A team at the Vall d'Hebron Research Institute (VHIR) has developed an algorithm that allows the measurement of fat accumulated in the liver using computed tomography (CT) images with and without contrast. This new tool allows non-invasive and rapid detection of metabolic-associated fatty liver disease (MAFLD) and could transform the diagnosis and monitoring of this pathology, which is becoming increasingly prevalent due to the rise in obesity and diabetes. The work has been published in the journal Medical Image Analysis.

MAFLD is characterised by the accumulation of fat in the liver, which can progress to more serious conditions such as steatohepatitis (MASH), liver cancer or cirrhosis. Currently, diagnosis of this disease involves non-invasive techniques such as ultrasound-based elastography, which is not valid for morbidly obese patients who often suffer the most from this pathology, or MRI, but the definitive method is liver biopsy.

The new algorithm overcomes the limitations of current methods by allowing a detailed and accurate assessment of liver fat from CT images in an automated way. The system analyses the radiological density of the liver and spleen to identify areas with accumulated fat. "With this algorithm, we can provide detailed information about the distribution of fat in the liver, which is crucial for accurate diagnosis and effective monitoring of MAFLD", explains Dr. J. Raul Herance, head of the Molecular Medical Imaging group at VHIR who led the study. "Our tool reduces the need for invasive biopsies, which improves the patient experience and facilitates early diagnosis. On the other hand, it also eliminates the limitations of ultrasound in morbid obese patients or the slowness of MRI that makes it almost inoperative for use as a routine diagnostic tool", says Dr. Herance.

The algorithm has been validated against elastography and patient biopsy data and has demonstrated high accuracy in measured liver fat values in both contrast and non-contrast CT images. This allows its application to a wide variety of clinical contexts. Furthermore, unlike biopsy, which only provides information from a specific area of the liver, CT provides data from the whole organ.

CT is a widely available and cheaper technique than other imaging techniques such as MRI. It is a method that is commonly used in clinical practice and, therefore, would allow assessment of the presence of fatty liver in the diagnosis of other abdominal or thoracic diseases without the need for additional radiation exposure.

The team will continue to work to validate this tool with larger cohorts of patients and hope that its implementation in clinical practice will contribute to the early detection and clinical management of fatty liver disease. In this way, they hope to prevent serious complications associated with this pathology and improve the health and quality of life of patients.

The work has been led by the Molecular Medical Imaging group at VHIR in collaboration with the Liver Diseases; Diabetes and Metabolism; and Clinical Biochemistry, Drug Delivery and Therapy groups at VHIR, Pompeu Fabra University, the CIBER of Bioengineering, Biomaterials and Nanomedicine (CIBER-BBN), the CIBER of Liver and Digestive Diseases (CIBEREHD) and the CIBER of Diabetes and Metabolic Diseases (CIBERDEM).