Optical Design And Applicaitons For Point-of-care Testing System

With the continuous improvement of health conditions and the increasing preven-tion measures for various diseases,the cure rate for various diseases has also been in-creasing in recent years.However,for some diseases,such as bacterial infection and myocardial infarction,death rates are still high.At the same time,in many develop-ing countries and resource-limited areas,there is an extreme lack of expensive medical resources for both human and animal health.Thus,rapid and accurate technologies are urgently needed to obtain on-site diagnosis results for both medical emergencies such as heart attacks,and in low-resource settings such as diagnosing parasite infec-tion in rural livestock.The development of point-of-care testing(POCT),combining optical,chemical,biosensor,and paper-based microfluidic technologies,among others,provides a good solution to this problem,especially for multiparametric testing.Studies have shown that current POCT diagnostics have a wide range of applications in different fields,such as the detection of key biomarkers in myocardial infarction;the detection of glucose in the blood,etc.Research has shown that the bottlenecks in the application of POCT are as follows:fluorescent labeling makes sample preparation more compli-cated,smart phones cannot achieve large-field high-resolution imaging,low diagnostic sensitivity,and a variety of optical designs(the performance is not the best,and the inter-system cannot be compared)etc.The research described:in this thesis focuses mainly on the optical portion of POCT,exploring two applications(parasitic infection and myocardial infarction)based on classic 4F optical systems,aiming to simplify the volume of the instrument while optimizing the detection accuracy through optimizing the optical design,and combining the design with automatic control and image process-ing to obtain high sensitivity without dependence on trained users.In the first application,we explore parasitic infections of livestock.Manual hand counting of parasites in fecal samples is the gold standard for diagnosis,but requires costly components and substantial expertise,limiting its use in resource-constrained set-tings and encouraging overuse of prophylactic medication.To address this issue,a cost-effective,automated microscope-based parasite diagnostic system that does not require special sample preparation or a trained user was developed.Fecal samples prepared using the McMaster flotation method were imaged,with the imaging region compris-ing the entire McMaster chamber.It is composed of an inexpensive,portable,robotic microscope that can scan over the size of an entire McMaster chamber(100 mm2)to cap-ture bright field images without need for user intervention.The microscope is a simple 4F optical system with a sufficiently high lateral resolution(1 ?m)for imaging parasite morphology.It can scan the entire counting chamber(100 mm2)to obtain a bright field map.The overall microscope cost is controlled to be about US $ 350.The images are au-tomatically segmented and analyzed using a trained convolution neural network(CNN)to robustly separate eggs from background debris.Simple post-processing of the CNN output yields both egg species and egg counts.The system was validated by compar-ing accuracy with hand-counts by a trained operator,with excellent performance.As a further demonstration of utility,the system was used to conveniently quantify drug response over time in a single animal,showing residual disease due to Anthelmintic resistance after 2 weeks.In researching the above system,we identified one key area for improvement.For a low-cost microscope,the field of view is limited by field-dependent aberrations in the low-cost lenses.Further,due to limited manufacturing precision of low-cost sam-ple chambers,the sample must be auto-focused using a highly precise(and thus costly)translation stage.Both of these issues arise from the limited depth of field of tradi-tional microscopes.To comprehensively address these issues,we designed a pupil-modification to the optical system to alter the phase of the collected light and extend the depth of field of the system.After our verification,this method can achieve about 10x improvement(that is,about 200 ?m)depth of field extension in the fluorescence channel,which can fully meet the focus issues brought by aberrations or manufactur-ing errors,drastically simplifying the instrument design and experimental procedure.Imaging can be performed over a larger field-of-view without additional z-direction ad-justment.In the bright field of view,a 10x improvement cannot be achieved,but if the pupil surface can be precisely modulated,it can also be 5x improved.In the later application,it can also focus on flow cytometry detection based on microfluidic chips without complex flow-focusing,which will greatly reduce the volume and cost of these systems.Finally,we began to explore the application effects of the 4F system in non-imaging applications,such as readout of light from a luminescent assay.Based on the previous system,we aimed optimize the sensitivity of the detection of key biomark-ers for myocardial infarction using chemiluminescence(CL)and electrochemilumines-cence(ECL),from the perspective of improved optical design.CL and ECL are often used in such assays due to their convenience that they do not require light sources or other components to complicate or add cost to the system,and the low limits of detection due to the lack of background light signal.Reports of these assays for POCT applica-tions often include readers built on a cellphone platform or constructed from low-cost components.Although the detection systems utilize a variety of optical designs,includ-ing 4F optical systems,the impact the optical design has on the limit of detection in these systems has not been explored.Here,based on the research of the 4F optical system our the prior work,our team reports a theoretical rubric to evaluate different optical designs in terms of maximizing the use of photons emitted from a CL or ECL assay to improve the limit of detection.We found that in measurement of CL assay,the requirements for the optical system are totally different from the microscope.Both systems prioritize collecting photons efficiently,but while microscopes are primarily concerned with res-olution(such that photons must be spread out among many pixels to adequately sample the image),luminescent assay detectors are more concerned with light concentration(i.e.photons should be compressed into only a few pixels).Based on the theoretical rubric we design a new portable reader built using off-the-shelf condenser optics,and demonstrate a nearly 10x performance enhancement compared to prior reports on an ECL assays running on a portable chip.In summary,we use optical design methods,combined with automatic control,image processing,deep learning and other methods,to improve the diagnostic stability and sensitivity of real-time detection of disease biomarkers,with our instruments having a broad application prospect.