Naifold capillaroscopy in mixed connective tissue disease patients.

INTRODUCTION: Mixed connective tissue disease (MCTD) is a rare systemic disease characterized by overlapping features of systemic lupus erythematosus (SLE), systemic sclerosis (SSc), dermato-/polymyositis (DM/PM), and rheumatoid arthritis (RA). Naifold capillaroscopy (NFC) is a non-invasive test for evaluating the capillaries of the nail shaft used in the diagnosis of rheumatic diseases. OBJECTIVES: To determine whether there are characteristic abnormalities in NFC in MCTD patients, and whether the type of NFC lesions correlates with organ involvement in these patients. METHODS: Clinical picture and NFC patterns were analyzed in 43 patients with MCTD. Capillaroscopic images were divided into scleroderma-like pattern (SD-like pattern) according to the Cutolo classification, non-specific lesions, and normal images. Relationships between the clinical aspects considered in the MCTD classification criteria and the changes in the capillaroscopic images were evaluated. RESULTS: SD-like pattern was present in 20 MCTD patients (46.51%) with a predominance of the "early" pattern. Giant, branched, dilated capillaries and reduced capillary density were found more frequently in MCTD patients compared to the control group (p-values 0.0005, 0.005, 0.02, < 0.0001 respectively). There were associations found between the presence of a reduced number of vessels, avascular areas, and SD-like pattern with the presence of sclerodactyly in MCTD patients (p = 0.002, p = 0.006, p = 0.02, respectively), alongside an association between the presence of branched vessels and the subpapillary plexus with pulmonary arterial hypertension (PAH) (p = 0.04 and p = 0.005, respectively). CONCLUSIONS: MCTD patients are significantly more likely to have abnormalities upon NFC. It is worthwhile to perform capillaroscopic examination in MCTD patients. Key Points • Scleroderma-like pattern was found in more than half of the MCTD patients. • Reduced capillary density was found to be a significant predictor of the diagnosis of MCTD. • There were relationships between the presence of reduced capillary density, avascular areas, and SD-like with the presence of sclerodactyly in the MCTD patients. • There was an association between the presence of branched vessels and the visibility of the subpapillary plexus and pulmonary arterial hypertension (PAH).

Microvascular status in juvenile Sjögren's disease: the first nailfold videocapillaroscopy investigation.

INTRODUCTION: Juvenile Sjögren's disease (jSjD) is a rare autoimmune disease characterized by exocrine gland involvement and systemic manifestations, including small vessel vasculitis and Raynaud's phenomenon (RP). We aimed to investigate the microvascular status in jSjD patients by nailfold videocapillaroscopy (NVC) and the potential correlations with clinical and serological features. METHODS: Clinical data from thirteen consecutive jSjD patients (11 females and 2 males), with a mean age of 16 ± 4 years, diagnosed before 16 years of age (mean age at diagnosis 12 ± 3) according to the 2016 American College of Rheumatology/EULAR criteria for adult SjD, were collected including age- and sex-matched healthy controls (HCs). Clinical, laboratory, and instrumental data were collected, together with NVC examination. Non-specific and specific NVC parameters were investigated, such as capillary density, capillary dilations, giant capillaries, microhaemorrhages and abnormal shapes. Associations between NVC findings and clinical/serological features were explored and analysed using parametrical and non-parametrical tests. RESULTS: Capillary density reduction correlated significantly with articular involvement (arthralgias) (p = 0.024). Microhaemorrhages correlated with lower C3 levels (p = 0.034). No specific NVC pattern for jSjD was identified, whereas abnormal capillary shapes were significantly higher in jSjD patients than HCs (p = 0.005). NVC abnormalities were not associated with SjD-specific instrumental tests (biopsy, imaging, Schirmer's test). RP was present in 8% of jSjD patients. CONCLUSIONS: The reduction of capillary density, as well as microhaemorrhages at NVC analysis, are significantly associated with some clinical aspects like articular involvement and serum biomarkers (C3 reduction). The NVC is suggested as safe and further analysis in jSjD patients.

