Aims of Study: Islet transplantation is a promising therapy for Type 1 diabetes. Islet mass is one of the parameters that determines transplant outcomes. In practice, islet mass is assessed by manual microscopic analysis, which is heavily operator-dependent and objective with significant inter- and intra-variability. One study indicated that islet mass of same sample by 36 operators had coefficients of variation over 20%. Digital imaging systems (DIA) analysis is an alternative, showing reduced variability. However, the DIA only measures islet mass and purity, but not other useful parameters. Additionally, it still needs human intervention leading to errors and bias that not meet cGMP requirement. The adaptation of smartphone in microfluidics has been explored for particle counting and analysis. Here, we are developing a simple microfluidic device with a high-power smartphone application for rapid, accurate, and automatic assessment of islet characteristics including IEq, purity, and other parameters.
Methods:  As shown in Fig.1 an iPhone 5S case was designed via AutoCAD and fabricated by 3D-Printing. Microfluidic channel (350 μm in length and 500 μm width) was fabricated using PDMS. A 4 X Aspheric lens was build onto the case to adjust optical aberration. Islets were loaded into microfluidic by syringe pump at 30 μl/min and recorded by iPhone 5S camera. Software development was under Matlab 2016a. Blob Analysis was used to distinguish the foreground and background pixels; Watershed transformation was used to detect islet objects and characterize islet properties. a final report was generated with all islet parameters and imaging of each islet.
Results: Human and rodent islet samples were evaluated, showing that the system is able to measure islet diameter, area, volume, purity, islet equivalents (IEQ), roundness, total islets count/IEQ, and fragmentation as shown in Fig.2. The final report was attached with each islet snapshots, allowing user to visualize islet quality beyond islet mass such as islet fragmentation, cell volume, and central necrosis. By comparing with manual counting, the system is more reliable and consistent (within 5% error rate).
Conclusions: The novel approach in microfluidic integrated with smartphone is consistent and reliable for providing fast and comprehensive evaluation of human islet quality and quantity. Furthermore, it can minimize the errors from manual counting. As Point of Care (POC) device. it allows physician to monitor human islet isolation process locally and remotely. The success application of this system will be significant for achieving Investigational New Drug (IND) application and a Biologics License Application (BLA) as human islets are defined as therapeutics.