AIVIS, a leading medical AI company specializing in digital pathology, will present its latest research and product innovations at the 114th Annual Meeting of the United States and Canadian Academy of Pathology (USCAP), taking place March 22–27, 2025, at the Boston Convention and Exhibition Center.

AIVIS will exhibit at Booth #440, offering attendees a firsthand look at its AI-powered diagnostic software, including Qanti IHC, a MFDS- approved tool for immunohistochemistry (IHC) biomarker quantification.

In addition to product demonstrations, AIVIS will participate in 4 poster sessions, presenting new research across several cancer types—including thyroid cancer, invasive urothelial carcinoma, and renal pathology—highlighting the growing potential of AI in diagnostic accuracy, workflow efficiency, and personalized treatment strategies.

AI Accuracy in Thyroid Cancer Diagnosis

Two poster studies will demonstrate the ability of AIVIS’s AI models to accurately classify benign versus malignant thyroid lesions using core needle biopsy slide images. Leveraging tissue contour-guided learning, the model achieved an average AUC of 94.72% and an F1-score of 89.18%, outperforming conventional methods by 5–10%. Notably, AI-generated attention heatmaps closely matched regions annotated by pathologists, emphasizing the tool’s interpretability and clinical value.

Muscle Invasion and Recurrence Risk Prediction in Urothelial Carcinoma

Another study focuses on the prognostic application of AI in invasive urothelial carcinoma. The model predicted muscle invasion with 80% accuracy and recurrence risk with 71.43% accuracy, addressing a long-standing challenge in manual pathology review.

Virtual Staining for Renal Pathology

AIVIS will also unveil its Virtual Staining technology, which utilizes generative AI to transform H&E-stained images into simulated special stains. The model successfully generated PAS, AFOG, and PM staining equivalents, with structural similarity indices (SSIs) ranging from 0.72 to 0.87, demonstrating high fidelity to actual stained slides and eliminating the need for additional chemical staining.

Join AIVIS at USCAP 2025

Visitors are invited to learn more about AIVIS’s digital pathology platform at Booth #440. For details on the company’s poster presentations, see below:

Poster Presentation

Abstract: 1566
Abstract Title: Diagnosis of Thyroid Nodules Using Dual-Attention Multiple Instance Learning in Core Needle Biopsy Whole Slide Images

Abstract: 2055
Abstract Title: A Novel Tissue-Contour Guided MIL Framework for Thyroid Core Needle Biopsy Images Classification

Abstract: 1159
Abstract Title: Weakly Supervised Learning for Predicting Recurrence and Invasion in Bladder Cancer from Transurethral Resection Specimens

Abstract: 1604
Abstract Title: Virtal Straining Using Morphological CycleGAN for Renal Pathology

About AIVIS

AIVIS is a South Korea-based medical AI company specializing in digital pathology. Its flagship product, Qanti IHC, is MFDS-approved and widely used for IHC biomarker quantification, including ER, PR, HER2 Ultralow, and Ki-67. The company has collaborated with over 20 leading hospitals in South Korea and is expanding globally through partnerships with Philips and AstraZeneca, focusing on diagnostic enhancement and AI-powered drug development.