Analysis of microvascular pattern in diabetes mellitus condition using the nailfold capillaroscopy images.

Diabetes is often considered a vascular disease due to its impact on blood vessels, it is a complex condition with various metabolic and autoimmune factors involved. One of the long term comorbidities of diabetes includes microvascular complications. The microvascular complications can be analyzed using the Nailfold capillaroscopy, a non-invasive technique that allows for the visualization and analysis of capillaries in the proximal nailfold area. Using advanced video capillaroscopy with high magnification, capillary images can be captured from and processed to analyze their morphology. The capillary images of normal group and diabetic group are acquired from 118 participants using nailfold capillaroscopy and the obtained images are preprocessed using image processing filters. The identification and segmentation of the capillaries are the challenges to be addressed in the processing of the images. Hence segmentation of capillaries is done using morphological operations, thresholding and convolutional neural networks. The performance of the filters and segmentation methods are evaluated using Mean Square Error (MSE), Peak signal to Noise Ratio (PSNR), Structural Similarity Index Measure (SSIM), Jaccard Index and Sorensen coefficient. By analyzing the morphological features namely the capillary diameter, density, distribution, presence of hemorrhage and the shape of the capillaries from both the groups, the capillary changes associated with diabetic condition were studied. It was found that the non diabetic participants considered in this study has capillary diameter in the range of 8-14 µm and the capillary density in the range of 10-30 capillaries per mm2 whereas the diabetic participants has capillary diameter greater than 30 µm and the capillary density is less than 10 capillaries per mm2. In addition to capillary density and diameter, the presence of hemorrhage, the orientation and distribution of the capillaries are also considered to differentiate the diabetic group from the non diabetic group. The classification of the participants are validated with the clinical history of the participants.

Microvascular changes on nailfold capillaroscopy in acute stage of Kawasaki disease: a new diagnostic paradigm for an enigmatic condition.

OBJECTIVES: Kawasaki disease (KD) is a medium vessel vasculitis with a predilection to involve coronary arteries. However, there is a paucity of literature on microvascular changes in patients with KD. METHODS: Children diagnosed with KD based on American Heart Association guidelines 2017 were enrolled prospectively. Demographic details and echocardiographic changes in coronaries were recorded. Nailfold capillaries were assessed using Optilia Video capillaroscopy and data were analysed using Optilia Optiflix Capillaroscopy software at acute (prior to IVIG administration) and subacute/convalescent phase. RESULTS: We enrolled 32 children with KD (17 boys) with a median age of 3 years. Nailfold capillaroscopy (NFC) was performed in 32 patients in the acute phase (compared with 32 controls) and in 17 during the subacute/convalescent phase at a median follow-up of 15 (15-90) days after IVIG treatment. The following findings were seen in NFC in the acute phase of KD: reduced capillary density (n = 12, 38.6%), dilated capillaries (n = 3, 9.3%), ramifications (n = 3, 9.3%) and capillary haemorrhages (n = 2, 6.2%). Capillary density was reduced significantly in the acute phase of KD (38.6%) as compared with the subacute/convalescent phase (25.4%) (P-value <0.001) and controls (0%) (P-value = 0.03). We observed no correlation between coronary artery involvement and mean capillary density (P = 0.870). CONCLUSION: Results show that patients with KD have significant nailfold capillary changes in the acute phase. These findings may provide a new diagnostic paradigm for KD and a window to predict coronary artery abnormalities.

Non-invasive detection of early microvascular changes in juveniles with type 1 diabetes.

