Abstract
In recent years, a type of biochip known as a Digital Microfluidic Biochip (DMFB) has been actively researched in the field of life sciences. DMFBs perform dilution operations by mixing reagent solutions and buffer solutions at a 1:1 ratio to generate droplets with the desired concentration. One of the challenges of DMFBs is that droplets may not always be evenly split during the droplet division process. To address this issue, an error correction method utilizing error cancellation has been proposed. This method modifies the dilution graph to minimize the impact of division errors on the target node. However, this approach has a significant drawback: when large division errors occur in nodes close to the target node, they can introduce substantial concentration errors at the target node. In this paper, we propose a method that duplicates nodes near the target node and performs re-dilution to correct errors. Furthermore, we present an efficient and accurate error correction approach by modifying the dilution graph so that the output nodes of the dilution operation are at equal levels relative to the target node. Through simulations conducted 10,000 times, we demonstrate that our method effectively reduces the average concentration error at the target node.