本計畫目的是想以深度學習的方式讓三維生物列印系統擁有支架製作品質判斷的能力。生物支架材料大致可分為天然材料跟人工合成兩大類。目前最為廣泛使用的天然材料有膠原蛋白、海藻酸鈉、明膠及殼聚醣等。由於上述材料係從動、植物取得，每批量配置出的生物墨水總有些許差距。生物列印之製造參數大多彼此相依，因此很難藉由理論推導出有效的預測公式。往往只能使用試誤法，導致需耗費大量時間在調整製造參數來達到預設的尺寸目標。卷積神經網路為深度學習的一支，其可透過大量過往實驗數據(轉換成二維影像)訓練出一個有效預測模型，在新的數據輸入此一模型後進而可預測結果，而這個能力可幫助改進生物支架的品質。本計畫擬以殼聚醣、明膠及海藻酸鈉為支架材料，細胞使用纖維母細胞。三年計畫規劃如下：第一年重點在建構深度學習模型雛形，並進行不含細胞溫感3D列印以獲得訓練資料。第二年著重在建立可商業部署之深度學習軟體，並進行不含細胞光感3D列印以獲得訓練資料。最後一年則開始進行大量的含細胞3D列印(溫感、光感及溫、光感混合交聯)，用來獲取訓練資料及驗證結果。在此階段，細胞存活率亦將做為支架品質評估標準。期望透過此三年計畫，將來製作出的含細胞支架不僅尺寸符合需求，細胞存活、增殖、型態等細胞屬性亦能符合需求。 ;The purpose of this project is to enable the 3D bioprinting system to have the ability to judge the qualities of bio-scaffolds in a deep learning manner. The materials of bio-scaffolds can roughly divide into two categories: natural and synthetic materials. The most widely used natural materials are collagen, sodium alginate, gelatin, and chitosan. As these natural materials obtained from animals and plants, there is always a slight difference between the bio-inks in each batch. Most of the fabrication parameters have dependence with each other, and it is difficult to derive effective predictive formulas for manufacturing scaffolds. Often can only use the trial and error method, resulting in the need to spend a lot of time to adjust the fabrication parameters for achieving the preset target. The convolutional neural network is a kind of Deep learning, and it can train a model through a large amount of historical data (converted into 2D images), and predictable results obtain after inputting new data into the model. The capability can help to improve the quality of the bio-scaffold.The project plans to use chitosan, gelatin, and sodium alginate as scaffold materials, and the cells use fibroblasts. The three-year plan is as follows: In the first year, the focus is on constructing the prototype of the deep learning model and performing thermo-sensitive 3D printing to obtain the training data. The second year focused on the establishment of commercial available deep learning software and UV- sensitive 3D printing to obtain the training data. In the last year, a large number of cell-laden 3Dscaffolds (thermo-sensitive, UV-sensitive, and hybrid) will print to obtain the training data and verify the result. In this stage, cell viability will add as the scaffold quality evaluation standard. Through this three year project, this project would develop an intelligent bio-printing system capable of producing high quality scaffolds.