AIMS/HYPOTHESIS: The study aimed to assess the usefulness of capillaroscopy and photoplethysmography in the search for early vascular anomalies in children with type 1 diabetes. METHODS: One hundred sixty children and adolescents aged 6-18, 125 patients with type 1 diabetes, and 35 healthy volunteers were enrolled in the study. We performed a detailed clinical evaluation, anthropometric measurements, nailfold capillaroscopy, and photoplethysmography. RESULTS: Patients with diabetes had more often abnormal morphology in capillaroscopy (68.60%, p = 0.019), enlarged capillaries (32.6%, p = 0.006), and more often more over five meandering capillaries (20.90%, p = 0.026) compared to healthy controls. Meandering capillaries correlated with higher parameters of nutritional status. In a photoplethysmography, patients with diagnosed neuropathy had a higher percentage of flow disturbance curves (p < 0.001) with a reduced frequency of normal curves (p = 0.050). CONCLUSIONS: Capillaroscopic and photoplethysmographic examinations are non-invasive, painless, fast, and inexpensive. They are devoid of side effects, and there are no limitations in the frequency of their use and repetition. The usefulness of capillaroscopy and photoplethysmography in the study of microcirculation in diabetic patients indicates the vast application possibilities of these methods in clinical practice.

Automated Stabilization, Enhancement and Capillaries Segmentation in Videocapillaroscopy.

Oral capillaroscopy is a critical and non-invasive technique used to evaluate microcirculation. Its ability to observe small vessels in vivo has generated significant interest in the field. Capillaroscopy serves as an essential tool for diagnosing and prognosing various pathologies, with anatomic-pathological lesions playing a crucial role in their progression. Despite its importance, the utilization of videocapillaroscopy in the oral cavity encounters limitations due to the acquisition setup, encompassing spatial and temporal resolutions of the video camera, objective magnification, and physical probe dimensions. Moreover, the operator's influence during the acquisition process, particularly how the probe is maneuvered, further affects its effectiveness. This study aims to address these challenges and improve data reliability by developing a computerized support system for microcirculation analysis. The designed system performs stabilization, enhancement and automatic segmentation of capillaries in oral mucosal video sequences. The stabilization phase was performed by means of a method based on the coupling of seed points in a classification process. The enhancement process implemented was based on the temporal analysis of the capillaroscopic frames. Finally, an automatic segmentation phase of the capillaries was implemented with the additional objective of quantitatively assessing the signal improvement achieved through the developed techniques. Specifically, transfer learning of the renowned U-net deep network was implemented for this purpose. The proposed method underwent testing on a database with ground truth obtained from expert manual segmentation. The obtained results demonstrate an achieved Jaccard index of 90.1% and an accuracy of 96.2%, highlighting the effectiveness of the developed techniques in oral capillaroscopy. In conclusion, these promising outcomes encourage the utilization of this method to assist in the diagnosis and monitoring of conditions that impact microcirculation, such as rheumatologic or cardiovascular disorders.

Automated nailfold capillary density measurement method based on improved YOLOv5.

Nailfold capillary density is an essential physiological parameter for analyzing nailfold health; however, clinical images of the nailfold are taken in many situations, and most clinicians subjectively analyze nailfold images. Therefore, based on the improved "you only look once v5" (YOLOv5) algorithm, this study proposes an automated method for measuring nailfold capillary density. The improved technique can effectively and rapidly detect distal capillaries by incorporating methods or structures such as 9mosaic, spatial pyramid pooling cross-stage partial construction, bilinear interpolation, and efficient intersection over union. First, the modified YOLOv5 algorithm was used to detect nailfold capillaries. Subsequently, the number of distal capillaries was filtered using the 90° method. Finally, the capillary density was calculated. The results showed that the Average Precision (AP)@0.5 value of the proposed approach reached 85.2 %, which was an improvement of 4.93 %, 5.24 %, and 107 % compared with the original YOLOv5, YOLOv6, and simple-faster rapid-region convolutional network (R-CNN), respectively. For different nailfold images, using the density calculated by nailfold experts as a benchmark, the calculated results of the proposed method were consistent with the manually calculated results and superior to those of the original YOLOv5.

