BackgroundCancer diagnosis in Colombia faces delays and regional inequities, particularly in rural and underserved areas where access to subspecialist pathology is limited. General pathologists in these contexts frequently manage complex oncological cases with limited expert support, which may increase diagnostic uncertainty and professional burden.MethodsA concurrent mixed-methods design was applied within the GLORIA (Globalization of a telepathology network with artificial intelligence applications) project framework. Quantitative data were collected from a national survey distributed to ASOCOLPAT-registered pathologists (n = 57). Qualitative data were obtained through conversational interviews with general pathologists from six small and mid-sized laboratories (n = 11).FindingsMost participants reported needing subspecialist support for complex oncological cases (100%) and reported high emotional stress (86%), mainly associated with diagnostic uncertainty, legal responsibility, and excessive workload. Consultation response times differed across contexts, ranging from 1 to 7 days in Bogotá to more than 30 days in small cities. While 86% reported familiarity with telepathology, 63.2% had never used it in routine practice; however, 94% believed slide scanners could improve diagnostic timeliness. Qualitative findings identified recurrent themes related to administrative authorization barriers, logistical delays, professional isolation, and the perceived value of telepathology for timely expert feedback and learning.InterpretationThis baseline assessment suggests that telepathology implementation in Colombia must address not only technical readiness but also structural barriers in referral workflows and the professional needs of general pathologists, including timely feedback, shared responsibility, and learning opportunities.
Adapting DeepLabV3+ for biopsy cervical cancer lesion segmentation
IntroductionCervical cancer remains a leading cause of cancer mortality in resource-constrained settings, where access to advanced digital pathology equipment is severely limited. Automated histopathological image


