基于光学触觉传感器的肿块检测研究

Research on Lump Detection Based on Optical Tactile Sensor

肿块检测可以帮助患者更早地发现潜在的疾病,提高患者的生存率。针对如何使用非侵入式方法检测组织样本中的肿块这一问题,制作了一款结构简单、校准方便的光学触觉传感器,并提出了一种基于表面重建的肿块判别算法。该方法对触觉传感器采集到的原始图像进行深度重建,并从重建图像中减去正常组织产生的基线图像,从而获得由肿块产生的异常触觉图像,并将处理完的图像送入神经网络中进行肿块检测。结果表明,该方法能够有效地分辨出直径在6~10 mm、嵌入深度在1~5 mm的肿块,验证了所述方法的实用性与可行性。最后,基于该方法与流程,讨论了肿块大小、肿块嵌入深度和正常组织硬度对肿块检测结果的影响。提出的非侵入式的肿块检测方法探索了光学触觉传感器在肿块检测领域中的可行性,为相关领域进一步的研究提供了新的思路。

Lump detection can help patients find potential diseases early and improve their survival rate.Aimming at the problem of how to use a non-invasive method to detect the lump in the tissue samples,an optical tactile sensor with simple fabrication and convenient calibration is developed,and a new algorithm for lumps identification based on surface reconstruction is proposed.In this method,the original images collected by the tactile sensor are reconstructed in depth,and the baseline images generated by normal tissue are subtracted from the reconstructed image,so as to obtain the abnormal tactile images generated by the lump,and the processed images are sent to the neural network for lump detection.The result shows that the method can effectively distinguish lumps with a diameter of 6~10 mm and an embedding depth of 1~5 mm,which verifies the practicability and feasibility of the method.Finally,based on the method and procedure,the effects of lump size,lump embedding depth and normal tissue hardness on lump detection results are discussed.The proposed non-invasive lump detection method explores the feasibility of optical tactile sensors in the field of lump detection,providing new ideas for further research in related fields.