Raynaud phenomenon and microvasculopathy in systemic sclerosis: multi-modality imaging for diagnosis and evaluation.

PURPOSE OF REVIEW: To describe the clinical significance of and the diagnostic approach to Raynaud phenomenon (RP) in the peripheral extremities and the heart. RECENT FINDINGS: Nailfold capillaroscopy has recently been standardized in an expert consensus paper. Abnormal capillaroscopy in combination with specific autoantibody profiles and clinical signs are highly predictive of progression of RP to systemic sclerosis (SSc). Magnetic resonance imaging (MRI) can also perform tissue characterization of both the extremities and the heart. Microvascular wall abnormalities detected using nailfold capillaroscopy in patients with SSc may lead to deposition of erythrocyte-derived iron, due to microhemorrhages, which may predispose to fibrosis. MRI can assess the presence of iron using T2∗ measurements. SUMMARY: RP is a hallmark of the microvasculopathy in SSc and can affect both the peripheral extremities and the heart. Nailfold capillaroscopy is the current gold standard for the evaluation of the peripheral microvasculature. Other imaging modalities include thermography, laser Doppler-derived methods, 99m Tc-pertechnetate hand perfusion scintigraphy, power Doppler ultrasonography, dynamic optical coherence tomography, MRI, and photoacoustic imaging, but these are currently not widely used. Cardiac RP can be investigated with positron emission tomography or cardiovascular magnetic resonance, with the latter offering the additional possibility of tissue characterization and iron content quantification secondary to microhemorrhages.

Multivariate Parametric Study of Nailfold Capillary Images for Disease Detection.

Nailfold capillaroscopy is a tool which is non-invasive in nature and can be useful for diagnosis, research, therapeutic study and prognosis. Research shows that specific capillary morphology patterns are identified for diabetic subjects, hypertensive subjects and normal controls. In this study, we have proposed RATHEW approach of classifying these three classes of subjects. RATHEW approach employs a three step process for classifying nailfold images: one, identify six abnormality parameters from the image dataset; two, score these abnormality parameters based on the defined scoring rules; and three, combine them mathematically to segregate them into three classes. This technique can be further enhanced to grade the severity of disease and organ involvement. This can bring in a paradigm shift to the disease detection and therapeutic study mechanism.

Microvascular damage in autoimmune connective tissue diseases: a capillaroscopic analysis from 20 years of experience in a EULAR training and research referral centre for imaging.

OBJECTIVE: Nailfold videocapillaroscopy (NVC) allows the detection of microvascular damage in autoimmune connective tissue diseases (CTDs). The prevalence of the morphological capillary findings was retrospectively evaluated in a wide cohort of patients with Raynaud's phenomenon secondary to a CTD at the time of the first single NVC, independently from their current treatment, autoantibody profile and comorbidities. METHODS: One-thousand-one-hundred-eighty-one patients affected by CTDs were included from 2001 to 2021. The considered CTDs were systemic sclerosis (SSc), undifferentiated connective tissue disease (UCTD), mixed connective tissue disease (MCTD), dermatomyositis (DM), systemic lupus erythematosus, Sjögren's syndrome and primary antiphospholipid syndrome (aPS). The capillaroscopic parameters were distinguished between scleroderma patterns and non-scleroderma patterns. RESULTS: Giant capillaries were significantly more frequent in SSc, DM and MCTD than in other CTDs (respectively, in 73%, 73% and 61% of patients, p<0.001 when comparing each rate vs the other CTDs). The mean capillary count was significantly lower in SSc, DM and MCTD (respectively, 7.04±0.18 vs 6.5±0.75 vs 7.7±2 capillaries/linear mm) compared with the other CTDs (p<0.001 for each rate vs the other CTDs). The non-specific abnormalities of capillary morphology were significantly more frequent in SSc, MCTD and aPS (respectively, in 48%, 41% and 36% of cases, all p<0.001 vs each other CTDs). CONCLUSION: This large size sample of patients with CTDs, collected over 20 years of analysis, confirms the highest prevalence of specific capillaroscopic alterations in patients with SSc, DM and MCTD, when compared with other CTDs.

