AI Tool HESpotEx Revolutionizes Cancer Research with Gene Prediction

A new deep learning tool called HESpotEx is transforming how researchers analyse tissue samples. By combining standard microscope images with advanced computational techniques, it predicts gene activity without needing complex lab tests. The system promises to speed up cancer research and improve personalised treatment options. HESpotEx works by linking whole-slide histopathological images (WSIs) with spatial transcriptomics (ST). Unlike traditional methods, it extracts gene expression data directly from stained tissue slides. This removes the need for extra molecular profiling steps, cutting both time and cost.

The tool’s dual-stream design merges image-based features with graph representations. This approach captures not just the visual structure of tissues but also their underlying molecular patterns. As a result, it can predict expression levels for up to 5,457 genes across individual spots within a sample. Tests on large datasets, including The Cancer Genome Atlas (TCGA), show HESpotEx maintains high accuracy. It outperforms existing models on both cancerous and noncancerous tissues. The system also ensures consistent results across different tissue depths and preparation methods. Beyond prediction, HESpotEx improves interpretability by generating gene expression heatmaps overlaid on tissue images. This helps pathologists spot diagnostic markers more easily. The visual output provides clearer insights into malignant changes and other critical features.

HESpotEx could reshape molecular profiling in research and clinical settings. Its ability to predict gene activity from standard slides may lead to faster, more precise cancer detection. The tool’s reliability across diverse datasets further supports its potential for widespread use in personalised medicine.