Gianluca Marchi
Pleural effusion represents a frequent and diagnostically challenging condition across multiple clinical settings. Conventional characterisation relies on invasive thoracentesis and biochemical analysis, with Light's criteria offering high sensitivity but limited specificity. Procedural risks and patient factors may limit fluid sampling, highlighting the need for complementary, noninvasive approaches. Recent advances in thoracic ultrasound (TUS) suggest that quantitative assessment of pleural fluid (PF) echogenicity, particularly pixel density analysis and hypoechogenicity index, may represent an investigational approach for objective and reproducible evaluation. Preliminary studies using standardised image analysis platforms, such as ImageJ, indicate that exudative effusions more frequently exhibit higher pixel density than transudates, in association with markers of cellularity, protein content, and inflammation. Hybrid scoring systems combining quantitative metrics with morphological sonographic features, such as septations, fibrin strands, and debris, have been explored and may enhance diagnostic specificity, although validation remains limited. However, methodological constraints, operator- and device-dependent variability and small sample sizes continue to restrict their applicability in routine clinical practice. Emerging approaches, including artificial intelligence, may be able to mitigate these limitations by standardising measurements, harmonising grayscale output, and integrating clinical, laboratory and imaging data to generate real-time risk scores. In the absence of robustly validated evidence, quantitative TUS should be regarded as an investigational adjunct, with potential to refine PF characterisation, support clinical decision-making, and inform future research on noninvasive diagnostic strategies.