Nailfold capillaroscopy.

Nailfold capillaroscopy is a safe and well-established method for the assessment of structural alterations of the microcirculation. It is a crucial tool in the investigation and monitoring of patients presenting with Raynaud's phenomenon. Detection of the characteristic "scleroderma pattern" on capillaroscopy may indicate an underlying rheumatic disease, particularly systemic sclerosis (SSc). Herein, we highlight the practical aspects of videocapillaroscopy, including image acquisition and analysis, with mention of dermoscopy. Special emphasis is placed on standardized use of terminology to describe capillary characteristics. Systematic evaluation of images in discerning the normal from the abnormal using the validated European Alliance of Associations for Rheumatology (EULAR) Study Group consensus reporting framework is paramount. In addition to the relevance of capillaroscopy in the (very) early diagnosis of SSc, its emerging predictive value (especially capillary loss) for new organ involvement and disease progression is underscored. We further provide capillaroscopic findings in selected other rheumatic diseases.

Peripheral nailfold capillary microscopic abnormalities in rheumatoid arthritis are associated with arterial stiffness: Results from a cross-sectional study.

Vascular injury eventually resulting in the establishment of cardiovascular disease is a serious complication in rheumatoid arthritis (RA). Nailfold videocapillaroscopy (NVC) is a non-invasive imaging modality that enables the quantitative and qualitative assessment of the peripheral microvasculature. Nevertheless, capillaroscopic patterns remain inadequately defined in RA, especially regarding their clinical significance as potential markers of systemic vascular impairment. Consecutive RA patients underwent NVC using a standardized protocol, to assess the following parameters: capillary density, avascular areas, capillary dimensions, microhemorrhages, subpapillary venous plexus, and presence of ramified, bushy, crossed and tortuous capillaries. Carotid-femoral pulse wave velocity (PWV) and pulse pressure were measured as well-acknowledged markers of large artery stiffening. The vast majority of our cohort (n = 44) presented a combination of non-specific and abnormal capillaroscopic parameters. Capillary ramification was associated with both PWV and pulse pressure, even after adjustment for cardiovascular risk factors and systemic inflammation. Our study highlights the high prevalence of a wide range of capillaroscopic deviations from the normal patterns in RA. Furthermore, it provides for the first time evidence of an association between structural disorders of the microcirculation and markers of macrovascular dysfunction, suggesting that NVC might have a role as an index of generalised vascular impairment in RA.

Nailfold capillaroscopy reveals early peripheral microcirculation abnormalities in children affected by heterozygous familial hypercholesterolemia.

BACKGROUND: nailfold capillaroscopy (NCF) is a non-invasive imaging technique to seek peripheral microcirculation abnormalities in children and adults. Familial hypercholesterolemia is a genetic disorder caused by mutations capable of increasing blood levels of low-density lipoproteins cholesterol (LDL-C), thus triggering early atherosclerosis. The study aims at evaluating peripheral microcirculation in children with heterozygous familial hypercholesterolemia (HeFH) by means of NFC in comparison with healthy peers and at searching for possible correlations between these abnormalities and patients' lipid panel. METHODS: thirty-six HeFH patients were enrolled (13 males and 23 females. Mean age 8 ± 3 years; age range 3-13 years). They had increased levels of total cholesterol (237.9 ± 34.2 mg/dl) and LDL-C (154.2 ± 37.6 mg/dl). Both values were ≥95th gender and age specific centile. All the subjects in the study underwent NFC. RESULTS: In 69.4 % of HeFH children nailfold capillaries were tortuous (p < 0.00001 compared to healthy controls). In 41.6 % the number of capillaries was markedly reduced (<7 capillaries/mm). The mean number of capillaries was 8.4 ± 2.6/mm in HeFH and 12.2 ± 1.4/mm in healthy controls (p < 0.00001). In 100 % of the sample size capillary blood flow was slowed down (p < 0.00001). In 50 % of the sample size a blood "sludge" phenomenon was seen (p < 0.00001). No gender differences were detected. Sludge phenomenon was seen only in those with LDL-C over 99th centile (p < 0.00001). CONCLUSION: NCF allows the identification of an early peripheral microvascular dysfunction in HeFH children which is similar to that already seen in atherosclerotic disease. Prompt identification of these capillary abnormalities may be crucial in implementing early prevention measures.

