Intravenous(IV) access is animportant daily clinical procedure that delivers fluids or medication into a patient’s vein. However, IV insertion is very challenging where clinicians aresuffering in locating the subcutaneous vein due to patients’physiological factorssuch as hairy forearm and thick dermis fat, and alsomedical staff’s level of fatigue. To resolve this issue, researchers have proposed autonomous machines to be usedfor IV access,but such equipmentarelackingcapabilityin detecting the vein accurately. Therefore, this project proposes an automatic veindetection algorithm using deep learning for IVaccess purpose. U-Net, a fully connected network (FCN) architecture is employed in thisproject due to its capability in detecting the near-infrared (NIR)subcutaneous vein. Data augmentation is applied toincrease the dataset size and reduce the bias from overfitting. The original U-Net architecture is optimized by replacing up-sampling with transpose convolution as well as theadditional implementation ofbatch normalizationbesides reducing the number oflayers to diminish the risk of overfitting.After fine-tuning and retraining the hypermodel, an unsupervised dataset is used to evaluate the hypermodel by selecting 10 checkpoints foreach forearm image and comparing the checkpoints onpredicted outputs to determine true positive vein pixels. The proposed lightweight U-Net has achieved slightly lower accuracy (0.8871) than the original U-Net architecture. Even so, the sensitivity, specificity, and precisionare greatly improved byachieving 0.7806, 0.9935,and 0.9918 respectively. This result indicates that the proposed algorithm can be applied into the venipuncture machine toaccurately locate the subcutaneous vein for intravenous (IV)procedures.
Read full abstract