 in those with obesity and diabetes and higher morbidity and mortality has received considerable attention. However, little attention has been given to the relationship between COVID-19 and microvascular complications. Indeed, microvascular complications are associated with an increased risk of cardiovascular disease (CVD) and mortality in diabetes. This review assesses the evidence for an association between diabetic microvascular complications (neuropathy, nephropathy, and retinopathy) and COVID-19. It draws parallels between the pathological changes occurring in the microvasculature in both diseases and assesses whether microvascular disease is a prognostic factor for COVID-19 outcomes in diabetes.

Loss of corneal nerves and brain volume in mild cognitive impairment and dementia

Introduction

This study compared the capability of corneal confocal microscopy (CCM) with magnetic resonance imaging (MRI) brain volumetry for the diagnosis of mild cognitive impairment (MCI) and dementia.

Methods

In this cross‐sectional study, participants with no cognitive impairment (NCI), MCI, and dementia underwent assessment of Montreal Cognitive Assessment (MoCA), MRI brain volumetry, and CCM.

Results

Two hundred eight participants with NCI (n = 42), MCI (n = 98), and dementia (n = 68) of comparable age and gender were studied. For MCI, the area under the curve (AUC) of CCM (76% to 81%), was higher than brain volumetry (52% to 70%). For dementia, the AUC of CCM (77% to 85%), was comparable to brain volumetry (69% to 93%). Corneal nerve fiber density, length, branch density, whole brain, hippocampus, cortical gray matter, thalamus, amygdala, and ventricle volumes were associated with cognitive impairment after adjustment for confounders (All P’s < .01).

Discussion

The diagnostic capability of CCM compared to brain volumetry is higher for identifying MCI and comparable for dementia, and abnormalities in both modalities are associated with cognitive impairment.

The Impact of Suprascapular Nerve Interventions in Patients with Frozen Shoulder: A Systematic Review and Meta-Analysis

BACKGROUND

Frozen shoulder is a common condition resulting in severe pain and restricted range of motion. An assessment of the effectiveness of interventions may provide an improved understanding of the development and management of frozen shoulder.

METHODS

A literature search was conducted using Embase, the Cumulative Index of Nursing and Allied Health (CINAHL), the Cochrane Central Register of Controlled Trials (CENTRAL), and National Center for Biotechnology Information PubMed using relevant terms. Studies were included if they assessed the outcomes of interventions on the suprascapular nerve that aim to improve the symptoms of frozen shoulder.

RESULTS

A database search returned 196 articles. After review, 9 articles met the inclusion criteria and were included in the analysis. Suprascapular nerve interventions (nerve block, pulsed radiofrequency lesioning) are associated with improvement in pain, motion, and function. Meta-analysis showed that pain (Hedges g, -3.084 [95% confidence interval (CI), -4.273 to -1.894]; p < 0.001) and range of motion (Hedges g, 2.204 [95% CI, 0.992 to 3.415]; p < 0.001) improved significantly following suprascapular nerve block (SSNB).

CONCLUSIONS

SSNB is associated with significant improvements in shoulder pain and range of motion in patients with frozen shoulder. Further randomized controlled trials comparing SSNB with intra-articular injection and other nonoperative treatments are required to fully define its role in the management of frozen shoulder.

LEVEL OF EVIDENCE

Therapeutic Level III. See Instructions for Authors for a complete description of levels of evidence.

Corneal nerve fiber loss relates to cognitive impairment in patients with Parkinson’s disease

Cognitive impairment in Parkinson’s disease (PD) adversely influences quality of life. There is currently no available biomarker to predict cognitive decline in PD. Corneal confocal microscopy (CCM) has been used as a non-invasive tool for quantifying small nerve damage in PD. The present study investigated whether corneal nerve measures were associated with cognitive function in PD. Patients with PD were classified into those with normal cognitive function (PD-CN), mild cognitive impairment (PD-MCI), and dementia (PDD). Corneal nerve fiber density (CNFD), corneal nerve branch density (CNBD), and corneal nerve fiber length (CNFL) were quantified with CCM and compared with a control group. Sixty-five PD patients and thirty controls were studied. CNFD was decreased and CNBD was increased in PD patients compared to controls (P < 0.05). CNBD and CNBD/CNFD ratio was higher in PD-CN compared to controls. CNFD was positively correlated with the Montreal cognitive assessment (MoCA) score (r = 0.683, P < 0.001), but negatively associated with unified Parkinson disease rating scale (UPDRS)-part III (r = −0.481, P < 0.001) and total UPDRS scores (r = −0.401, P = 0.001) in PD patients. There was no correlation between CNFD and Levodopa equivalent daily dose (LEDD) (r = 0.176, P = 0.161). CNFD, CNBD, CNFL, and CNBD/CNFD ratio was lower with increasing Hoehn and Yahr stage. PD patients show evidence of corneal nerve loss compared with controls and corneal nerve parameters are associated with the severity of cognitive and motor dysfunction in PD. CCM could serve as an objective in vivo ophthalmic imaging technique to assess neurodegeneration in PD.

