人工智能技术在口腔颌面部肿瘤诊断中应用的研究进展

Research progress on the application of artificial intelligence technology in the diagnosis of oral and maxillofacial tumors

随着计算机运算能力的提升,以及医疗数据的积累,人工智能技术(artificial intelligence,AI)在口腔颌面部肿瘤诊断工作中逐步得到应用,可以辅助医生的诊断工作,提升诊疗效率和诊断准确性。近年来的研究更多集中在医疗图像的识别上,常用的方法是由专家对大量图像进行标注,供机器学习图像特征,从而实现智能化的诊断。现有研究已经能通过AI技术对影像学图像、病理切片和肿瘤外观照片等进行智能化分析,实现对肿瘤的诊断。现阶段的研究存在的主要问题是标注数据质量参差不齐、样本量过小、研究问题局限和数据模态单一等。这些问题需要通过算法的不断完善以及大量优质数据的积累逐步解决。未来AI技术发展的方向应该是综合多种来源医疗数据,辅助医生进行诊断,探索利用各种无创、易行的新方法早期发现肿瘤,彻底改变现有诊疗模式。

With the improvement of computer computing capability and the accumulation of a large amount of medical data, artificial intelligence is gradually being applied in the diagnosis of oral and maxillofacial tumors. Artificial intelligence technology can assist doctors in clinical diagnosis and improve the efficiency of clinical work and the accuracy of diagnosis. In recent years, researchers have focused primarily on the recognition of medical images. The commonly used method is to annotate a large number of images by experts for learning image features by machines. The available literature has been able to utilize artificial intelligence technology to diagnose tumors by analyzing medical images, pathological sections, and tumor photos. The main issues in the current research are uneven labeling data quality, small data size, limited research problems, and single data modalities. These problems need to be solved through the continuous improvement of algorithms and the accumulation of high-quality data. The future direction of artificial intelligence applications should be to integrate medical data from multiple sources, assist doctors in diagnosis, and explore a variety of noninvasive and easy-to-use new methods for the early diagnosis of tumors. This may completely change the existing diagnosis and treatment model of oral and maxillofacial tumors.