The emergence of whole slide imaging technology allows for pathology diagnosis on a computer screen. The applications of digital pathology are expanding, from supporting remote institutes suffering from a shortage of pathologists to routine use in daily diagnosis including that of lung cancer. Through practice and research large archival databases of digital pathology images have been developed that will facilitate the development of artificial intelligence (AI) methods for image analysis. Currently, several AI applications have been reported in the field of lung cancer; these include the segmentation of carcinoma foci, detection of lymph node metastasis, counting of tumor cells, and prediction of gene mutations. Although the integration of AI algorithms into clinical practice remains a significant challenge, we have implemented tumor cell count for genetic analysis, a helpful application for routine use. Our experience suggests that pathologists often overestimate the contents of tumor cells, and the use of AI-based analysis increases the accuracy and makes the tasks less tedious. However, there are several difficulties encountered in the practical use of AI in clinical diagnosis. These include the lack of sufficient annotated data for the development and validation of AI systems, the explainability of black box AI models, such as those based on deep learning that offer the most promising performance, and the difficulty in defining the ground truth data for training and validation owing to inherent ambiguity in most applications. All of these together present significant challenges in the development and clinical translation of AI methods in the practice of pathology. Additional research on these problems will help in resolving the barriers to the clinical use of AI. Helping pathologists in developing knowledge of the working and limitations of AI will benefit the use of AI in both diagnostics and research.

Background: Transforming growth factor-β (TGF-β) plays a key role in bone metastasis formation; we hypothesized the possible involvement of TGF-β in the induction of cancer stem cells (CSCs) in the bone microenvironment (micro-E), which may be responsible for chemo-resistance. Methods: Mouse mammary tumor cells were implanted under the dorsal skin flap over the calvaria and into a subcutaneous (subQ) lesions in female mice, generating tumors in the bone and subQ micro-Es. After implantation of the tumor cells, mice were treated with a TGF-β R1 kinase inhibitor (R1-Ki). Results: Treatment with R1-Ki decreased tumor volume and cell proliferation in the bone micro-E, but not in the subQ micro-E. R1-Ki treatment did not affect the induction of necrosis or apoptosis in either bone or subQ micro-E. The number of cells positive for the CSC markers, SOX2, and CD166 in the bone micro-E, were significantly higher than those in the subQ micro-E. R1-Ki treatment significantly decreased the number of CSC marker positive cells in the bone micro-E but not in the subQ micro-E. TGF-β activation of the MAPK/ERK and AKT pathways was the underlying mechanism of cell proliferation in the bone micro-E. BMP signaling did not play a role in cell proliferation in either micro-E. Conclusion: Our results indicated that the bone micro-E is a key niche for CSC generation, and TGF-β signaling has important roles in generating CSCs and tumor cell proliferation in the bone micro-E. Therefore, it is critically important to evaluate responses to chemotherapeutic agents on both cancer stem cells and proliferating tumor cells in different tumor microenvironments in vivo.

Context.— The accurate identification of different lung adenocarcinoma histologic subtypes is important for determining prognosis but can be challenging because of overlaps in the diagnostic features, leading to considerable interobserver variability. Objective.— To provide an overview of the diagnostic agreement for lung adenocarcinoma subtypes among pathologists and to create a ground truth using the clustering approach for downstream computational applications. Design.— Three sets of lung adenocarcinoma histologic images with different evaluation levels (small patches, areas with relatively uniform histology, and whole slide images) were reviewed by 18 international expert lung pathologists. Each image was classified into one or several lung adenocarcinoma subtypes. Results.— Among the 4702 patches of the first set, 1742 (37%) had an overall consensus among all pathologists. The overall Fleiss κ score for the agreement of all subtypes was 0.58. Using cluster analysis, pathologists were hierarchically grouped into 2 clusters, with κ scores of 0.588 and 0.563 in clusters 1 and 2, respectively. Similar results were obtained for the second and third sets, with fair-to-moderate agreements. Patches from the first 2 sets that obtained the consensus of the 18 pathologists were retrieved to form consensus patches and were regarded as the ground truth of lung adenocarcinoma subtypes. Conclusions.— Our observations highlight discrepancies among experts when assessing lung adenocarcinoma subtypes. However, a subsequent number of consensus patches could be retrieved from each cluster, which can be used as ground truth for the downstream computational pathology applications, with minimal influence from interobserver variability.

No abstract

The histopathological distinction of lung adenocarcinoma (LADC) subtypes is subject to high inter-observer variability, which can compromise the optimal assessment of the patient prognosis. Therefore, this study developed convolutional neural networks (CNNs) capable of distinguishing LADC subtypes and predicting disease-specific survival, according to the LADC tumour grades established recently by the International Association for the Study of Lung Cancer pathology committee. Consensus LADC ground truth histopathological images were obtained from seventeen expert pulmonary pathologists and one pathologist in training. Two deep learning models (AI-1 and AI-2) were trained with EfficientNet b3 architecture to predict eight different LADC classes (lepidic, acinar, papillary, micropapillary, solid, invasive mucinous adenocarcinoma, other carcinoma types, and no carcinoma cells). Furthermore, the trained models were tested on an independent cohort of 133 patients. The models achieved high precision, recall, and F1-scores exceeding 0.90 for most of the LADC classes. Clear stratification of the three LADC grades was reached in predicting the disease-specific survival by the two models. Moreover, the grading prediction of one of the trained models was more accurate than those of 14 out of 15 pulmonary pathologists involved in the study (p=0.0003). Both trained models showed high stability in the segmentation of each pair of predicted grades with low variation in the hazard ratio across 200 bootstrapped samples. These findings indicate that the trained CNNs improve the diagnostic accuracy of the pathologist, standardise LADC subtype recognition, and refine LADC grade assessment. Thus, the trained models are promising tools that may assist in the routine evaluation of LADC subtypes and grades in clinical practice.