A mathematical model for treatment using chemo-immunotherapy

MiR-301b-3p can be used as a Potential Marker for the Diagnosis of Lung Adenocarcinoma

BACKGROUND

The involvement of aberrantly expressed miR-301b-3p has been discovered in diverse human tumors. Our study was primarily centered around the role of miR-301b-3p in diagnosing lung adenocarcinoma (LUAD).

METHOD

We used the TCGA database to download the TCGA-LUAD dataset and selected miR- 301b-3p as the object of our study by differential expression analysis of miRNAs combined with previous studies. The LUAD diagnostic model was constructed utilizing machine learning based on miR-301b-3p expression. The predictive performance of the diagnostic model was found to be excellent by ROC curves combined with the clinical information of the dataset samples. GSEA, GO, and KEGG enrichment analyses demonstrated that miR-301b-3p may mediate the cell cycle by regulating the expression of hormones. Subsequently, combined with tumor immunity and mutation analysis, it was found that patients in the low-expression group had better immune infiltration, indicating that their response rate to immunotherapy may be relatively high. Finally, a mouse xenograft model was constructed to verify how miR-301b-3p affected LUAD progression in mice.

RESULT

The results illustrated that overexpressed miR-301b-3p could cause faster tumor growth in mice. On the contrary, the growth of LUAD could be impeded by the downregulated miR-301b-3p expression. It was suggested that miR-301b-3p had a crucial part in LUAD progression.

CONCLUSION

Overall, the diagnostic performance of the LUAD diagnostic model constructed based on miR-301b-3p is great, and the model can be used as a potential diagnostic marker for LUAD to provide new ideas for clinical diagnosis.

Optimizing cancer therapy for individuals based on tumor-immune-drug system interaction

BACKGROUND AND AIM

Chemotherapy is a crucial component of cancer therapy, albeit with significant side effects. Chemotherapy either damages or inhibits the immune system; therefore, its efficacy varies according to the patient's immune state. Currently, there is no efficient model that incorporates tumor-immune-drug (TID) interactions to guide clinical medication strategies. In this study, we compared five different types of existing TID models with the aim to integrate them into a single, comprehensive model; our goal was to accurately reflect the reality of TID interactions to guide personalized cancer therapy.

METHODS

We studied four different drug treatment profiles: direct function, normal distribution function, sine function, and trapezoid function. We developed a platform capable of plotting all combinations of parameter sets and their corresponding treatment efficiency scores. Subsequently, we generated 10,000 random parameter combinations for an individual case and plotted two polygon graphs using a seismic colormap to depict efficacy of treatment. Then, we developed a platform providing treatment suggestions for all stages of tumors and varying levels of self-immunity. We created polygons demonstrating successful treatments according to parameters related to tumor and immune status.

RESULTS

The trapezoid drug treatment function achieved the best inhibitory effect on the tumor cell density. The treatment can be optimized with a high score indicating that the drug delivery interval had exceeded a specific value. More efficient parameter combinations existed when the immunity was strong compared to when it was weak, thus indicating that increasing the patient's self-immunity can make treatment much more effective.

CONCLUSIONS

In summary, we created a comprehensive model that can provide quantitative recommendations for a gentle, yet efficient, treatment customized according to the individual's tumor and immune system characteristics.

Earlier Detection of Brain Tumor by Pre-Processing Based on Histogram Equalization with Neural Network

MRI is an influential diagnostic imaging technology specifically worn to detect pathological changes in tissues with organs early. It is also a non-invasive imaging method. Medical image segmentation is a complex and challenging process due to the intrinsic nature of images. The most consequential imaging analytical approach is MRI, which has been in use to detect abnormalities in tissues and human organs. The portrait was actualized for CAD (computer-assisted diagnosis) utilizing image processing techniques with deep learning, initially to perceive a brain tumor in a person with early signs of brain tumor. Using AHCN-LNQ (adaptive histogram contrast normalization with learning-based neural quantization), the first image is preprocessed. When compared to extant techniques, the simulation outcome shows that this proposed method achieves an accuracy of 93%, precision of 92%, and 94% of specificity.