UTSAAH Lab, Center for Product Design and Manufacturing, Indian Institute of Science, Bangalore, India.
J Diabetes Sci Technol. 2024 May;18(3):618-624. doi: 10.1177/19322968221115749. Epub 2022 Aug 5.
A wide suite of methods are available to evaluate delivery accuracy of insulin pumps. However, these methods do not capture any temporal information, which may be critical for design of artificial pancreas (AP) systems. We propose a novel video microscopy method to understand the delivery accuracy and temporal nature for a new durable pump under development (IFP), and a commercially available pump (Medtronic 722G, M722G).
The cannula tip of an infusion set is inserted into a graduated pipette placed under a digital microscope. A video of the delivery is captured to track the fluid meniscus, to measure volumetric delivery rate and accuracy. This was done for a programmed value of 0.5 and 1 U. A similar procedure was adopted to track linear motion of the piston rod, which actuates the reservoir plunger, for a programmed value of 10 U.
It was observed that the commercially available pump delivers insulin in pulses of 0.05 U every two seconds. The mean absolute volumetric delivery error (MAE) for both pumps was found to be within the values reported previously. More importantly, it was found that a significant fraction of the programmed value is delivered, after completion of the planned bolus duration (IFP: 14.31% vs M722G: 9.38% for 1 U delivery).
The methods presented in this article help understand the delivery dynamics of liquid drug delivery devices. Our results indicate that a significant fraction of insulin delivery happens after the planned bolus duration, which might be important consideration for design of AP systems.
有许多方法可用于评估胰岛素泵的输送准确性。然而,这些方法没有捕捉到任何时间信息,这对于人工胰腺 (AP) 系统的设计可能至关重要。我们提出了一种新的视频显微镜方法,以了解正在开发的新型耐用泵(IFP)和市售泵(美敦力 722G,M722G)的输送准确性和时间特性。
将输注套件的管尖插入放置在数字显微镜下的刻度吸管中。拍摄输送过程的视频以跟踪流体弯月面,以测量体积输送率和准确性。对于编程值为 0.5 和 1 U 进行了此操作。采用类似的程序来跟踪驱动储液器柱塞的活塞杆的线性运动,对于编程值为 10 U 进行了此操作。
观察到市售泵以每两秒 0.05 U 的脉冲输送胰岛素。发现两种泵的平均绝对体积输送误差(MAE)均在先前报道的值范围内。更重要的是,发现计划的推注持续时间完成后,输送了相当一部分的编程值(IFP:14.31%比 M722G:1 U 输送时的 9.38%)。
本文介绍的方法有助于了解液体药物输送装置的输送动力学。我们的结果表明,在计划的推注持续时间之后,胰岛素输送发生了相当大的一部分,这对于 AP 系统的设计可能是一个重要的考虑因素。