Search Results
Semantic segmentation of non-linear multimodal images for disease grading of inflammatory bowel disease: A segnet-based application
Non-linear multimodal imaging, the combination of coherent anti-stokes Raman scattering (CARS), two-photon excited fluorescence (TPEF) and second harmonic generation (SHG), has shown its potential to assist the diagnosis of different inflammatory bowel diseases (IBDs). This label-free imaging technique can support the ‘gold-standard’ techniques such as colonoscopy and histopathology to ensure an IBD diagnosis in clinical environment. Moreover, non-linear multimodal imaging can measure biomolecular changes in different tissue regions such as crypt and mucosa region, which serve as a predictive marker for IBD severity. To achieve a real-time assessment of IBD severity, an automatic segmentation of the crypt and mucosa regions is needed. In this paper, we semantically segment the crypt and mucosa region using a deep neural network. We utilized the SegNet architecture (Badrinarayanan et al., 2015) and compared its results with a classical machine learning approach. Our trained SegNet mod el achieved an overall F1 score of 0.75. This model outperformed the classical machine learning approach for the segmentation of the crypt and mucosa region in our study.
Nonresonant Raman spectroscopy of isolated human retina samples complying with laser safety regulations for in vivo measurements
Retinal diseases, such as age-related macular degeneration, are leading causes of vision impairment, increasing in incidence worldwide due to an aging society. If diagnosed early, most cases could be prevented. In contrast to standard ophthalmic diagnostic tools, Raman spectroscopy can provide a comprehensive overview of the biochemical composition of the retina in a label-free manner. A proof of concept study of the applicability of nonresonant Raman spectroscopy for retinal investigations is presented. Raman imaging provides valuable insights into the molecular composition of an isolated ex vivo human retina sample by probing the entire molecular fingerprint, i.e., the lipid, protein, carotenoid, and nucleic acid content. The results are compared to morphological information obtained by optical coherence tomography of the sample. The challenges of in vivo Raman studies due to laser safety limitations and predefined optical parameters given by the eye itself are explored. An in-house built setup simulating the optical pathway in the human eye was developed and used to demonstrate that even under laser safety regulations and the above-mentioned optical restrictions, Raman spectra of isolated ex vivo human retinas can be recorded. The results strongly support that in vivo studies using nonresonant Raman spectroscopy are feasible and that these studies provide comprehensive molecular information of the human retina. © The Authors. Published by SPIE.