跟骨是人體足中最大骨，最重要的骨骼。它的作用是保持人體重量和其施加的力量。然而，這種骨骼也是最常受傷的骨。在大多數情況下，發生損傷由於過度的軸向載荷。 跟骨骨折分類用於將受損的跟骨骨分類為每種類型需要不同類型的治療。在幾種不同的分類系統中，Sanders分類是最常用的，因為它基於跟骨骨折的冠狀計算機斷層掃描（CT），顯示跟骨最寬的下表面。然而，儘管其受歡迎程度，但其解釋存在高度的可變性和不一致性，其用戶之間僅具有公平到中等的一致性，這主要是由於經驗的差異造成的。這種不一致是重要的，因為治療中的錯誤可能對患者造成不可挽回的傷害。 目前，醫生手動檢查所選擇的CT掃描圖像，其中包含最完整的跟骨結構，然後根據跟骨骨折分類對其進行分類。這容易出錯，不僅是因為上述問題，還因為有時CT掃描圖像模糊並且沒有顯示清晰的斷裂線。 本論文試圖通過檢測跟骨碎片並顯示骨折線來幫助識別跟骨骨折類型。第一部分是將每個骨碎片與圖像中的其他骨碎片區分開來。第二部分是指出跟骨碎片並顯示其邊緣。 ;Calcaneus bone is the biggest and the most important bone in human foot. Its role is to hold human weight and other forces that they applied to it. However, this bone is also the most frequently injured tarsal bone. In most cases, the injuries happen because of excessive axial load such as a fall from height or in a motor vehicle accident. Calcaneus bone fracture classification is used to classify the injured calcaneus bone into types with each type needing a different kind of treatment. Among the several different classification systems, Sanders classification is the most commonly used because it is based on the coronal computed tomography (CT) scan of the calcaneal fracture that shows the widest undersurface of the posterior facet of the calcaneus. However, in spite of its popularity, there is a high degree of variability and inconsistency in its interpretation with only fair to moderate consistency among its users, which is mostly caused by the difference in experience. Such inconsistency is significant because a mistake in treatment may cause irreparable harm to the patient. Currently, doctors manually examine the chosen CT scan image, which contains the most complete calcaneus structure, then classify it according to calcaneus fracture classification. This is prone to mistakes, not only because of the aforementioned problem but also because sometimes the CT scan image is blurred and does not show clear fracture line. This thesis attempts to help in identifying the calcaneus fracture type by detecting calcaneus bone fragments and showing the fracture lines. The first part is to differentiate each bone fragment from the other bone fragments in the image. The second part is to point out the calcaneus bone fragments and show their edges. This system successfully automatically determine which fragments belong to calcaneus bone.
