Spectral histology of breast tissue using mid-infrared spectroscopic imaging

Histopathologic recognition is the gold standard in breast cancer diagnoses and is a primary determinant tool for cancer research. Unfortunately, the manual nature of histopathologic recognition leads to low throughput analysis, delays in decision-making and errors. Here, we present an automated means to accurate histologic recognition using mid-infrared molecular spectroscopy. Fourier transform infrared (FT-IR) spectroscopic imaging is combined with statistical pattern recognition and high throughput sampling to provide automated tissue segmentation into constituent cell types. The method does not need dyes or probes and dispenses with human input. Results demonstrate that the technique is capable of accurate histologic segmentation that can potentially become competitive with that attained by conventional immunohistochemical analyses.