Chemotherapy-Induced Peripheral Neuropathy: Epidemiology, Pathomechanisms and Treatment

PURPOSE

This review provides an update on the current clinical, epidemiological and pathophysiological evidence alongside the diagnostic, prevention and treatment approach to chemotherapy-induced peripheral neuropathy (CIPN).

FINDINGS

The incidence of cancer and long-term survival after treatment is increasing. CIPN affects sensory, motor and autonomic nerves and is one of the most common adverse events caused by chemotherapeutic agents, which in severe cases leads to dose reduction or treatment cessation, with increased mortality. The primary classes of chemotherapeutic agents associated with CIPN are platinum-based drugs, taxanes, vinca alkaloids, bortezomib and thalidomide. Platinum agents are the most neurotoxic, with oxaliplatin causing the highest prevalence of CIPN. CIPN can progress from acute to chronic, may deteriorate even after treatment cessation (a phenomenon known as coasting) or only partially attenuate. Different chemotherapeutic agents share both similarities and key differences in pathophysiology and clinical presentation. The diagnosis of CIPN relies heavily on identifying symptoms, with limited objective diagnostic approaches targeting the class of affected nerve fibres. Studies have consistently failed to identify at-risk cohorts, and there are no proven strategies or interventions to prevent or limit the development of CIPN. Furthermore, multiple treatments developed to relieve symptoms and to modify the underlying disease in CIPN have failed.

IMPLICATIONS

The increasing prevalence of CIPN demands an objective approach to identify at-risk patients in order to prevent or limit progression and effectively alleviate the symptoms associated with CIPN. An evidence base for novel targets and both pharmacological and non-pharmacological treatments is beginning to emerge and has been recognised recently in publications by the American Society of Clinical Oncology and analgesic trial design expert groups such as ACTTION.

Subclinical Corneal Nerve Fiber Damage and Immune Cell Activation in Systemic Lupus Erythematosus: A Corneal Confocal Microscopy Study

Purpose

The purpose of this study was to evaluate the utility of corneal confocal microscopy (CCM) in identifying small nerve fiber damage and immune cell activation in patients with systemic lupus erythematosus (SLE).

Methods

This cross-sectional comparative study included 39 consecutive patients with SLE and 30 healthy control participants. Central corneal sensitivity was assessed using a Cochet-Bonnet contact corneal esthesiometer and a laser scanning CCM (Heidelberg, Germany) was used to quantify corneal nerve fiber density (CNFD), nerve branch density (CNBD), nerve fiber length (CNFL), and Langerhans cell (LC) density.

Results

Age was comparable among patients with SLE (33.7 ± 12.7) and controls (35.0 ± 13.7 years, P = 0.670) and the median duration of disease was 3.0 years (2.0–10.0 years). CNBD (P = 0.003) and CNFL (P = 0.019) were lower and mature LC density (P = 0.002) was higher, but corneal sensitivity (P = 0.178) and CNFD (P = 0.198) were comparable in patients with SLE compared with controls. The SELENA-SLEDAI score correlated with CNFD (ρ = −0.319, P = 0.048) and CNFL (ρ = −0.373, P = 0.019), and the total and immature LC densities correlated with CNBD (ρ = −0.319. P = 0.048, and ρ = −0.328, P = 0.041, respectively). Immature LC density was higher (P = 0.025), but corneal sensitivity and nerve fiber parameters were comparable between patients with (33%) and without neuropsychiatric symptoms and SLE.

Conclusions

Corneal confocal microscopy identifies distal corneal nerve fiber loss and increased immune cell density in patients with SLE and corneal nerve loss was associated with disease activity.

Translational Relevance

Corneal confocal microscopy may enable the detection of subclinical corneal nerve loss and immune cell activation in SLE.

