Abstract | ||
---|---|---|
Surgical instrument sorting, sterilization, and inspection in the hospital are highly time-consuming and labor-intensive due to the sheer volume and variety of tools. It is not easy to find the automation replacement since the diversity and similarity of these instruments bring numerous challenges to their identification, and hence the problem remains largely unresolved. In this paper, we design a system incorporating supervised deep learning networks and conventional methods to realize surgical instruments’ fast and robust identification to facilitate robot picking. Our approach overcomes the difficulties of manually determining the region of interest (ROI) for surgical instruments. We fasten this process through a proposed labeling strategy and hence avoid the high manual labor and time costs. Two types of surgical instruments datasets are created through the design for the first time and are openly available. Robot experiments are performed to demonstrate the effectiveness of our strategy in facilitating surgical instrument automation. |
Year | DOI | Venue |
---|---|---|
2022 | 10.1109/AIM52237.2022.9863381 | 2022 IEEE/ASME International Conference on Advanced Intelligent Mechatronics (AIM) |
Keywords | DocType | ISSN |
robot picking,surgical instrument sorting,sheer volume,automation replacement,deep learning networks,robot experiments,surgical instrument automation,surgical instrument identification system,region of interest,ROI | Conference | 2159-6247 |
ISBN | Citations | PageRank |
978-1-6654-1309-1 | 0 | 0.34 |
References | Authors | |
11 | 6 |
Name | Order | Citations | PageRank |
---|---|---|---|
Rongzihan Song | 1 | 0 | 0.68 |
Pius Zhen Ye Lim | 2 | 0 | 0.34 |
Albert Causo | 3 | 0 | 0.34 |
Gnanaprakasam Naveen | 4 | 0 | 0.34 |
Zhiping Lin | 5 | 839 | 83.62 |
I-Ming Chen | 6 | 567 | 87.28 |