Deep Reinforcement Learning Model for Blood Bank Vehicle Routing Multi-Objective Optimization

The overall healthcare system has been prioritized within development top lists worldwide.Since many national populations are aging,combined with the availability of sophisticated medical treatments,healthcare expenditures are rapidly growing.Blood banks are a major component of any healthcare system,which store and provide the blood products needed for organ transplants,emergency medical treatments,and routine surgeries.Timely delivery of blood products is vital,especially in emergency settings.Hence,blood delivery process parameters such as safety and speed have received attention in the literature,as well as other parameters such as delivery cost.In this paper,delivery time and cost are modeled mathematically and marked as objective functions requiring simultaneous optimization.A solution is proposed based on Deep Reinforcement Learning(DRL)to address the formulated delivery functions as Multi-objective Optimization Problems(MOPs).The basic concept of the solution is to decompose the MOP into a scalar optimization sub-problems set,where each one of these sub-problems is modeled as a separate Neural Network(NN).The overall model parameters for each sub-problem are optimized based on a neighborhood parameter transfer and DRL training algorithm.The optimization step for the subproblems is undertaken collaboratively to optimize the overall model.Paretooptimal solutions can be directly obtained using the trained NN.Specifically,the multi-objective blood bank delivery problem is addressed in this research.Onemajor technical advantage of this approach is that once the trainedmodel is available,it can be scaled without the need formodel retraining.The scoring can be obtained directly using a straightforward computation of the NN layers in a limited time.The proposed technique provides a set of technical strength points such as the ability to generalize and solve rapidly compared to othermulti-objective optimizationmethods.The model was trained and tested on 5 major hospitals in Saudi Arabia’s Riyadh region,and the simulation results indicated that time and cost decreased by 35%and 30%,respectively.In particular,the proposed model outperformed other state-of-the-art MOP solutions such as Genetic Algorithms and Simulated Annealing.

A Drone-Based Blood Donation Approach Using an Ant Colony Optimization Algorithm

This article presents an optimized approach of mathematical techniques in themedical domain by manoeuvring the phenomenon of ant colony optimization algorithm(also known as ACO).A complete graph of blood banks and a path that covers all the blood banks without repeating any link is required by applying the Travelling Salesman Problem(often TSP).The wide use promises to accelerate and offers the opportunity to cultivate health care,particularly in remote or unmerited environments by shrinking lab testing reversal times,empowering just-in-time lifesaving medical supply.

Management strategy of prevention and control of nosocomial infection of COVID-19 in blood purification center

There are many patients in the blood purification center who need maintenance hemodialysis to maintain life. Those patients generally havelow resistance and are easily exposed to coronavirus because they go back and forth the hospital and residence three times a week andclosely contact with family, caregivers, community personnel, people in various means of transportation, medical staff, and other patientsvisiting hospital. Therefore, the blood purification center has become a high‑risk environment for the spread of COVID-19 infection. In viewof this, our center quickly responded to the formulation and implementation of infection prevention and control measures suitable for thecharacteristics of the blood purification center and continuous renal replacement therapy (CRRT) emergency plan for fever and suspectedpatients. According to these measures, we have a positive effect on preventing and controlling nosocomial infection in the blood purificationcenter.

Dextrose in the banked blood products does not seem to affect the blood glucose levels in patients undergoing liver transplantation

AIM: Hyperglycemia commonly seen in liver transplantation (LT) has often been attributed to the dextrose in the storage solution of blood transfusion products. The purpose of the study is to compare the changes of the blood glucose levels in transfused and non-transfused patients during LT. METHODS: A retrospective study on 60 biliary pediatric patients and 16 adult patients undergoing LT was carried out. Transfused pediatric patients were included in Group Ⅰ (GI), those not transfused in Group Ⅱ (GⅡ). Twelve adult patients were not given transfusion and assigned to Group Ⅲ (GⅢ); whereas, four adult patients who received massive transfusion were assigned to Group Ⅳ (GⅣ). The blood glucose levels, volume of blood transfused, and the volume of crystalloid infused were recorded, compared and analyzed. RESULTS: Results showed that the changes in bloodglucose levels during LT for both non-transfused and minimally transfused pediatric groups and non-transfused and massively-transfused adult groups were almost the same. CONCLUSION: We conclude that blood transfusion does not cause significant changes in the blood glucose levels inthis study.