Superior Non-Invasive Glucose Sensor Using Bimetallic CuNi Nanospecies Coated Mesoporous Carbon

The assessment of blood glucose levels is necessary for the diagnosis and management of diabetes. The accurate quantification of serum or plasma glucose relies on enzymatic and nonenzymatic methods utilizing electrochemical biosensors. Current research efforts are focused on enhancing the non-invasive detection of glucose in sweat with accuracy, high sensitivity, and stability. In this work, nanostructured mesoporous carbon coupled with glucose oxidase (GO_x) increased the direct electron transfer to the electrode surface. A mixed alloy of CuNi nanoparticle-coated mesoporous carbon (CuNi-MC) was synthesized using a hydrothermal process followed by annealing at 700 °C under the flow of argon gas. The prepared catalyst's crystal structure and morphology were explored using X-ray diffraction and high-resolution transmission electron microscopy. The electrocatalytic activity of the as-prepared catalyst was investigated using cyclic voltammetry (CV) and amperometry. The findings show an excellent response time of 4 s and linear range detection from 0.005 to 0.45 mM with a high electrode sensitivity of 11.7 ± 0.061 mA mM cm^-2 in a selective medium.

Corneal confocal microscopy demonstrates axonal loss in different courses of multiple sclerosis

Axonal loss is the main determinant of disease progression in multiple sclerosis (MS). This study aimed to assess the utility of corneal confocal microscopy (CCM) in detecting corneal axonal loss in different courses of MS. The results were confirmed by two independent segmentation methods. 72 subjects (144 eyes) [(clinically isolated syndrome (n = 9); relapsing–remitting MS (n = 20); secondary-progressive MS (n = 22); and age-matched, healthy controls (n = 21)] underwent CCM and assessment of their disability status. Two independent algorithms (ACCMetrics; and Voxeleron deepNerve) were used to quantify corneal nerve fiber density (CNFD) (ACCMetrics only), corneal nerve fiber length (CNFL) and corneal nerve fractal dimension (CNFrD). Data are expressed as mean ± standard deviation with 95% confidence interval (CI). Compared to controls, patients with MS had significantly lower CNFD (34.76 ± 5.57 vs. 19.85 ± 6.75 fibers/mm², 95% CI − 18.24 to − 11.59, P < .0001), CNFL [for ACCMetrics: 19.75 ± 2.39 vs. 12.40 ± 3.30 mm/mm², 95% CI − 8.94 to − 5.77, P < .0001; for deepNerve: 21.98 ± 2.76 vs. 14.40 ± 4.17 mm/mm², 95% CI − 9.55 to − 5.6, P < .0001] and CNFrD [for ACCMetrics: 1.52 ± 0.02 vs. 1.45 ± 0.04, 95% CI − 0.09 to − 0.05, P < .0001; for deepNerve: 1.29 ± 0.03 vs. 1.19 ± 0.07, 95% − 0.13 to − 0.07, P < .0001]. Corneal nerve parameters were comparably reduced in different courses of MS. There was excellent reproducibility between the algorithms. Significant corneal axonal loss is detected in different courses of MS including patients with clinically isolated syndrome.

Insulin resistance limits corneal nerve regeneration in patients with type 2 diabetes undergoing intensive glycemic control

AIMS/INTRODUCTION

This study aimed to investigate whether insulin resistance (IR) in individuals with type 2 diabetes undergoing intensive glycemic control determines the extent of improvement in neuropathy.

MATERIALS AND METHODS

This was an exploratory substudy of an open-label, randomized controlled trial of individuals with poorly controlled type 2 diabetes treated with exenatide and pioglitazone or insulin to achieve a glycated hemoglobin <7.0% (<53 mmol/mol). Baseline IR was defined using homeostasis model assessment of IR, and change in neuropathy was assessed using corneal confocal microscopy.

RESULTS

A total of 38 individuals with type 2 diabetes aged 50.2 ± 8.5 years with (n = 25, 66%) and without (n = 13, 34%) IR were studied. There was a significant decrease in glycated hemoglobin (P < 0.0001), diastolic blood pressure (P < 0.0001), total cholesterol (P < 0.01) and low-density lipoprotein (P = 0.05), and an increase in bodyweight (P < 0.0001) with treatment. Individuals with homeostasis model assessment of IR <1.9 showed a significant increase in corneal nerve fiber density (P ≤ 0.01), length (P ≤ 0.01) and branch density (P ≤ 0.01), whereas individuals with homeostasis model assessment of IR ≥1.9 showed no change. IR was negatively associated with change in corneal nerve fiber density after adjusting for change in bodyweight (P < 0.05).

CONCLUSIONS

Nerve regeneration might be limited in individuals with type 2 diabetes and IR undergoing treatment with pioglitazone plus exenatide or insulin to improve glycemic control.

