1. Date of download: 9/18/2016 Copyright © ASME. All rights reserved. From: Development of a Non-Blood Contacting Cardiac Assist and Support Device: An In Vivo Proof of Concept Study J. Med. Devices. 2011;5(4):041007-041007-9. doi:10.1115/1.4005281 Normal, null, and inverted curvature in apex-to-base, radial plane (long axis) Figure Legend:

2. Date of download: 9/18/2016 Copyright © ASME. All rights reserved. From: Development of a Non-Blood Contacting Cardiac Assist and Support Device: An In Vivo Proof of Concept Study J. Med. Devices. 2011;5(4):041007-041007-9. doi:10.1115/1.4005281 Contour comparison of device cross-section when fully activated. Note curvature inversion of the Anstadt Cup (left), which is by design as illustrated in Fig. of the Anstadt patent (US 5,119,804) and more recently in Anstadt []. Ejection fraction is superb with both devices, but the Criscione device (right) achieves restoration of circulation without inverting the curvature. Figure Legend:

3. Date of download: 9/18/2016 Copyright © ASME. All rights reserved. From: Development of a Non-Blood Contacting Cardiac Assist and Support Device: An In Vivo Proof of Concept Study J. Med. Devices. 2011;5(4):041007-041007-9. doi:10.1115/1.4005281 Criscione Device Drawing. The device was constructed with six independent chambers in a spiral orientation. The chambers were sutured to a nitinol scaffold, which was mounted on hub at the base of the device. Figure Legend:

4. Date of download: 9/18/2016 Copyright © ASME. All rights reserved. From: Development of a Non-Blood Contacting Cardiac Assist and Support Device: An In Vivo Proof of Concept Study J. Med. Devices. 2011;5(4):041007-041007-9. doi:10.1115/1.4005281 The implanted device. A median sternotomy was performed and the device placed over the cardiac apex. The device drivelines and chest tube were routed caudal to the sternum. The sternum was closed with wire and the fascia was closed tightly with suture to create a pneumatic seal. Figure Legend:

5. Date of download: 9/18/2016 Copyright © ASME. All rights reserved. From: Development of a Non-Blood Contacting Cardiac Assist and Support Device: An In Vivo Proof of Concept Study J. Med. Devices. 2011;5(4):041007-041007-9. doi:10.1115/1.4005281 Cross-section of the Criscione device with chambers in the activated and deactivated state. Note that the device is designed to avoid device induced aberrant curvature inversion of the heart when assist is applied. Figure Legend:

6. Date of download: 9/18/2016 Copyright © ASME. All rights reserved. From: Development of a Non-Blood Contacting Cardiac Assist and Support Device: An In Vivo Proof of Concept Study J. Med. Devices. 2011;5(4):041007-041007-9. doi:10.1115/1.4005281 (a) Comparison of CO and SW in the healthy baseline, esmolol induced failure, and 60 mm Hg active assist in the esmolol induced failure model. Here it is evident that assist recovers much of the CO and SW that is lost in failure. (b) Overlay of pressure-volume loops for the healthy baseline and the 60 mm Hg assist in the esmolol failure model. Here it is evident that while assist produces cardiac performance metrics that are similar to the healthy baseline, the PV relationship has unique characteristics. Though SW and SV are similar in magnitude, the PV profile is altered. Note: SV = max LV volume – min LV volume, SW = area of the PV loop. Figure Legend:

7. Date of download: 9/18/2016 Copyright © ASME. All rights reserved. From: Development of a Non-Blood Contacting Cardiac Assist and Support Device: An In Vivo Proof of Concept Study J. Med. Devices. 2011;5(4):041007-041007-9. doi:10.1115/1.4005281 Comparison of PV Loops pre- and post-deployment. Note that the deployment of the device had no significant effect on cardiac function and performance as evidenced by the acquired PV data. The PV relationship is unaffected by device placement. Figure Legend:

8. Date of download: 9/18/2016 Copyright © ASME. All rights reserved. From: Development of a Non-Blood Contacting Cardiac Assist and Support Device: An In Vivo Proof of Concept Study J. Med. Devices. 2011;5(4):041007-041007-9. doi:10.1115/1.4005281 Application of 30 mm Hg active assist to the healthy heart. Note that though SV, EF, CO, and SW were significantly increased (p = 0.05), assist did not substantially alter the characteristic features of the PV relationship. Compare this result with the esmolol failure model, Fig., where application of 30 mm Hg active assist resulted in a different PV relationship due to the difference in cardiac function. Figure Legend:

9. Date of download: 9/18/2016 Copyright © ASME. All rights reserved. From: Development of a Non-Blood Contacting Cardiac Assist and Support Device: An In Vivo Proof of Concept Study J. Med. Devices. 2011;5(4):041007-041007-9. doi:10.1115/1.4005281 Fluoroscopic images of the device at end-diastole (left) and end-systole (right). Here the nitinol scaffold is visible as well as the segmented PV loop catheter. While the image quality is not sufficient for reliable quantitative analysis, visual inspection revealed no evidence of device induced aberrant curvature inversion of the heart when assist was applied. Figure Legend: