人工智能在儿童脊柱侧弯诊断中的研究进展

Research advances in artificial intelligence assisted diagnosis of scoliosis in children

摘要:

儿童脊柱侧弯是常见的三维脊柱畸形，早期准确诊断和干预可有效阻止侧弯进展。近年来，人工智能在图像识别任务上取得了显著进展，其通过数据收集、特征提取和模型构建等步骤处理医学图像，已被初步应用于脊柱侧弯的智能化诊断，并有望改善手工测量依赖医生经验、耗时费力、可重复性差的缺点。本文综述了人工智能在儿童脊柱侧弯诊断上的研究进展，为今后开发全自动、智能化脊柱侧弯诊疗工具提供研究思路和参考。

Abstract:

Scoliosis is a common three-dimensional spinal deformity in children. Early and accurate diagnosis and intervention can ef-fectively prevent the progression of scoliosis. In recent years, artificial intelligence has made remarkable progress in image recognition tasks,processing medical images through steps such as data collection, feature extraction and model construction, and has been initially applied tothe intelligent diagnosis of scoliosis, and is expected to overcome the shortcomings of manual measurement relying on doctor experience,time-consuming and laborious, and poor repeatability. This paper reviews the research progress of artificial intelligence in the diagnosis ofscoliosis in children, and provides research ideas and references for the development of automatic and intelligent scoliosis diagnosis andtreatment tools in the future.