Malignancies are still responsible for a large share of lethalities. Macroscopical evaluation of the surgical resection margins is uncertain. Big data based imaging approaches have emerged in the recent decade (mass spectrometry, two-photon microscopy, infrared and Raman spectroscopy). Indocianine green labelled MS is the most common approach, however, label free mid-infrared imaging is more promising for future practical application. We aimed to identify and separate different transformed (A-375, HT-29) and non-transformed (CCD986SK) cell lines by a label-free infrared spectroscopy method. Our approach applied a novel set-up for label-free mid-infrared range classification method. Transflection spectroscopy was used on aluminium coated glass slides. Both whole range spectra (4000–648 cm⁻¹) and hypersensitive fingerprint regions (1800–648 cm⁻¹) were tested on the imaged areas of cell lines fixed in ethanol. Non-cell spectra were possible to be excluded based on mean transmission values being above 90%. Feasibility of a mean transmission based spectra filtering method with principal component analysis and linear discriminant analysis was shown to separate cell lines representing different tissue types. Fingerprint region resulted the best separation of cell lines spectra with accuracy of 99.84% at 70–75 mean transmittance range. Our approach in vitro was able to separate unique cell lines representing different tissues of origin. Proper data handling and spectra processing are key steps to achieve the adaptation of this dye-free technique for intraoperative surgery. Further studies are urgently needed to test this novel, marker-free approach.

恶性肿瘤仍是造成大量杀伤力的原因。手术切除切缘的宏观评估尚不确定。近十年来出现了基于大数据的成像方法（质谱，双光子显微镜，红外和拉曼光谱）。印度绿标记的MS是最常用的方法，但是，无标记的中红外成像对于未来的实际应用更有希望。我们旨在通过无标记红外光谱法鉴定和分离不同的转化细胞（A-375，HT-29）和未转化细胞（CCD986SK）。我们的方法为无标签的中红外范围分类方法应用了一种新颖的设置。透射光谱法用于涂铝的载玻片上。整个光谱范围（4000–648 cm ^-1）和超敏指纹区域（1800–648 cm ^-1在固定在乙醇中的细胞系成像区域上测试）。基于平均透射率高于90％，可以排除非细胞光谱。具有基于主成分分析和线性判别分析的基于平均透射的光谱过滤方法的可行性被证明可以分离代表不同组织类型的细胞系。指纹图谱区域能最好地分离细胞系光谱，在70-75的平均透射率范围内，准确度为99.84％。我们的体外方法能够分离代表不同来源组织的独特细胞系。正确的数据处理和光谱处理是实现这种无染料技术适用于术中手术的关键步骤。迫切需要进一步研究以测试这种新颖的，无标记的方法。