From slides to insights: Harnessing deep learning for prognostic survival prediction in human colorectal cancer histology

Prognostic survival prediction in colorectal cancer (CRC) plays a crucial role in guiding treatment decisions and improving patient outcomes. In this research, we explore the application of deep learning techniques to predict survival outcomes based on histopathological images of human colorectal cancer. We present a retrospective multicenter study utilizing a dataset of 100,000 nonoverlapping image patches from hematoxylin & eosin-stained histological images of CRC and normal tissue. The dataset includes diverse tissue classes such as adipose, background, debris, lymphocytes, mucus, smooth muscle, normal colon mucosa, cancer-associated stroma, and colorectal adenocarcinoma epithelium. To perform survival prediction, we employ various deep learning architectures, including convolutional neural network, DenseNet201, InceptionResNetV2, VGG16, VGG19, and Xception. These architectures are trained on the dataset using a multicenter retrospective analysis approach. Extensive preprocessing steps are undertaken, including image normalization using Macenko's method and data augmentation techniques, to optimize model performance. The experimental findings reveal promising results, demonstrating the effectiveness of deep learning models in prognostic survival prediction. Our models achieve high accuracy, precision, recall, and validation metrics, showcasing their ability to capture relevant histological patterns associated with prognosis. Visualization techniques are employed to interpret the models' decision-making process, highlighting important features and regions contributing to survival predictions. The implications of this research are manifold. The accurate prediction of survival outcomes in CRC can aid in personalized medicine and clinical decision-making, facilitating tailored treatment plans for individual patients. The identification of important histological features and biomarkers provides valuable insights into disease mechanisms and may lead to the discovery of novel prognostic indicators. The transparency and explainability of the models enhance trust and acceptance, fostering their integration into clinical practice. Research demonstrates the potential of deep learning models for prognostic survival prediction in human colorectal cancer histology. The findings contribute to the understanding of disease progression and offer practical applications in personalized medicine. By harnessing the power of deep learning and histopathological analysis, we pave the way for improved patient care, clinical decision support, and advancements in prognostic prediction in CRC.

Increased Tenascin-C expression contributes to cardiac dysfunction and fibrosis in Duchenne muscular dystrophy

Introduction and aims

Cardiac fibrosis is characterized by the net accumulation of extracellular matrix (ECM) proteins in the cardiac interstitium and contributes to cardiac contractile dysfunction. In Duchenne muscular dystrophy (DMD), cardiomyopathy develops as a result of a dystrophin deficiency causing fibrofatty replacement of the myocardium, however the underlying mechanisms are not fully understood. There is a growing collection of evidence that ECM proteins, including Tenascin C (TN-C), plays a maladaptive role in left ventricular (LV) remodelling and cardiac fibrosis in ischemic heart disease. The aims of our study were 1) to assess TN-C levels, fibrosis and cardiac dysfunction in DMD patients, and 2) to clarify the role of TN-C in cardiovascular dysfunction and fibrosis using male mdx (n=10) and mdx TN-C KO mice (n=8).

Results

In male patients with DMD (n=18) and age matched controls (n=12) undergoing cardiac MRI, we detected greater myocardial fibrosis than in control hearts. In addition, we observed an elevation of TN-C plasma levels [median concentration (3.55); interquartile range (0.61–7.43) ng/mL] in DMD patients, and its expression negatively correlated to LV ejection fraction (EF) [median LVEF (45); interquartile range (37.5–51.5) %]. Male wt, mdx and mdx TN-C KO age-matched (10 months) mice were used. Transthoracic echocardiography was performed and fibrosis was assessed on cardiac tissue sections. Wire myography was used to assess vascular endothelial function. To explore the signalling pathways contributing to cardiac fibrosis, human cardiac fibroblasts (hCFs) were treated with recombinant human TN-C or TGF-β and gene expression and epigenetic regulation of NF-kB/p65 were assessed. Mdx mice showed significantly increased cardiac fibrosis which was accompanied with markedly elevated TN-C level in cardiac tissue and plasma compared to wt animals (p<0.05, respectively). Moreover, TN-C level in plasma correlated positively with the degree of cardiac dilation in dystrophic mice. In addition, vascular endothelial function was notably impaired in mdx mice. In contrast, we observed preserved vascular function in mdx- TN-C KO mice, this was accompanied by a significant reduction in cardiac fibrosis in compared to age-matched mdx mice (p<0.05, respectively). hCFs treated with TN-C or TGF-β showed increased collagen and α-SMA expressions which could be prevented by application of siRNA against TN-C. In addition, both TN-C and TGF-β caused p65/NF-κB promoter demethylation and subsequently triggered pro-inflammatory and pro-fibrotic signalling, which could be reversed by applying p38 MAPK inhibitor in hCFs.

