The purpose of this study was to investigate the association between the mechanical properties of plantar soft tissue and diabetes status. 51 (M/F: 21/30) participants with prediabetes onset (fasting blood sugar [FBS] level > 100 mg/dL), age >18 years, and no lower limb amputation were recruited after ethical approval was granted from Pontificia Universidad Catolica del Peru ethical review board. Ultrasound reverberant shear wave elastography was used to assess the soft tissue stiffness at the 1st metatarsal head (MTH), 3rd MTH, and the heel at both feet. Spearman’s rank-order correlation (rho) test indicated a significant (P < .05) positive correlations between FBS level and the plantar soft tissue shear wave speed at the 1st MTH: rho = 0.402 (@400 Hz), rho = 0.373 (@450 Hz), rho = 0.474 (@500 Hz), rho= 0.395 (@550 Hz), and rho = 0.326 (@600 Hz) in the left foot and rho = 0.364 (@450 Hz) in the right foot. Mann-Whitney U test indicated a significantly (P < .05) higher shear wave speed in the plantar soft tissue with the following effect sizes (r) at the 1st MTH of the left foot at all tested frequencies: r = 0.297 (@450 Hz), r = 0.345 (@500 Hz), r = 0.322 (@550 Hz), and r = 0.275 (@600 Hz), and at the 1st MTH of right foot r = 0.286 (@400 Hz) in diabetes as compared with the age and body mass index matched prediabetes group. An association between fasting blood sugar level and the stiffness of the plantar soft tissue with higher values of shear wave speed in diabetes versus prediabetes group was observed. This indicated that the proposed approach can improve the assessment of the severity of diabetic foot complications with potential implications in patient stratification.
Wound area segmentation really progressed with the emergence of deep learning, due to its robustness in uncontrolled lighting and no need to design hand-crafted features but two limits have still to be overcome: firstly, its performance relies on the size and quality of the training dataset in the medical field, where data annotation is costly and time-consuming; secondly the accuracy of the segmentation depends highly on the camera distance and angle and moreover perspective effects prevent measuring real surfaces in single views. To address concurrently these two issues, we propose to apply multi-view modeling: an image sequence is acquired around the wound site and enables wound 3D reconstruction. Then, a segmentation step is run to extract roughly the wound from the background in each view and to select the best view with an original strategy. This view provides the most accurate segmentation and the real wound bed area even on non planar wounds. Finally, this segmentation is backprojected in each view to generate a complete set of well annotated real images to reinforce the learning step of the neural network. In our experiments, we compare several strategies to select the best view in the image sequence. The proposed method, tested on a dataset of 270 images, outperforms standard deep learning approach based on a single view, as recorded with DICE index and IoU score which rise respectively from 36.53% to 86.3% and 29.48% to 77.09% for the wound class to achieve an overall DICE and IoU score of 93.04% and 86.61% including background class. These results attest to the robustness of our method and its improved accuracy in the wound segmentation task.
Along with geometric and color indicators, thermography is another valuable source of information for wound monitoring. The interaction of geometry with thermography can provide predictive indicators of wound evolution; however, existing processes are focused on the use of high-cost devices with a static configuration, which restricts the scanning of large surfaces. In this study, we propose the use of commercial devices, such as mobile devices and portable thermography, to integrate information from different wavelengths onto the surface of a 3D model. A handheld acquisition is proposed in which color images are used to create a 3D model by using Structure from Motion (SfM), and thermography is incorporated into the 3D surface through a pose estimation refinement based on optimizing the temperature correlation between multiple views. Thermal and color 3D models were successfully created for six patients with multiple views from a low-cost commercial device. The results show the successful application of the proposed methodology where thermal mapping on 3D models is not limited in the scanning area and can provide consistent information between multiple thermal camera views. Further work will focus on studying the quantitative metrics obtained by the multi-view 3D models created with the proposed methodology.
Fusion des données thermiques et tridimensionnelles pour le suivi des plaies chroniques: Les plaies chroniques constituent un problème de santé grave car elles ne suivent pas un processus de guérison normal et prennent beaucoup de temps à guérir. Le suivi de l'évolution de la plaie est vital pour évaluer le traitement actuel et déterminer la nécessité d'une approche différente. Traditionnellement, cette évaluation est basée sur des mesures manuelles et une observation visuelle, qui peuvent être subjectives et non quantitatives, entraînant une incertitude dans le suivi du patient. Cette thèse propose d'exploiter l'imagerie couleur et thermique des caméras embarquées dans un smartphone pour fournir un outil non invasif d'évaluation des plaies. Des appareils standard et peu coûteux garantissent son applicabilité et son évolutivité en milieu clinique, avec une acquisition rapide et simple. Une méthodologie a été proposée pour créer des modèles thermiques 3D réalistes et précis à partir d'un ensemble limité d'images capturées avec des appareils portables et en utilisant la prise de vue à main levée.En outre, la méthodologie proposée a été testée dans plusieurs cas de blessures, et une preuve de concept des mesures obtenues avec le modèle thermique 3D a été démontrée pour l'évaluation des blessures dans une étude clinique dans un centre médical à Lima, au Pérou.