Image Dataset of Paper-Based Biosensor for CA19-9 Detection Using Melanin Nanoparticles: Machine Learning and Deep Learning Analysis for Pancreatic Cancer Biomarker Monitoring

This dataset focuses on the development and evaluation of a paper-based colorimetric biosensor utilizing natural melanin nanoparticles derived from cuttlefish for the detection of the pancreatic cancer biomarker CA 19-9. The melanin nanoparticles were extracted, purified, and functionalized with glutaraldehyde to enable antibody immobilization, as confirmed by ATR-FTIR and XPS analyses. Anti-CA 19-9 antibodies were conjugated to the functionalized surface, and the biosensor's interaction with varying concentrations of CA 19-9 solutions—specifically 0.025%, 0.05%, 0.075%, 0.1%, 1%, 2%, 3%, 4%, and 5%—produced distinct color changes. These changes were analyzed using optical readers and digital image processing techniques. The dataset includes images captured via smartphone, featuring both control (represents the blank, unmodified surface) and target regions (the modified surface that interacts with the CA 19-9 solution) to minimize environmental variability, and provides quantitative measurements of color intensity changes corresponding to the CA 19-9 concentrations. The biosensor demonstrated high selectivity, reliability, and sensitivity, validated through repeated measurements and cross-reactivity assays. This dataset supports the development of machine learning and image processing algorithms for accurate, portable, and cost-effective biomarker detection, with potential applications in early cancer diagnosis and monitoring.

本数据集聚焦于一款基于纸质比色生物传感器的开发与评估，该传感器利用源自乌贼的天然黑色素纳米颗粒，用于检测胰腺癌生物标志物CA 19-9。研究人员对该黑色素纳米颗粒进行提取、纯化，并通过戊二醛对其进行功能化修饰以实现抗体固定，该修饰效果经衰减全反射傅里叶变换红外光谱（ATR-FTIR）与X射线光电子能谱（XPS）分析得以验证。将抗CA 19-9抗体偶联至功能化修饰后的传感器表面后，当该生物传感器与不同浓度的CA 19-9溶液（浓度分别为0.025%、0.05%、0.075%、0.1%、1%、2%、3%、4%及5%）发生反应时，会产生可区分的颜色变化。研究人员通过光学读数仪与数字图像处理技术对上述颜色变化进行分析。本数据集包含智能手机拍摄的图像，其中涵盖对照组（代表空白未修饰表面）与靶区域（即与CA 19-9溶液发生反应的修饰后表面），以尽可能降低环境变量带来的影响；同时数据集还提供了与CA 19-9浓度相对应的颜色强度变化的定量测量数据。经多次重复测量与交叉反应性实验验证，该生物传感器展现出优异的选择性、可靠性与灵敏度。本数据集可为机器学习与图像处理算法的开发提供支撑，助力实现精准、便携且低成本的生物标志物检测，在癌症早期诊断与病情监测领域具备潜在应用价值。