Assessment of optical coherence tomography angiography and multifocal electroretinography in eyes with and without nonproliferative diabetic retinopathy

Purpose:

To examine (i) the retinal structure and function using optical coherence tomography angiography (OCTA) and multifocal electroretinography (mfERG), respectively, in eyes with and without nonproliferative diabetic retinopathy (NPDR), (ii) and their interrelationship between retinal structure (OCTA) and function (mfERG) in the two groups independently.

Methods:

This was a prospective observational study. One hundred twenty-one eligible participants with type 2 diabetes with No DR (n = 89), or with mild or moderate NPDR (n = 32) underwent ophthalmic examination, ultrawide field-view fundus photography, OCTA, and mfERG. Group differences were assessed using a Mann–Whitney U test. Correlations were assessed using Spearman's rho.

Results:

There were no significant differences in OCTA measures between the two groups. The mfERG P1 implicit times (rings 1–6) were significantly delayed and P1 response densities in rings 5 and 6 were significantly lower in participants with NPDR compared to those with No DR. In those with No DR, P1 implicit times in almost all rings were delayed in relation to lower vessel density and perfusion (maximum variance noted was 13%). In individuals with NPDR, the P1 response density in rings 2 and 3 showed a positive nonsignificant correlation with macular perfusion.

Conclusion:

In those with diabetes with No DR, retinal neuronal function is influenced by lower macular vessel density and perfusion. The retinal neuronal function is abnormal in individuals with NPDR compared to those with No DR and is not correlated with OCT angiometric measures, suggesting the likelihood of a different retinal structural correlate.

Corneal nerve loss is related to the severity of painful diabetic neuropathy

Background and purpose

Previously it has been shown that patients with painful diabetic neuropathy (PDN) have greater corneal nerve loss compared to patients with painless diabetic neuropathy. This study investigated if the severity of corneal nerve loss was related to the severity of PDN.

Methods

Participants with diabetic neuropathy (n = 118) and healthy controls (n = 38) underwent clinical and neurological evaluation, quantitative sensory testing, nerve conduction testing and corneal confocal microscopy and were categorized into those with no (n = 43), mild (n = 34) and moderate‐to‐severe (n = 41) neuropathic pain.

Results

Corneal nerve fibre density (p = 0.003), corneal nerve fibre length (p < 0.0001) and cold perception threshold (p < 0.0001) were lower and warm perception threshold was higher (p = 0.002) in patients with more severe pain, but there was no significant difference in the neuropathy disability score (p = 0.5), vibration perception threshold (p = 0.5), sural nerve conduction velocity (p = 0.3) and amplitude (p = 0.7), corneal nerve branch density (p = 0.06) and deep breathing heart rate variability (p = 0.08) between patients with differing severity of PDN. The visual analogue scale correlated significantly with corneal nerve fibre density (r = −0.3, p = 0.0002), corneal nerve branch density (r = −0.3, p = 0.001) and corneal nerve fibre length (r = −0.4, p < 0.0001). Receiver operating curve analysis showed that corneal nerve fibre density had an area under the curve of 0.78 with a sensitivity of 0.73 and specificity of 0.72 for the diagnosis of PDN.

Conclusions

Corneal confocal microscopy reveals increasing corneal nerve fibre loss with increasing severity of neuropathic pain and a good diagnostic outcome for identifying patients with PDN.

A machine learning model for early detection of diabetic foot using thermogram images

Diabetes foot ulceration (DFU) and amputation are a cause of significant morbidity. The prevention of DFU may be achieved by the identification of patients at risk of DFU and the institution of preventative measures through education and offloading. Several studies have reported that thermogram images may help to detect an increase in plantar temperature prior to DFU. However, the distribution of plantar temperature may be heterogeneous, making it difficult to quantify and utilize to predict outcomes. We have compared a machine learning-based scoring technique with feature selection and optimization techniques and learning classifiers to several state-of-the-art Convolutional Neural Networks (CNNs) on foot thermogram images and propose a robust solution to identify the diabetic foot. A comparatively shallow CNN model, MobilenetV2 achieved an F1 score of ∼95% for a two-feet thermogram image-based classification and the AdaBoost Classifier used 10 features and achieved an F1 score of 97%. A comparison of the inference time for the best-performing networks confirmed that the proposed algorithm can be deployed as a smartphone application to allow the user to monitor the progression of the DFU in a home setting.

