机器学习策略识别形态学衰老亚型 ：研究人员采用基于 Xception 卷积神经网络架构的机器学习策略，对细胞的原始纹理图像进行分析，计算单个细胞的衰老得分。通过训练，该算法在测试和验证数据集上的平均准确率达到 89%，受试者工作特征曲线下面积（AUC）为 0.95。根据衰老得分，研究人员确定了 C7、C10 和 C11 为衰老亚型，这些亚型具有较高的衰老得分和独特的形态学特征，而 C6 和 C8 ...

Cardiac fibroblasts (CF) can become activated and produce excessive extracellular matrix, contributing to fibrotic remodeling ...

Not all zombie skin cells are the same. New research reveals three unique subtypes with different roles in aging and disease, ...

which influenced fibroblast distribution and organization. Cells near the coverslip-scaffold interface displayed a spread 2D morphology. Those deeper within the scaffold assumed a compact shape ...