AI uses chromatin as a biomarker for the (early) diagnosis and treatment evaluation of cancer
December 2023
Paul-Scherrer Institute, Villigen, Switzerland
Results of previous studies show that tumors influence the chromatin conformation in peripheral blood mononuclear cells (PBMCs) through so-called secretome signals. In addition, the concentration of the secretomes depends on the stage of the disease. Following this, the present study evaluates the potential of chromatin biomarkers for early diagnosis and treatment evaluation of cancer. For this purpose, the three-dimensional chromatin organization from blood samples from 10 healthy volunteers and various tumor patients was first imaged and analysed using fluorescence methods. Characteristic chromatin patterns and features were transformed into subject-specific data sets and fed into a machine learning system for prediction. In a first series of tests, the data sets from 10 healthy volunteers were compared with the information from 10 tumor patients. The AI was able to distinguish between healthy and sick people with a high level of sensitivity based on the chromatic characteristics. In further series of tests, it was also shown that the AI could distinguish three tumor groups from one another with up to 89% accuracy. Finally, the chromatin conformation of 30 tumor patients (10 glioma, 10 meningioma and 10 head and neck tumor patients) who underwent proton irradiation was examined. The blood samples were taken and analysed at three different stages during therapy. The results showed characteristic cellular changes that indicated a response to therapy and will be further developed in future clinical studies. In summary, the AI-based chromatin biomarker analysis proves to be a new and valuable method that can help to diagnose cancer at an early stage, to predict the success or response of patient-specific therapeutic approaches and to avoid the use of invasive diagnostic procedures in seriously ill patients.
Imaging and AI based chromatin biomarkers for diagnosis and therapy evaluation from liquid biopsies
damien.weber@psi.ch, Damien C. Weber
Added on: 12-19-2023
[1] https://www.nature.com/articles/s41698-023-00484-8[2] https://www.bionity.com/en/news/1182300/enabling-early-detection-of-cancer-ai-helps.html?utm_source=newsletter&utm_medium=email&utm_campaign=bionityde&WT.mc_id=ca0264
