Abstract: The medicinal agent effect prediction system includes: a storage unit that records a learned model based on a plurality of teaching data obtained by associating input data for learning, which is based on images of neurites in culture spaces respectively growing from nerve cells cultured in a state where each of known substances is individually added to a culture solution, with output data for learning which relates to information about impaired nerve sites resulting from the known substances; a data input receiving unit that receives input of input data for assessment based on a microscope image of the neurites; and an assessment unit that outputs information derived from probability of occurrence of impairment of each nerve site belonging to the group of impaired site candidates resulting from introduction of the test substance, by applying the input data for assessment to the learned model.

Abstract: A method for predicting the characteristics of a target compound according to the present invention comprises: (1) a step for receiving activity data on nerve cells for the target compound; (2) a step for converting the activity data into an image; (3) a step for inputting the image into an image recognition model trained by a training data set, the training data set including a plurality of training images obtained by converting the activity data on nerve cells for a plurality of known compounds; and (4) a step for processing the image in the image recognition model and outputting at least one characteristic of the target compound.

Abstract: The present invention provides a method for measuring the propagation of electric activity in neural network and/or axon by using a nerve cell device produced by seeding nerve cells on an oriented fiber sheet. According to the present invention, synapse function can be assessed by using propagation speed of neural network as an index.

Abstract: This method for predicting a characteristic of a target includes: (1) a step of obtaining first data relating to a first object; (2) a step of obtaining second data relating to a second object; (3) a step of identifying a prediction parameter set by performing multivariate analysis of the first data and the second data for all combinations of candidate parameters in a candidate parameter set; and (4) a step of predicting the characteristic on the basis of the prediction parameter set, from target data relating to the target.

Abstract: A surface-coated tungsten carbide-based cemented carbide blade member including a tungsten carbide-based cemented carbide substrate, and a hard coating of an average layer thickness of 3 to 30 .mu.m formed thereon is disclosed. The hard coating includes a first layer deposited on the substrate, a second layer deposited on the first layer, and a third layer deposited on the second layer. The first layer has a granular crystal structure, while the second layer has an elongated crystal structure. Among the constituents forming the substrate, at least tungsten and cobalt are diffused into the grain boundaries of the first and second layers.