基于深度学习的白血病语义细分影响了白细胞的白血病细分
Zahoor Jan1, Muhammad Shabir1, Haleem Farman2
1Department of Computer Science, Islamia College University, Peshawar, Pakistan.
PloS one
|May 8, 2025
概括
这项研究引入了一种新的方法,用于在医学图像中对白细胞 (WBC) 进行细分,使用UNet++,标记分流和神经普通微分方程 (ODE). 该方法实现了高精度,改善了用于疾病诊断的自动化血细胞分析.
科学领域:
- 医疗图像分析 医学图像分析
- 计算病理学计算病理学
- 生物医学工程 生物医学工程
背景情况:
- 精确的白细胞细分对于诊断各种疾病至关重要.
- 在WBC细分的挑战包括重叠的细胞,大小/形状的变化,和不成熟的细胞边界.
- 现有的方法往往难以应对血涂片图像的复杂性.
研究的目的:
- 开发和评估一种新的,强大的方法,从血液涂抹图像对白细胞进行细分.
- 为了提高自动化血细胞分析的准确性和可靠性.
- 整合先进的深度学习和图像处理技术,以改善诊断成像.
主要方法:
- 一种混合方法,将UNet++用于预分割,标记分水算法用于分离,以及神经普通微分方程 (ODE) 用于细分细分.
- UNet++生成概率灰度图像,然后使用标记分水来解决重叠的细胞.
- ODE 在卷积后应用,以尽量减少在训练和推理过程中的错误传播.
主要成果:
- 提出的方法实现了高细分精度,平均交叉与结合 (贾卡德指数) 为97.73%,子相似系数为98.36%,平均像素精度为98.97%.
- 与现有系统相比,UNet++,标记分水和ODE的独特组合表现出卓越的性能.
- 该方法有效地根据结构变异将WBC与红细胞 (RBC) 和血小板区分开来.
结论:
- 集成的UNet++,标记分水和ODE方法在自动化WBC细分方面取得了重大进展.
- 这种技术有望增强自动化血细胞分析,诊断成像和疾病监测的临床应用.
- 未来的工作重点应该是优化该模型,以便在低资源的诊断点诊断设备上部署.
更多相关视频
09:57Comprehensive Protocol to Sample and Process Bone Marrow for Measuring Measurable Residual Disease and Leukemic Stem Cells in Acute Myeloid Leukemia
Published on: March 5, 2018
29.1K
05:24Two Flow Cytometric Approaches of NKG2D Ligand Surface Detection to Distinguish Stem Cells from Bulk Subpopulations in Acute Myeloid Leukemia
Published on: February 21, 2021
4.1K
相关概念视频
Structure and Function of Leukocytes
5.2K
An adult in good health typically has between 4,500 and 11,000 leukocytes, or white blood cells, per microliter of blood, which constitutes about 1% of the total blood volume. Unlike red blood cells, white blood cells contain a nucleus and other cellular organelles but do not have hemoglobin. Most white blood cells reside in connective tissues, particularly in lymphatic organs such as the lymph nodes, with only a small fraction present in circulating blood.
White blood cells protect the body...
White blood cells protect the body...
5.2K
Classification of Leukocytes
9.3K
Leukocytes are classified into two groups based on the presence or absence of cytoplasmic granules. Granular leukocytes, which contain granules, belong to the myeloid lineage and are divided into three subtypes: neutrophils, eosinophils, and basophils. These cells are roughly spherical and characterized by the granules in their cytoplasm.
Neutrophils are the most abundant type of granular leukocytes, comprising 50-70% of all leukocytes. They feature small, evenly distributed granules and a...
Neutrophils are the most abundant type of granular leukocytes, comprising 50-70% of all leukocytes. They feature small, evenly distributed granules and a...
9.3K
Disorders of Leukocytes
2.4K
Leukocyte disorders can lead to either leukopenia, characterized by an abnormally low leukocyte count, or leukocytosis, marked by a very high leukocyte number.
Leukopenia may result from bone marrow disorders, autoimmune diseases, and infectious diseases. For example, conditions such as multiple myeloma and aplastic anemia can impair the bone marrow's ability to produce adequate leukocytes. Similarly, autoimmune diseases like lupus and viral infections such as HIV can prompt the immune...
Leukopenia may result from bone marrow disorders, autoimmune diseases, and infectious diseases. For example, conditions such as multiple myeloma and aplastic anemia can impair the bone marrow's ability to produce adequate leukocytes. Similarly, autoimmune diseases like lupus and viral infections such as HIV can prompt the immune...
2.4K