Corneal Nerve and Brain Imaging in Mild Cognitive Impairment and Dementia

Background:

Visual rating of medial temporal lobe atrophy (MTA) is an accepted structural neuroimaging marker of Alzheimer’s disease. Corneal confocal microscopy (CCM) is a non-invasive ophthalmic technique that detects neuronal loss in peripheral and central neurodegenerative disorders.

Objective:

To determine the diagnostic accuracy of CCM for mild cognitive impairment (MCI) and dementia compared to medial temporal lobe atrophy (MTA) rating on MRI.

Methods:

Subjects aged 60–85 with no cognitive impairment (NCI), MCI, and dementia based on the ICD-10 criteria were recruited. Subjects underwent cognitive screening, CCM, and MTA rating on MRI.

Results:

182 subjects with NCI (n = 36), MCI (n = 80), and dementia (n = 66), including AD (n = 19, 28.8%), VaD (n = 13, 19.7%), and mixed AD (n = 34, 51.5%) were studied. CCM showed a progressive reduction in corneal nerve fiber density (CNFD, fibers/mm²) (32.0±7.5 versus 24.5±9.6 and 20.8±9.3, p < 0.0001), branch density (CNBD, branches/mm²) (90.9±46.5 versus 59.3±35.7 and 53.9±38.7, p < 0.0001), and fiber length (CNFL, mm/mm²) (22.9±6.1 versus 17.2±6.5 and 15.8±7.4, p < 0.0001) in subjects with MCI and dementia compared to NCI. The area under the ROC curve (95% CI) for the diagnostic accuracy of CNFD, CNBD, CNFL compared to MTA-right and MTA-left for MCI was 78% (67–90%), 82% (72–92%), 86% (77–95%) versus 53% (36–69%) and 40% (25–55%), respectively, and for dementia it was 85% (76–94%), 84% (75–93%), 85% (76–94%) versus 86% (76–96%) and 82% (72–92%), respectively.

Conclusion:

The diagnostic accuracy of CCM, a non-invasive ophthalmic biomarker of neurodegeneration, was high and comparable with MTA rating for dementia but was superior to MTA rating for MCI.

Corneal Confocal Microscopy Predicts the Development of Diabetic Neuropathy: A Longitudinal Diagnostic Multinational Consortium Study

OBJECTIVE

Corneal nerve fiber length (CNFL) has been shown in research studies to identify diabetic peripheral neuropathy (DPN). In this longitudinal diagnostic study, we assessed the ability of CNFL to predict the development of DPN.

RESEARCH DESIGN AND METHODS

From a multinational cohort of 998 participants with type 1 and type 2 diabetes, we studied the subset of 261 participants who were free of DPN at baseline and completed at least 4 years of follow-up for incident DPN. The predictive validity of CNFL for the development of DPN was determined using time-dependent receiver operating characteristic (ROC) curves.

RESULTS

A total of 203 participants had type 1 and 58 had type 2 diabetes. Mean follow-up time was 5.8 years (interquartile range 4.2-7.0). New-onset DPN occurred in 60 participants (23%; 4.29 events per 100 person-years). Participants who developed DPN were older and had a higher prevalence of type 2 diabetes, higher BMI, and longer duration of diabetes. The baseline electrophysiology and corneal confocal microscopy parameters were in the normal range but were all significantly lower in participants who developed DPN. The time-dependent area under the ROC curve for CNFL ranged between 0.61 and 0.69 for years 1-5 and was 0.80 at year 6. The optimal diagnostic threshold for a baseline CNFL of 14.1 mm/mm² was associated with 67% sensitivity, 71% specificity, and a hazard ratio of 2.95 (95% CI 1.70-5.11; P < 0.001) for new-onset DPN.

CONCLUSIONS

CNFL showed good predictive validity for identifying patients at higher risk of developing DPN ∼6 years in the future.

Painful diabetic neuropathy is associated with increased nerve regeneration in patients with type 2 diabetes undergoing intensive glycemic control

AIMS/INTRODUCTION

Painful diabetic peripheral neuropathy (pDPN) is associated with small nerve fiber degeneration and regeneration. This study investigated whether the presence of pDPN might influence nerve regeneration in patients with type 2 diabetes undergoing intensive glycemic control.

MATERIALS AND METHODS

This exploratory substudy of an open-label randomized controlled trial undertook the Douleur Neuropathique en 4 questionnaire and assessment of electrochemical skin conductance, vibration perception threshold and corneal nerve morphology using corneal confocal microscopy in participants with and without pDPN treated with exenatide and pioglitazone or basal-bolus insulin at baseline and 1-year follow up, and 18 controls at baseline only.

