Weighted least squares fitting with multiple templates for detection of small spontaneous signals.

Cultured neurons have been used for investigating synaptic plasticity due to their accessibility to biochemical and immunochemical assays. In addition, recording spontaneous postsynaptic miniature events provides important information about the mechanisms involved in the modulation of synaptic transmission. To automatically detect the spontaneous events, we developed a technique in which a weighted least squares algorithm was used instead of an ordinary one. In addition, multiple templates were used simultaneously to scan the data to increase the accuracy of the fit between the data and templates. An important outcome of the weighted template technique is that the detection rate is not as sensitive to the length of templates and this technique is capable of detection overlapping events reliably.