The investigators investigate the heterogeneity fatty liver disease (FLD) as well as risk factors associated with progression of the liver diseases or development of cardiovascular complication. The overarching aim is to develop a AI-based prediction model that could be integrated into the electronic medical record system.
To identify large numbers of unselected individuals with FLD, the investigators developed a phrase recognition algorithm, which identifies FLD using radiology reports. This allows them to study also individuals who have had abdominal imaging for any reason (usually not for suspected liver disease). Within the data lake of Helsinki University Hospital, the investigators identified 61,271 individuals with FLD and 155,521 without. The investigators have obtained national register data on their diagnosed diseases and medication.
The investigators will now invite a random sample of 1000-1500 individuals with FLD (aged 20-60 years at the time of the previous imaging) to a follow-up study. Progression of liver disease will be studied by elastography, measurement of fat content and a blood sample for fibrosis markes (ELF score). The participants' metabolism will be evaluated with several laboraroty tests and an i.v. glucagon tolerance test and amino acid response. Questionnaires will be used to obtain life-style information.
