Search

Ovarian cancer, particularly high-grade serous carcinoma, is highly lethal, and accurate survival prediction is essential for treatment planning. However, traditional prognostic models rely on limited clinical and histologic features, while deep learning approaches require expensive pixel-level annotations of whole-slide histopathology images, limiting scalability. We propose a weakly supervised attention-based multiple instance learning (MIL) framework that predicts ovarian cancer survival using only slide-level survival labels. Each whole-slide image is treated as a bag of patches, where a patch encoder extracts features using a pre-trained CNN or vision transformer. An attention-based MIL aggregator assigns importance weights to patches, and a survival head outputs a risk score via a deep Cox model. The attention mechanism enhances interpretability by identifying prognostically relevant regions such as aggressive tumor morphology, stromal patterns, and immune infiltration. This reduces the need for manual annotation while preserving clinical relevance. The framework provides a scalable and interpretable approach for survival prediction and can be evaluated on datasets such as TCGA-OV and CPTAC for clinical translation.

Pancreatic cancer is highly lethal, and surgical resection is the only curative option. Preoperative assessment using contrast-enhanced CT is essential for determining tumor resectability based on involvement of key vessels such as the superior mesenteric artery, celiac trunk, and portal vein. Accurate pancreatic tumor segmentation is difficult due to unclear boundaries, low contrast with surrounding tissue, and proximity to major vessels. Manual segmentation is slow, subjective, and inconsistent, especially in borderline cases, while tumor-associated fibrosis further obscures lesion margins. We propose a deep learning-based framework using an attention-enhanced U-Net with multi-scale feature fusion and deep supervision for tumor segmentation and resectability assessment. The model incorporates attention gates, atrous spatial pyramid pooling, and auxiliary losses at multiple decoder levels to improve feature learning and gradient flow. A pre-trained encoder extracts hierarchical features refined by attention mechanisms in skip connections. A multi-scale decoder reconstructs segmentation maps, supported by deep supervision at different resolutions. A parallel branch models tumor–vessel spatial relationships using distance maps to improve resectability classification. This framework enables automated pancreatic tumor segmentation and resectability evaluation from CT scans, improving accuracy, interpretability, and clinical utility. Validation on datasets such as Pancreas-CT and Medical Segmentation Decathlon is recommended.