AI in Atomic Force Microscopy: Advancing Biological Nanoscale Imaging and Autonomous Discovery

Abstract

Atomic force microscopy (AFM) enables label-free nanoscale imaging and nanomechanical profiling but remains constrained by low throughput, operator dependence, and variability in data interpretation. Artificial intelligence (AI) transforms AFM into a scalable and adaptive platform. Initially applied in materials science for super-resolution imaging, tip deconvolution, segmentation, and force-curve analysis, AI approaches are now being extended to biological AFM. These methods support robust denoising of soft matter maps, automated recognition of heterogeneous structures, and three-dimensional reconstruction of biomolecular assemblies. This review provides an end-to-end workflow of AI-enabled AFM─from probe optimization and adaptive control to multimodal data integration─highlighting advances relevant to mechanobiology and biomedical engineering. By surveying studies with amyloid fibrils, extracellular vesicles, membranes, and living cells, we show how AI-AFM convergence enhances reproducibility, throughput, and clinical utility. AI-driven AFM is poised to enable disease modeling, therapeutic screening, and precision diagnostics, establishing itself as a next-generation tool for biomedical discovery.