Artificial Intelligence in Biliary Atresia Diagnosis: A Bibliometric Analysis and Categorization of Approaches for the Use of Low-Complexity Techniques in Primary Care

Biliary Atresia (BA) is a rare but serious neonatal cholestatic disease that requires early diagnosis to avoid undesirable long-term prognoses. Although Artificial Intelligence (AI) has shown promise in aiding the diagnosis of various pathologies, its application in BA remains limited. This study conducted a bibliometric and content analysis to map the current landscape of AI approaches for the diagnosis of biliary atresia and to identify research gaps, with particular attention to low-complexity techniques suitable for use in primary care centers. Searches were performed in the Scopus database (2014-2024). An initial comprehensive search on AI applications in the healthcare sector resulted in 8,253 publications. Subsequently, more specific searches identified 92 studies related to the use of Case-Based Reasoning (CBR) in healthcare and 266 to the use of CBR combined with fuzzy logic. Only eight studies specifically addressed the diagnosis of biliary atresia. Bibliometric analysis using VOSviewer revealed a strong concentration of studies on deep learning models (particularly convolutional neural networks) applied to gallbladder ultrasound images and conventional machine learning models using clinical and laboratory data. Content analysis of the eight articles related to biliary atresia confirmed that the proposed approaches rely heavily on the availability of large image samples or multiple laboratory analyses, making them impractical for use in resource-limited primary care settings. Additionally, no studies were found related to the use of expert systems, fuzzy inference systems, or case-based reasoning in the diagnosis of biliary atresia. These less complex techniques require fewer input variables and smaller training samples and represent underexplored alternatives with research potential to support the early detection of biliary atresia in primary care, especially in resource-limited settings.