Tumor-associated macrophages (TAMs), one of the most abundant immune components in gastric cancer (GC), are difficult to characterize due to their heterogeneity. Multiple approaches have been used to elucidate the issue, however, due to the tissue-destructive nature of most of these methods, the spatial distribution of TAMs in situ remains unclear. Here we probe the relationship between tumor context and TAM heterogeneity by multiplex immunohistochemistry of 56 human GC cases. Using distinct expression marker profiles on TAMs, we report seven predominant populations distributed between tumor and non-tumor tissue. TAM population-associated gene signatures reflect their heterogeneity and polarization in situ. Increased density of CD163+ (CD206−) TAMs with concurrent high CD68 expression is associated with upregulated immune-signaling and improved patient survival by univariate, but not multivariate analysis. CD68-only and CD206+ TAMs are correlated with high PDL1 expression.

Segmentation of the prostate gland from magnetic resonance images is rapidly becoming a standard of care in prostate cancer radiotherapy treatment planning. Automating this process has the potential to improve accuracy and efficiency. However, the performance and accuracy of deep learning models varies depending on the design and optimal tuning of the hyper-parameters. In this study, we examine the effect of loss functions on the performance of deep-learning-based prostate segmentation models. A U-Net model for prostate segmentation using T2-weighted images from a local dataset was trained and performance compared when using nine different loss functions, including: Binary Cross-Entropy (BCE), Intersection over Union (IoU), Dice, BCE and Dice (BCE + Dice), weighted BCE and Dice (W (BCE + Dice)), Focal, Tversky, Focal Tversky, and Surface loss functions. Model outputs were compared using several metrics on a five-fold cross-validation set. Ranking of model performance was found to be dependent on the metric used to measure performance, but in general, W (BCE + Dice) and Focal Tversky performed well for all metrics (whole gland Dice similarity coefficient (DSC): 0.71 and 0.74; 95HD: 6.66 and 7.42; Ravid 0.05 and 0.18, respectively) and Surface loss generally ranked lowest (DSC: 0.40; 95HD: 13.64; Ravid −0.09). When comparing the performance of the models for the mid-gland, apex, and base parts of the prostate gland, the models’ performance was lower for the apex and base compared to the mid-gland. In conclusion, we have demonstrated that the performance of a deep learning model for prostate segmentation can be affected by choice of loss function. For prostate segmentation, it would appear that compound loss functions generally outperform singles loss functions such as Surface loss.

Background: There are currently no methods to estimate cell density in the prostate. This study aimed to develop predictive models to estimate prostate cell density from multiparametric magnetic resonance imaging (mpMRI) data at a voxel level using machine learning techniques. Material and methods: In vivo mpMRI data were collected from 30 patients before radical prostatectomy. Sequences included T2-weighted imaging, diffusion-weighted imaging and dynamic contrast-enhanced imaging. Ground truth cell density maps were computed from histology and co-registered with mpMRI. Feature extraction and selection were performed on mpMRI data. Final models were fitted using three regression algorithms including multivariate adaptive regression spline (MARS), polynomial regression (PR) and generalised additive model (GAM). Model parameters were optimised using leave-one-out cross-validation on the training data and model performance was evaluated on test data using root mean square error (RMSE) measurements. Results: Predictive models to estimate voxel-wise prostate cell density were successfully trained and tested using the three algorithms. The best model (GAM) achieved a RMSE of 1.06 (± 0.06) × 10³ cells/mm² and a relative deviation of 13.3 ± 0.8%. Conclusion: Prostate cell density can be quantitatively estimated non-invasively from mpMRI data using high-quality co-registered data at a voxel level. These cell density predictions could be used for tissue classification, treatment response evaluation and personalised radiotherapy.

Objective: To investigate the association between multiparametric MRI (mpMRI) imaging features and hypoxia-related genetic profiles in prostate cancer. Methods: In vivo mpMRI was acquired from six patients prior to radical prostatectomy. Sequences included T 2 weighted (T2W) imaging, diffusion-weighted imaging, dynamic contrast enhanced MRI and blood oxygen-level dependent imaging. Imaging data were co-registered with histology using three-dimensional deformable registration methods. Texture features were extracted from T 2W images and parametric maps from functional MRI. Full transcriptome genetic profiles were obtained using next generation sequencing from the prostate specimens. Pearson correlation coefficients were calculated between mpMRI data and hypoxia-related gene expression levels. Results were validated using glucose transporter one immunohistochemistry (IHC). Results: Correlation analysis identified 34 candidate imaging features (six from the mpMRI data and 28 from T 2W texture features). The IHC validation showed that 16 out of the 28 T 2W texture features achieved weak but significant correlations (p < 0.05). Conclusions: Weak associations between mpMRI features and hypoxia gene expressions were found. This indicates the potential use of MRI in assessing hypoxia status in prostate cancer. Further validation is required due to the low correlation levels. Advances in knowledge: This is a pilot study using radiogenomics approaches to address hypoxia within the prostate, which provides an opportunity for hypoxia-guided selective treatment techniques.