基于深度学习的领域几何优化,用于用体积调制弧线疗法传递的全髓辐射
Nicola Lambri1,2, Giorgio Longari3, Daniele Loiacono3
1Department of Biomedical Sciences, Humanitas University, Pieve Emanuele, Milan, Italy.
Medical physics
|April 18, 2024
概括
卷积神经网络 (CNN) 自动生成总骨髓辐射 (TMI/TMLI) 场地几何,证明在临床上可接受全原血造干细胞移植. 这种人工智能驱动的方法与专家医学物理学家的表现相匹配,提高了治疗计划的效率.
科学领域:
- 医学物理 医学物理
- 辐射疗法 辐射疗法
- 人工智能在医学中的应用
背景情况:
- 总骨髓 (淋巴细胞) 照射 (TMI/TMLI) 对于全源造血干细胞移植调节方案至关重要.
- 使用体积调制弧疗法 (VMAT) 为TMI/TMLI生成复杂的场地几何是临床上具有挑战性和耗时的.
- 专业的医学物理学家 (MP) 目前正在执行TMI/TMLI领域几何创建.
研究的目的:
- 开发和验证卷积神经网络 (CNN) 用于自动生成TMI/TMLI字段几何.
- 为了减少与手动TMI/TMLI领域几何规划相关的临床负担.
主要方法:
- 用TMI/TMLI治疗的117名患者的数据集用于训练两种基于CT,PTV和器官面罩的角平面投影的CNN模型.
- CNN被训练来预测同中心坐标和孔,并通过局部优化提炼输出.
- 模型的性能使用根平均平方误差 (RMSE) 和三名经验丰富的国会议员盲目定性评估进行了评估.
主要成果:
- 优化的CNN实现了13±3mm (CNN-1) 和18±4mm (CNN-2) 的RMSE值.
- 与手动配置相比,生成的场地几何没有显著差异,质量的中位数为4 (充足).
- 自2023年10月以来,人工智能生成的场地几何已经集成到临床实践中.
结论:
- 基于CNN的TMI/TMLI自动场地几何生成在临床上是可以接受的,甚至对于经验丰富的国会议员来说也是足够的.
- 将MP专业知识集成到CNN开发周期中,对于模型优化至关重要,特别是在有限的数据的情况下.
- 这种人工智能方法简化了TMI/TMLI规划,支持高效准确的放射治疗.
相关概念视频
01:26Optimization Problems
220
Optimization problems often involve identifying maximum or minimum values under specific constraints. A well-known example is determining the longest horizontal pipe that can be moved around a right-angled corner, where a 3-meter-wide hallway meets a 2-meter-wide hallway. This scenario, common in architectural design and industrial transport, can be understood conceptually through geometric and trigonometric reasoning.To visualize the problem, consider the pipe as a straight line that touches...
220
01:26Area Problem
312
Determining the area of a region with straight edges is straightforward, as geometric formulas for rectangles, triangles, and polygons can be applied directly. However, traditional geometric methods are insufficient when a region has a curved boundary, such as the area under a function.fromThe area problem involves finding a systematic way to measure such regions. One approach to solving this problem is through approximation. Instead of attempting to compute the area exactly at the outset, the...
312
01:28Methods of Medium Optimization
70
Optimizing growth media enhances microbial proliferation and maximizes product yield. Statistical experimental design methodologies provide structured and reproducible approaches, offering progressively higher levels of robustness and efficiency.The One-Factor-at-a-Time (OFAT) MethodThe One-Factor-at-a-Time (OFAT) method involves adjusting a single variable while keeping all others constant. However, it cannot detect interactions between variables, often leading to suboptimal outcomes when...
70