The role of nailfold capillaroscopy in pediatric patients with Kawasaki disease.

BACKGROUND: Nailfold capillaroscopy has been used as a non-invasive diagnostic method for microvasculature evaluation in various rheumatological disorders. The present study aimed to determine the utility of nailfold capillaroscopy in the diagnosis of Kawasaki Disease (KD). METHOD: In this case-control study nailfold capillaroscopy was performed in 31 patients with KD and 30 healthy controls. All nailfold images were evaluated for capillary distribution and capillary morphology such as enlargement, tortuosity, and dilatation of the capillaries. RESULT: Abnormal capillaroscopic diameter was identified in 21 patients from the KD group and 4 patients in the control group. The most common abnormality in capillary diameter was irregular dilatation in 11 (35.4%) KD patients and in 4 people (13.3%) in the control group. Distortions of the normal capillary architecture was commonly seen in the KD group (n=8). A positive correlation was observed between coronary involvement and abnormal capillaroscopic results (r=.65, P<.03). The sensitivity and specificity of capillaroscopy for the diagnosis of KD were 84.0% (95%CI: 63.9-95.5%) and 72.2% (95%CI: 54.8-85.8%), respectively. The PPV and NPV of capillaroscopy for KD were 67.7% (95%CI: 48.6-83.3) and 86.7% (95% CI: 69.3-96.2), respectively. CONCLUSION: Capillary alterations are more common in KD patients compared to control group. Thus, nailfold capillaroscopy can be useful in detecting these alterations. Capillaroscopy is a sensitive test for detecting capillary alternations in KD patients. It could be used as a feasible diagnostic modality for evaluating microvascular damage in KD.

External clinical validation of automated software to identify structural abnormalities and microhaemorrhages in nailfold videocapillaroscopy images.

OBJECTIVES: Automated systems to analyse nailfold videocapillaroscopy (NVC) images are needed to promptly and comprehensively characterise patients with systemic sclerosis (SSc) or Raynaud's phenomenon (RP). We previously developed, and validated in-house, a deep convolutional neural network-based algorithm to classify NVC-captured images according to the presence/absence of structural abnormalities and/or microhaemorrhages. We present its external clinical validation. METHODS: A total of 1,164 NVC images of RP patients were annotated by 5 trained capillaroscopists according to the following categories: normal capillary; dilation; giant capillary; abnormal shape; tortuosity; microhaemorrhage. The images were also presented to the algorithm. Matches and discrepancies between algorithm predictions and those annotations obtained by consensus of ≥3 or ≥4 interobservers were analysed. RESULTS: Consensus among ≥3 capillaroscopists was achieved in 86.9% of images, 75.8% of which were correctly predicted by the algorithm. Consensus among ≥4 experts occurred in 52.0% of cases, in which 87.1% of the algorithm's results matched with those of the expert panel. The algorithm's positive predictive value was >80% for microhaemorrhages and unaltered, giant or abnormal capillaries. Sensitivity was >75% for dilations and tortuosities. Negative predictive value and specificity were >89% for all categories. CONCLUSIONS: This external clinical validation suggests that this algorithm is useful to assist in the diagnosis and follow-up of SSc or RP patients in a timely manner. It may also be helpful in the management of patients with any pathology presenting with microvascular changes, as the algorithm has been designed to also be useful for research aiming at extending the usage of nailfold capillaroscopy to more conditions.

