近日，美国约翰霍普金斯大学教授Denis Wirtz及其研究团队提出了InterpolAI：基于深度学习的光流插值和生物医学图像恢复，以改进3D组织映射。该项研究成果发表在2025年5月28日出版的《自然—方法学》上。

研究小组介绍了InterpolAI，这是一种利用大图像运动的帧插值（基于光流的人工智能（AI）模型），在图像堆栈中对真实图像之间插值合成图像的方法。InterpolAI优于线性插值和最先进的基于光流的方法XVFI，保留微观解剖特征和细胞计数，以及图像对比度，方差和亮度。InterpolAI可修复组织损伤并减少缝合伪影。研究组通过多种成像方式、物种、染色技术和像素分辨率验证了InterpolAI。这项工作证明了人工智能在提高图像数据集的分辨率、吞吐量和质量方面的潜力，从而改善了3D成像。

据了解，成像和计算的最新进展使大型三维（3D）生物数据集的分析成为可能，揭示空间组成、形态、细胞相互作用和罕见事件。然而，这些分析的准确性受到图像质量的限制，图像质量可能会受到数据缺失、组织损伤或由于机械、时间或财务限制而导致的低分辨率的影响。

附：英文原文

Title: InterpolAI: deep learning-based optical flow interpolation and restoration of biomedical images for improved 3D tissue mapping

Author: Joshi, Saurabh, Forjaz, Andr, Han, Kyu Sang, Shen, Yu, Queiroga, Vasco, Selaru, Florin A., Grard, Marie, Xenes, Daniel, Matelsky, Jordan, Wester, Brock, Barrutia, Arrate Muoz, Kiemen, Ashley L., Wu, Pei-Hsun, Wirtz, Denis

Issue&Volume: 2025-05-28

Abstract: Recent advances in imaging and computation have enabled analysis of large three-dimensional (3D) biological datasets, revealing spatial composition, morphology, cellular interactions and rare events. However, the accuracy of these analyses is limited by image quality, which can be compromised by missing data, tissue damage or low resolution due to mechanical, temporal or financial constraints. Here, we introduce InterpolAI, a method for interpolation of synthetic images between pairs of authentic images in a stack of images, by leveraging frame interpolation for large image motion, an optical flow-based artificial intelligence (AI) model. InterpolAI outperforms both linear interpolation and state-of-the-art optical flow-based method XVFI, preserving microanatomical features and cell counts, and image contrast, variance and luminance. InterpolAI repairs tissue damages and reduces stitching artifacts. We validated InterpolAI across multiple imaging modalities, species, staining techniques and pixel resolutions. This work demonstrates the potential of AI in improving the resolution, throughput and quality of image datasets to enable improved 3D imaging.

DOI: 10.1038/s41592-025-02712-4

Source: https://www.nature.com/articles/s41592-025-02712-4