The value of solid wood products is largely determined by the sizes, types and distribution of the knots in the products. Hence, there is a great interest in describing the internal knot structure of individual logs. The Swedish Stem Bank has been extensively used for modelling the interior knot structure of Scots pine (Pinus sylvestris L.) and for simulating the outcome of sawing operations. The stem bank holds parametric descriptions, extracted from computed tomographic (CT) imagery, of mature trees. A method for extracting parametric descriptions, in compliance with the stem bank, from young Scots pine sawlogs is presented in this study. A key step in the algorithm is the use of an artificial neural network to find knots in the CT images. The accuracy of the extracted descriptions was evaluated by comparing the size and position of knots measured on 10 real boards with corresponding boards simulated based on the description. The study showed that the number of knots on the real boards was well predicted (R2=0.90). The differences in tangential and longitudinal position were 0.3±3.6 mm and 1.6±4.2 mm, respectively. The differences in tangential and longitudinal diameter were 0.6±4.0 mm and −0.6±3.9 mm, respectively. Knot diameters were more accurately predicted on boards distant from the pith than on boards close to pith.