Conclusion

TN-C is a critical component of cardiac fibrosis and cardiac dysfunction in DMD. The activation of NF-κB p65 signalling pathway may play a role in TN-C induced fibrosis. Thus, TN-C may be a mediator and potential target for therapy in DMD-associated cardiovascular complications.

Funding Acknowledgement

Type of funding sources: Foundation. Main funding source(s): Österreichische MuskelforschungFWF - Austrian Science Found P 35878

Influence of Temperature and Humidity on Pregnancy Rate of Murrah Buffaloes under Subtropical Climate

Heat stress has adverse effects on fertility of dairy animals. Decline in fertility is linearly associated with an increase in combination of both temperature and humidity. The purpose of this study was to investigate the relationship between temperature humidity index (THI) and the pregnancy rate of Murrah buffaloes in a subtropical climate. The effects of genetic and non-genetic factors viz., sire, parity, period of calving and age group at first calving were found non-significant on pregnancy rate. The effect of THI was found significant (p<0.001) on pregnancy rate of Murrah buffaloes calved for first time and overall pregnancy rate. The threshold THI affecting the pregnancy rate was identified as THI 75. The months from October to March showed THI<75 and considered as non heat stress zone (NHSZ), while months from April to September were determined as heat stress zone (HSZ) with THI≥75. The lowest overall pregnancy rate (0.25) was obtained in July with THI 80.9, while the highest overall pregnancy rate (0.59) was found in November with THI 66.1. May and June were identified as critical heat stress zone (CHSZ) within the HSZ with maximum decline (-7%) in pregnancy rate with per unit increase in THI. The highest overall pregnancy rate was estimated as 0.45 in NHSZ with THI value 56.7 to 73.2. The pregnancy rate was found to have declined to 0.28 in HSZ with THI 73.5 to 83.7. However, the lowest pregnancy rate was estimated as 0.27 in CHSZ with THI value 80.3 to 81.6.

Design and evaluation of biodegradable, biosensitive in situ gelling system for pulsatile delivery of insulin

Biodegradable glucose-sensitive in situ gelling system based on chitosan for pulsatile delivery of insulin was developed. The sols/gels were thoroughly characterized for swelling properties, rheology, texture analysis and water content. The developed glucose-sensitive gels responded to varied glucose concentrations in vitro indicating their ability to function as environment-sensitive systems. Insulin load onto the gels was optimized and was found to affect the rheological behavior of these gels, the final preparation used for in vitro contained 1IU/200mul of the sol. These gels released the entrapped insulin in a pulsatile manner in response to the glucose concentration in vitro. Furthermore, the formulations when evaluated for their in vivo efficacy in streptozotocin-induced diabetic rats at a dose of 3IU/kg, demonstrated their ability to release insulin in response to glucose concentration and were preferred much better against subcutaneously given plain insulin formulation used as the control. Together, these preliminary results indicate that biosensitive chitosan in situ gelling systems have substantial potential as pulsatile delivery systems for insulin.

Hydrogels for pharmaceutical and biomedical applications

Hydrogels are crosslinked hydrophilic polymer structures that can imbibe large amounts of water or biological fluids. Hydrogels are one of the upcoming classes of polymer-based systems that embrace numerous biomedical and pharmaceutical applications. This review discusses various parameters of hydrogels such as surface properties, water content and swelling behavior, effect of nature of polymer, ionic content, and thermodynamics, all of which can influence the biomedical usage of hydrogels. Meanwhile, intelligent or environment-sensitive hydrogels and bioadhesive hydrogels continue to be important materials for medical applications; therefore, a part of this review is devoted to some of their important classes. Hydrogels are extensively used for various biomedical applications--tissue engineering, molecular imprinting, wound dressings materials, immunoisolation, drug delivery, etc. Thus, this review aims to throw light on the numerous applications that hydrogels have in the biomedical arena.