[Nailfold capillaroscopy-Principles and clinical application]. / Kapillarmikroskopie ­ Grundlagen und klinische Anwendung.

Nailfold capillaroscopy is a rapid and easily applicable differential diagnostic technique that allows direct visualization of the microcirculation. Abnormal findings in nailfold capillaroscopy are closely associated with connective tissue diseases, such as systemic sclerosis. The clinical manifestation of impaired microcirculation is Raynaud's phenomenon, which is a classical symptom of connective tissue diseases. Nailfold capillaroscopy is increasingly used in various fields of medicine, therefore it is important to define methods for the acquisition and analysis of the results of nailfold capillary and to have a uniform definition of abnormal capillaries. This article discusses image acquisition and analysis, various capillaroscopic techniques, normal and abnormal capillaroscopic features and their significance, scoring systems and reliability of image acquisition and interpretation.

Nailfold capillaroscopy and deep learning in diabetes.

OBJECTIVE: To determine whether nailfold capillary images, acquired using video capillaroscopy, can provide diagnostic information about diabetes and its complications. RESEARCH DESIGN AND METHODS: Nailfold video capillaroscopy was performed in 120 adult patients with and without type 1 or type 2 diabetes, and with and without cardiovascular disease. Nailfold images were analyzed using convolutional neural networks, a deep learning technique. Cross-validation was used to develop and test the ability of models to predict five5 prespecified states (diabetes, high glycosylated hemoglobin, cardiovascular event, retinopathy, albuminuria, and hypertension). The performance of each model for a particular state was assessed by estimating areas under the receiver operating characteristics curves (AUROC) and precision recall curves (AUPR). RESULTS: A total of 5236 nailfold images were acquired from 120 participants (mean 44 images per participant) and were all available for analysis. Models were able to accurately identify the presence of diabetes, with AUROC 0.84 (95% confidence interval [CI] 0.76, 0.91) and AUPR 0.84 (95% CI 0.78, 0.93), respectively. Models were also able to predict a history of cardiovascular events in patients with diabetes, with AUROC 0.65 (95% CI 0.51, 0.78) and AUPR 0.72 (95% CI 0.62, 0.88) respectively. CONCLUSIONS: This proof-of-concept study demonstrates the potential of machine learning for identifying people with microvascular capillary changes from diabetes based on nailfold images, and for possibly identifying those most likely to have diabetes-related complications.

A deep learning system for quantitative assessment of microvascular abnormalities in nailfold capillary images.

OBJECTIVES: Nailfold capillaroscopy is key to timely diagnosis of SSc, but is often not used in rheumatology clinics because the images are difficult to interpret. We aimed to develop and validate a fully automated image analysis system to fill this gap. METHODS: We mimicked the image interpretation strategies of SSc experts, using deep learning networks to detect each capillary in the distal row of vessels and make morphological measurements. We combined measurements from multiple fingers to give a subject-level probability of SSc.We trained the system using high-resolution images from 111 subjects (group A) and tested on images from subjects not in the training set: 132 imaged at high-resolution (group B); 66 imaged with a low-cost digital microscope (group C). Roughly half of each group had confirmed SSc, and half were healthy controls or had primary RP ('normal'). We also estimated the performance of SSc experts. RESULTS: We compared automated SSc probabilities with the known clinical status of patients (SSc versus 'normal'), generating receiver operating characteristic curves (ROCs). For group B, the area under the ROC (AUC) was 97% (94-99%) [median (90% CI)], with equal sensitivity/specificity 91% (86-95%). For group C, the AUC was 95% (88-99%), with equal sensitivity/specificity 89% (82-95%). SSc expert consensus achieved sensitivity 82% and specificity 73%. CONCLUSION: Fully automated analysis using deep learning can achieve diagnostic performance at least as good as SSc experts, and is sufficiently robust to work with low-cost digital microscope images.