Plantar fascia segmentation and thickness estimation in ultrasound images

Abstract

Ultrasound (US) imaging offers significant potential in diagnosis of plantar fascia (PF) injury and monitoring treatment. In particular US imaging has been shown to be reliable in foot and ankle assessment and offers a real-time effective imaging technique that is able to reliably confirm structural changes, such as thickening, and identify changes in the internal echo structure associated with diseased or damaged tissue. Despite the advantages of US imaging, images are difficult to interpret during medical assessment. This is partly due to the size and position of the PF in relation to the adjacent tissues. It is therefore a requirement to devise a system that allows better and easier interpretation of PF ultrasound images during diagnosis. This study proposes an automatic segmentation approach which for the first time extracts ultrasound data to estimate size across three sections of the PF (rearfoot, midfoot and forefoot). This segmentation method uses artificial neural network module (ANN) in order to classify small overlapping patches as belonging or not-belonging to the region of interest (ROI) of the PF tissue. Features ranking and selection techniques were performed as a post-processing step for features extraction to reduce the dimension and number of the extracted features. The trained ANN classifies the image overlapping patches into PF and non-PF tissue, and then it is used to segment the desired PF region. The PF thickness was calculated using two different methods: distance transformation and area-length calculation algorithms. This new approach is capable of accurately segmenting the PF region, differentiating it from surrounding tissues and estimating its thickness.

Anatomical variations of the plantar fascia's origin with respect to age and sex—an MRI based study

“…The PF plays a significant role in both static and dynamic function of the foot by supporting the arch (Boussouar et al, ; Guo et al, ; Orner et al, ; Park et al, ). Originating from CB in close proximity to the AT and expanding toward the metatarsal bones where it inserts, the PF may also highly influence the biomechanical properties of the surrounding structures (Aydogan et al, ; Angin et al, ).…”

“…). By supporting the arch of the foot, this connective tissue plays a significant role in terms of foot biomechanical properties and its mobility (Boussouar et al, ; Guo et al, ; Park et al, ). While standing, the arch of the foot together with PF may highly influence the foot's ability to carry load from the body weight (Cheung et al, ; Aydogan et al, ; Angin et al, ; Guo et al, ; Park et al, ).…”

“…Pathologies involving PF such as flat foot deformity in children, acquired flat foot deformity in adults, and plantar fasciitis are common and affect both athletes and the general population (Boussouar et al, ; Draghi et al, ; Bowser et al, ; de Cesar Netto et al, ; Granado et al, ; Orner et al, ; Ramskov et al, ; Kerr et al, ). These pathologies may influence not only the functional properties of the foot itself, but also the overall biomechanical properties and mobility of the lower limbs (Silvester, ; Park et al, ).…”

Plantar Fascia Ultrasound Images Characterization and Classification Using Support Vector Machine

“…Automated segmentation is one of the most important tasks in medical image processing and analysis, including, pattern recognition, supervised or unsupervised subjects classification and novelty detection; it is mainly used to locate the desired region of interest objects in the input images dataset. As reported in [3], an automated ANNs supervised segmentation approach was introduced in this study to segment different PF regions. The proposed segmentation approach uses the radial basic function neural network (RBF-NN) classifier [9] to automatically segment the PF region and estimate its thickness.…”

“…The proposed PF classification model consists of the following modules as shown in Fig 2. (i) preprocessing phase employing speckle noise reduction filtering and image enhancement operations to reduce the effects of undesirable speckle noise phenomenon and improve the contrast of the PF US images using dual tree complex wavelet transform with soft thresholding (DT-CWT_S) and contrast-limited adaptive histogram equalization filter (CLAHE), respectively; (ii) artificial neural networks supervised segmentation phase applying different features measures, a features ranking module and trained radial basic function neural network (RBF-NN) classifier as discussed in [3], to automatically segment the PF region and calculate its thickness; (iii) texture features extraction and analysis introducing 6 sets of feature extraction measures (for extracting a total of 40 features), features ranking and selection operation using an unsupervised infinity feature selection method [17] to select and analyse the most discriminating and suitable features for the classification process; (iv) the classifier module using Linear-SVM and Kernel-SVM to distinguish between asymptomatic and symptomatic plantar fascia subjects; and (v) classification performance analysis using different performance measures such as recall, specificity, balanced accuracy, precision, F-score and MCC.…”