Correlation of nailfold capillaroscopy findings with history of digital ulcer on same finger: Results of SCLEROCAP study

Digital ulcers in systemic sclerosis are strongly associated with digital arterial disease: a finger-by-finger analysis of finger brachial pressure index measurements

OBJECTIVES

The digital ulcers (DU) of SSc are disabling and frequent. Their pathogenesis involves a capillary microangiopathy and a digital arterial disease that few studies were able to quantify up to now. A multicentre observational study about the predictive value of capillaroscopy in SSc offered us the opportunity to evaluate further the complementary information provided by both capillary and arterial evaluations.

METHODS

During the SCLEROCAP study, five out of the nine centres performed a systematic evaluation of the finger brachial pressure index (FBPI) in the last four fingers of both hands at baseline, using the same laser-Doppler device. In the present work, FBPI measurements were compared between fingers with vs without DU or scars, before and after adjusting for the capillaroscopic pattern and systemic factors.

RESULTS

FBPI measurements were performed in 2537 fingers from 326 patients. Active ulcers or scars were found in 10.8% of those fingers, more often on the right hand, and in the second and third fingers. FBPI was lower than 0.70 in 26% of all fingers and in 57.5% of those with ulcers. A strong association was found between a low FBPI and the presence of DU, even after adjusting for capillaroscopic pattern, ulcer location and the patient himself.

CONCLUSION

These results confirm the importance of digital arterial disease in the pathogenesis of DU of SSc, which is independent from the microangiopathy. FBPI measurements complement the information provided by capillaroscopy and might have an important predictive value for subsequent DU.

Analysis of nailfold capillaroscopy images with artificial intelligence: Data from literature and performance of machine learning and deep learning from images acquired in the SCLEROCAP study

OBJECTIVE

To evaluate the performance of machine learning and then deep learning to detect a systemic scleroderma (SSc) landscape from the same set of nailfold capillaroscopy (NC) images from the French prospective multicenter observational study SCLEROCAP.

METHODS

NC images from the first 100 SCLEROCAP patients were analyzed to assess the performance of machine learning and then deep learning in identifying the SSc landscape, the NC images having previously been independently and consensually labeled by expert clinicians. Images were divided into a training set (70 %) and a validation set (30 %). After features extraction from the NC images, we tested six classifiers (random forests (RF), support vector machine (SVM), logistic regression (LR), light gradient boosting (LGB), extreme gradient boosting (XGB), K-nearest neighbors (KNN)) on the training set with five different combinations of the images. The performance of each classifier was evaluated by the F1 score. In the deep learning section, we tested three pre-trained models from the TIMM library (ResNet-18, DenseNet-121 and VGG-16) on raw NC images after applying image augmentation methods.

RESULTS

With machine learning, performance ranged from 0.60 to 0.73 for each variable, with Hu and Haralick moments being the most discriminating. Performance was highest with the RF, LGB and XGB models (F1 scores: 0.75-0.79). The highest score was obtained by combining all variables and using the LGB model (F1 score: 0.79 ± 0.05, p < 0.01). With deep learning, performance reached a minimum accuracy of 0.87. The best results were obtained with the DenseNet-121 model (accuracy 0.94 ± 0.02, F1 score 0.94 ± 0.02, AUC 0.95 ± 0.03) as compared to ResNet-18 (accuracy 0.87 ± 0.04, F1 score 0.85 ± 0.03, AUC 0.87 ± 0.04) and VGG-16 (accuracy 0.90 ± 0.03, F1 score 0.91 ± 0.02, AUC 0.91 ± 0.04).

CONCLUSION

By using machine learning and then deep learning on the same set of labeled NC images from the SCLEROCAP study, the highest performances to detect SSc landscape were obtained with deep learning and in particular DenseNet-121. This pre-trained model could therefore be used to automatically interpret NC images in case of suspected SSc. This result nevertheless needs to be confirmed on a larger number of NC images.

Association between oxygen delivery and digital ulcers in systemic sclerosis

OBJECTIVE

Tissue hypoxia due to microvasculopathy is the main cause of digital ulcers (DUs) in systemic sclerosis (SSc). Reduced oxygen delivery (DO2) to the tissues may also contribute to the development of DU. This study was conducted to investigate the association between DO2 and DUs in patients with SSc.

METHODS

In all, 111 patients and 30 healthy controls were enrolled. DO2 was calculated by using the formula; DO2 = Cardiac output × arterial oxygen saturation (SpO2) × serum haemoglobin level × 1.39 × 10. Both right index finger SpO2 measurements (index-SpO2) and highest value of SpO2 (maximum SpO2) obtained among the fingers of the subjects were used for the calculations and DO2 results were adjusted both for weight and body surface area (BSA).

RESULTS

Mean DO2 was lower in SSc patients as compared to controls in all 4 different calculations but the difference was only statistically significant when using index-SpO2 and adjusting for BSA (498 mL/min/m² vs 549 mL/min/m², p = 0.03). There was a strong positive correlation between cardiac output and DO2 calculated by using the index-SpO2 (r = 0.903; p < 0.001). Of the SSc patients, 46 (41.4 %) had DUs within the last 12 months. Patients with DUs had higher mean mRSS, lover mean FVC and more frequently diffuse disease, interstitial lung disease, anti-SCL70 antibody positivity (p < 0.05 for all). No difference was observed in DO2 among DU positive or DU negative groups by any calculation (p > 0.05 for all).

CONCLUSIONS

DO2 in SSc patients seems to be lower than healthy controls. However, DO2 is similar between the patients with and without DUs. Our results suggest that the contribution of DO2 is negligible to the development of DU and support the major role of microvasculopathy in SSc patients with DUs.