Artificial intelligence‐assisted histopathological diagnosis of endocervical gastric‐type adenocarcinoma: a multicenter model development and validation study

Vollständiger Abstract

Worum geht es in dieser Arbeit?

Abstract Endocervical gastric‐type adenocarcinoma (GAS) is one of the most aggressive subtypes of cervical cancer and is frequently underdiagnosed due to morphological ambiguity, leading to delayed diagnosis. Despite the availability of molecular and genomic assays, their high cost, complexity, and limited reproducibility restrict clinical use. This study therefore proposes a highly sensitive artificial intelligence (AI)–assisted diagnostic system for GAS based exclusively on H&E‐stained histopathological images. We included 309 slides from 96 GAS cases collected at Peking University Third Hospital from January 2018 to January 2025, representing the largest GAS cohort reported to date for AI research. In addition, we incorporated other morphologically analogous diseases, encompassing a total of 1,320 slides sourced from four categories: normal cervical mucosa (NORM), benign endocervical lesion entities (BELE), HPV‐associated adenocarcinoma (HPVA), and endometrioid carcinoma with mucinous differentiation (ECMD). We developed GASPath, based on a novel multiple instance learning framework that efficiently captures fine‐grained morphological variations from H&E‐stained images. Beyond internal validation, GASPath was evaluated across 12 independent retrospective cohorts and further subjected to large‐scale real‐world validation on more than 7,000 samples from March 2024 to April 2025. Across three stages, GASPath demonstrated high performance. In internal validation (Stage I), it achieved an accuracy of 0.980 (95% CI 0.977–0.983) and an ROC‐AUC of 0.995 (95% CI 0.994–0.997). In external validation (Stage II), the sensitivity reached 0.902 and improved to 0.968 with proposed strategies. For biopsy samples, GASPath achieved an ROC‐AUC of 0.990 (95% CI 0.984–0.997). In large‐scale real‐world deployment (Stage III, n = 7,056), GASPath achieved a balanced accuracy of 0.953, with 100% sensitivity for GAS (45/45 cases correctly identified). The heatmaps highlight morphological features of GAS that are easily underestimated, such as irregular, angulated glands, subtle loss of nuclear polarity, and mild cytologic atypia, which show substantial morphological overlap with other diagnostic categories. GASPath enables high‐sensitivity detection of GAS in routine H&E‐stained slides, obviating the need for extensive auxiliary testing while preventing underdiagnosis and misdiagnosis. This advancement addresses a critical gap by streamlining diagnostic workflows without compromising accuracy. Its implementation could enable cost‐effective, scalable AI‐assisted diagnostics, potentially transforming the early detection and management of this aggressive cancer subtype.