RESULTS

Participants with type 2 diabetes, with (n = 13) and without (n = 28) pDPN had comparable corneal nerve fiber measures, electrochemical skin conductance and vibration perception threshold at baseline, and pDPN was not associated with the severity of DPN. There was a significant glycated hemoglobin reduction (P < 0.0001) and weight gain (P < 0.005), irrespective of therapy. Participants with pDPN showed a significant increase in corneal nerve fiber density (P < 0.05), length (P < 0.0001) and branch density (P < 0.005), and a decrease in the Douleur Neuropathique en 4 score (P < 0.01), but no change in electrochemical skin conductance or vibration perception threshold. Participants without pDPN showed a significant increase in corneal nerve branch density (P < 0.01) and no change in any other neuropathy measures. A change in the severity of painful symptoms was not associated with corneal nerve regeneration and medication for pain.

CONCLUSIONS

This study showed that intensive glycemic control is associated with greater corneal nerve regeneration and an improvement in the severity of pain in patients with painful diabetic neuropathy.

Optimal glycaemic and blood pressure but not lipid targets are related to a lower prevalence of diabetic microvascular complications

BACKGROUND

Diabetic microvascular complications are a major cause of morbidity and are related to glycaemic control and cardiovascular risk factors.

AIMS

We sought to determine the association of microvascular complications in relation to control of glycemia, blood pressure and lipids in T2DM patients attending secondary care in Qatar.

METHODS

This is a cross-sectional study undertaken in patients with T2DM attending Qatar's National Diabetes Centres. Patients underwent assessment of glycemia, blood pressure and lipids and prevalence of diabetic peripheral neuropathy (DPN), retinopathy and microalbuminuria.

RESULTS

We included 1114 subjects aged 52.1 ± 11.3 years with a duration of diabetes 10.0 ± 7.6 years and had a prevalence of 25.8% for DPN, 34.3% for painful DPN, 36.8% for microalbuminuria and 25.1% for retinopathy. Patients who achieved an HbA1c ≤ 7.0% compared to >7% had a significantly lower prevalence of DPN (P < 0.01), painful DPN (P < 0.01), retinopathy (P < 0.01) and microalbuminuria (P < 0.007). Patients who achieved a systolic BP ≤ 140 mmHg compared to >140 mmHg had a significantly lower prevalence of DPN (P < 0.001), painful DPN (P < 0.001), retinopathy (P < 0.001) and microalbuminuria (P < 0.001). Patients who achieved an LDL ≤2.6 mmol/l compared to >2.6 mmol/l had a significantly higher prevalence of DPN (P < 0.03), but no difference in other outcomes. There was no difference in microvascular complications between those who achieved a HDL-C ≥ 1.02 mmol/l, and among those who achieved triglycerides ≤1.7 mmol/l.

CONCLUSIONS

Optimal control of glycemia and blood pressure, but not lipids is associated with a lower prevalence of diabetic microvascular complications.

Spinal Inhibitory Dysfunction in Patients With Painful or Painless Diabetic Neuropathy

OBJECTIVE

Impaired rate-dependent depression of the Hoffman reflex (HRDD) is a marker of spinal inhibitory dysfunction and has previously been associated with painful neuropathy in a proof-of-concept study in patients with type 1 diabetes. We have now undertaken an assessment of HRDD in patients with type 1 or type 2 diabetes.

RESEARCH DESIGN AND METHODS

A total of 148 participants, including 34 healthy control subjects, 42 patients with painful diabetic neuropathy, and 62 patients with diabetic neuropathy without pain, underwent an assessment of HRDD and a detailed assessment of peripheral neuropathy, including nerve conduction studies, corneal confocal microscopy, and thermal threshold testing.

RESULTS

Compared with healthy control subjects (P < 0.001) and patients without pain (P < 0.001), we found that HRDD is impaired in patients with type 1 or type 2 diabetes with neuropathic pain. These impairments are unrelated to diabetes type and the presence or severity of neuropathy. In contrast, patients without neuropathic pain (P < 0.05) exhibited enhanced HRDD compared with control subjects.

CONCLUSIONS

We suggest that loss or impairment of HRDD may help to identify a subpopulation of patients with painful diabetic neuropathy mediated by impaired spinal inhibitory systems who may respond optimally to therapies that target spinal or supraspinal mechanisms. Enhanced RDD in patients without pain may reflect engagement of spinal pain-suppressing mechanisms.