RESUMEN
Lung malignancies accounted for 11% of cancers worldwide in 2020 and remained the leading cause of cancer deaths. About 80% of lung cancers belong to non-small cell lung cancer (NSCLC), which is characterized by extremely high clonal and morphological heterogeneity of tumors and development of multidrug resistance. The improvement of current therapeutic strategies includes several directions. First, increasing knowledge in cancer biology results in better understanding of the mechanisms underlying malignant transformation, alterations in signal transduction, and crosstalk between cancer cells and the tumor microenvironment, including immune cells. In turn, it leads to the discovery of important molecular targets in cancer development, which might be affected pharmaceutically. The second direction focuses on the screening of novel drug candidates, synthetic or from natural sources. Finally, "personalization" of a therapeutic strategy enables maximal damage to the tumor of a patient. The personalization of treatment can be based on the drug screening performed using patient-derived tumor xenografts or in vitro patient-derived cell models. 3D multicellular cancer spheroids, generated from cancer cell lines or tumor-isolated cells, seem to be a helpful tool for the improvement of current NSCLC therapies. Spheroids are used as a tumor-mimicking in vitro model for screening of novel drugs, analysis of intercellular interactions, and oncogenic cell signaling. Moreover, several studies with tumor-derived spheroids suggest this model for the choice of "personalized" therapy. Here we aim to give an overview of the different applications of NSCLC spheroids and discuss the potential contribution of the spheroid model to the development of anticancer strategies.
RESUMEN
We have developed a deep learning-based computer algorithm to recognize and predict retinal differentiation in stem cell-derived organoids based on bright-field imaging. The three-dimensional "organoid" approach for the differentiation of pluripotent stem cells (PSC) into retinal and other neural tissues has become a major in vitro strategy to recapitulate development. We decided to develop a universal, robust, and non-invasive method to assess retinal differentiation that would not require chemical probes or reporter gene expression. We hypothesized that basic-contrast bright-field (BF) images contain sufficient information on tissue specification, and it is possible to extract this data using convolutional neural networks (CNNs). Retina-specific Rx-green fluorescent protein mouse embryonic reporter stem cells have been used for all of the differentiation experiments in this work. The BF images of organoids have been taken on day 5 and fluorescent on day 9. To train the CNN, we utilized a transfer learning approach: ImageNet pre-trained ResNet50v2, VGG19, Xception, and DenseNet121 CNNs had been trained on labeled BF images of the organoids, divided into two categories (retina and non-retina), based on the fluorescent reporter gene expression. The best-performing classifier with ResNet50v2 architecture showed a receiver operating characteristic-area under the curve score of 0.91 on a test dataset. A comparison of the best-performing CNN with the human-based classifier showed that the CNN algorithm performs better than the expert in predicting organoid fate (84% vs. 67 ± 6% of correct predictions, respectively), confirming our original hypothesis. Overall, we have demonstrated that the computer algorithm can successfully recognize and predict retinal differentiation in organoids before the onset of reporter gene expression. This is the first demonstration of CNN's ability to classify stem cell-derived tissue in vitro.