Journal of NeuroEngineering and Rehabilitation

included the Southampton Hand Assessment Protocol (SHAP), the Jebsen-Taylor Test of Hand Function, three repetitions of the Box and Blocks test, and three repeti- tions of the Clothespin Relocation task. These measures were chosen to evaluate hand, wrist, and elbow function and were activities that could be reasonably completed with a physical prosthesis. Individuals then took the device home for a minimum of 42 days (6 weeks) of home use. If the prosthesis needed to be repaired, or if the user had a valid, docu- mented reason for not wearing a myoelectric pros- thesis — e.g., an extreme sports competitions, a beach vacation, being sunburned — then additional time was added to ensure at least 6 weeks of usage. The controller logged the amount of time that the prosthesis was pow- ered on and the number of times the patient recalibrated the control system. After completion of the home trial, subjects returned to the laboratory and repeated the vir- tual and physical outcome measures described above (post-trial testing), in addition to completing the ACMC. For the SHAP and the classification error rate mea- sures, where only a single pre- and post-trial score was available, a one-tailed paired = t-test was used to com- pare differences. For other outcome measures where multiple trials were recorded, a repeated measures ANOVA with the subject as random factor, and pre/post trial condition and trial number were fixed factors. A correlation analysis using the Pearson coefficient was used to determine relationships between virtual and physical outcome measures. Results All subjects wore the device at home, and could success- fully recalibrate the device (Table 2 ) . Subject P2 typically removed the prosthesis while it was still powered on, and thus it was not possible to accurately determine wear-time for this subject. The occasional recalibration failures observed were primarily due to broken wires. Table 2 Prosthesis usage during home trial Patient Number of successful/ attempted PGT sessions Total number of days worn Total wear time (hrs) P1 7/7 9 45 P2 39/39 18 – P3 73/77 41 181 P4 56/57 58 365 P5 10/10 36 88 P6 20/20 14 28 P7 18/18 20 127 P8 38/38 28 69 P9 60/60 32 88 Designates statistical significance at the p < 0.05 level. CRT Clothespin Relocation Test. PGT Prosthesis Guided Training Fig. 1 Representative data from a prosthesis-guided training sequence. Data labels are provided by prosthesis movement; the resulting EMG patterns are used to train a pattern recognition system as described by Kuiken et al. [ 16 ] Hargrove et al. Journal of NeuroEngineering and Rehabilitation 2018, 15 (Suppl 1):60 Page 24 of 72