ポスター発表
- 第3日 5月13日(水) P会場(1008/09)
-
3P-07 PDF
PESI-MSと機械学習による乳がん診断
Detection in the early stages of the breast cancer (BC) assures good prognosis. For this purpose, imaging techniques sometimes show less sensitivity and pathological examination needs multiple steps and the duration of preparation. For more accurate and less time-consuming diagnostic method using small specimen, we have applied probe electrospray ionization mass spectrometry (PESI-MS) and machine learning (Logistic regression) to biological systems. This system enables us to distinguish a cancerous tissue from a non-cancerous tissue using almost raw samples without specific laborious pretreatments. In this study, we undertook two directions of study in BC. In both cases, mass spectra were acquired from non- cancerous and cancerous tissues. In the first part of study using PESI-MS, we established a diagnostic algorithm by machine learning using mass spectra. In the second part of study, we used LC-MS/MS to identify several specific molecules that were markedly increased in BC and used them as a collective biomarker. Collectively, we demonstrated that PESI-MS combined with machine learning has the potential to establish a lipid-based diagnosis of BC with higher accuracy using a simpler approach.