Fig. 11

Performance evaluation of nuclear classification at different threshold values and different development stages. a-c The graphs compare the performance of nuclear classification at different thresholds using following parameters i.e. accuracy (a), false positive rate (b) and false negative rate (c). Each series in the graph represents a different muscle cell. The black dotted line indicate the threshold value at which accuracy is highest and error rates are lowest. This threshold value is used for nuclear classification. d-f The graphs compare the performance of nuclear classification at different stages of pupal development using following parameters i.e. accuracy (d), false positive rate (e) and false negative rate (f). Large movements of internal nuclei results in low accuracy and high error during between +20 and +50 h. In comparison, accuracy is high after +50 h due to slow movement of internal nuclei