研究目的
To develop a label-free classification method for small cell lung cancer (SCLC) and poorly differentiated lung adenocarcinoma (PDLAC) cells using two-dimensional (2D) light scattering static cytometric technique and machine learning.
研究成果
The combination of 2D light scattering technique and machine learning algorithms offers a powerful strategy to differentiate SCLC cells reliably from PDLAC cells with high accuracy. This method may serve as a complementary lung cancer screening method, reducing the need for IHC staining.
研究不足
The classification of H209 cells and H69 cells was not effectively solved, possibly due to the limited angular range collected by the static cytometer. Wide-angle 2D light scattering measurements might provide better resolution.
1：Experimental Design and Method Selection:
The study used a 2D light scattering static cytometer to analyze the light scattering properties of SCLC and PDLAC cells. SVM classifier with LOO-CV was employed for automatic classification.
2：Sample Selection and Data Sources:
SCLC cells (H209 and H69) and PDLAC cells (SK-LU-1) were obtained from the Cell Resource Center, Peking Union Medical College.
3：List of Experimental Equipment and Materials:
A diode pumped solid state (DPSS) laser, optical fiber, glass slides, coverslips, CMOS detector, and conventional flow cytometer (BD FACS Canto II) were used.
4：Experimental Procedures and Operational Workflow:
Cells were cultured, fixed, and resuspended in PBS. The 2D light scattering patterns were obtained using the static cytometer, and data were analyzed using watershed and particle analysis algorithms.
5：Data Analysis Methods:
The number of blobs and their average area in each 2D pattern were extracted and used as inputs for the SVM algorithm for classification.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
