AI-Assisted Fusion of Scanning Electrochemical Microscopy Images Using Novel Soft Probe

Abstract

Scanning electrochemical microscopy (SECM) is one of the scanning probe techniques that has attracted considerable attention because of its ability to interrogate surface morphology or electrochemical reactivity. However, the quality of SECM images generally depends on the sizes of the electrodes and many uncontrollable factors. Furthermore, manipulating fragile glass ultramicroelectrodes and blurred images sometimes frustrate researchers. To overcome the challenges of modern SECM, we developed novel soft gold probes and then established the AI-assisted methodology for image fusion. A novel gold microelectrode probe with high softness was developed to scan fragile samples. The distribution of EGFR (protein biomarker) in oral cancer was investigated. Then, we fused the optical microscopic images and SECM images to enhance the image quality using Matlab software. However, by changing the parameters for image fusion, thousands of fused images were generated, which is annoying for researchers. Thus, a deep learning model was built to select the best-fused images according to the contrast and clarity of the fused images. Therefore, the quality of the SECM images was improved by using a novel soft probe followed by combining the image fusion technique. In the future, a new scanning probe with AI-assisted fused SECM image processing may be interpreted more preciously and contribute to the early detection of cancers.