Genetic screening of 5 blood group systems of Li ethnic group in Hainan province and establishment of rare blood group bank

Objective:To screen the blood group system genes of Duffy,Lutheran,Kidd,Diego,Dombrock blood group systems of Li ethnic group in Hainan Province and provide laboratory data for the rare blood group database in this area.Methods:The alleles of Duffy,Lutheran,Kidd,Diego,Dombrock blood group systems of 300 voluntary participants of Li ethnic group in Hainan were detected by sequence-specific primer polymerase chain reaction,and the polymorphism was analyzed.Results:The allele frequencies of Duffy,Lutheran,Kidd,Diego,Dombrock blood group systems of Li ethnic groups in Hainan Province are 0.9583 for Fy^(a),0.0417 for Fy^(b),0.8350 for Au^(a),0.1650 for Au^(b),0.4500 for Jk^(a),0.5500 for Jk^(b),0.0667 for Di^(a),0.9333 for Di^(b),0.1017 for Doa and 0.8983 for Dob,respectively.The antigen incompatibility rates of Fy^(a)/Fy^(b),Au^(a)/Au^(b),Jk^(a)/Jk^(b),Di^(a)/Di^(b),Doa/Dob of Duffy,Lutheran,Kidd,Diego,Dombrock blood group systems were 7.67%,23.76%,37.25%,11.67%and 16.60%,respectively.Conclusion:The gene frequencies of Duffy,Lutheran,Kidd,Diego,Dombrock blood group systems of Li ethnic group in Hainan Province are polymorphic,and the antigen incompatibility rates of alleles are higher,which is quite different from that of other nationalities in China and with unique ethnic distribution characteristics.It is of great significance to establish the rare blood group database in this region.

基于血站管理信息系统的业务全流程监督管理体系构建

目的 血站管理者可以实时对血站全流程关键业务运行态势进行全流程掌握,实现对血站业务全流程的监督管理,提升业务工作效率和保障血液质量。方法 从编制依据、血站业务范围、岗位监督管理内容、总体框架设计、管理系统和监督管理界面设计、物理数据库设计、程序开发和上线调试等多个方面,建立业务全流程监督管理体系,并将其与血站管理信息系统进行对接,及时、全面地记录从采血到供血全过程的业务流程关键指标信息,通过手机APP的方式进行展示和管理,于2023年投入血站使用后对2023年和2022年总采集量、总制备量、总供应量,以及检验不合格报废率和非检验不合格报废率(不含乳糜)进行统计分析。结果 建立了基于血站管理信息系统的业务全流程监督管理体系的手机APP,实现了“移动办公”式监督管理,2023年正式投入使用后,2023年总采集量、总制备量和总供应量均高于2022年,分别增长5.88%(13 247/225 454 U)、4.73%(24 156/510 698 U)、6.70%(34 814/519 914 U),检验报废率(0.54%,2 868/534 854 U)显著低于2022年(0.60%,3 047/510 698 U)(P<0.01),非检验报废率(不含乳糜)(0.12%,649/534 854 U)显著低于2022年(0.19%,991/510 698 U)(P<0.01)。结论 基于血站管理信息系统的业务全流程监督管理体系的构建,可以满足血站不同管理者不限时间和地点的标准化管理服务需求,实现对业务全流程的监督管理,还可以持续提升血站业务工作和血站管理的规范化、科学化水平,保障血液资源供应。

山东省血站血液质量监测指标体系的构建

目的构建山东省血站血液质量监测指标体系,以期持续提升血液质量和标准化管理。方法通过文献、标准等检索,结合采供血实际工作,以采供血业务过程关键控制点为导向,拟定血液质量监测指标体系。经2轮德尔菲法专家征询和论证,运用主观加权和层次分析法分析各指标权重,又经6个月的试行,完成对指标内容的修订和完善。结果构建了覆盖采供血全过程的血液质量监测指标体系,包括一级指标5项,分别为献血服务、成分制备、血液检测、血液供应和质量控制;二级指标72项,包含指标含义、计算公式等。2轮专家征询意见和2轮论证会议,修订指标17项,并通过主观加权和层次分析法获得指标权重,经试行的局部微调,形成了血液质量监测指标体系。结论首次构建覆盖采供血业务全过程的血液质量监测指标体系,可有效评价血站质量管理水平,为打造山东省血液质量管理“一盘棋”发展模式、提升标准化管理水平奠定基础。

山东省血站献血服务、成分制备及血液供应过程质量监测指标体系的应用

目的建立有效的献血服务、成分制备及血液供应过程质量指标监测体系,客观地评价血液采集、制备和供应过程的质量控制水平,促进血站采供血过程的持续优化和标准化管理。方法建立覆盖采供血全过程的献血服务、成分制备、血液检测、血液供应和质量控制的质量监测指标体系,向山东省17家血站发放《采供血过程质量监测指标统计表》,明确指标定义和计算公式,收集各血站2022年1—12月质量监测指标数据,利用SPSS25.0软件对其中的献血服务(10个)、成分制备(7个)和血液供应(4个)3个方面21个质量监测指标数据进行统计学分析,并比较了不同规模血站质量监测指标的差异。结果17家血站献血服务过程质量监测指标平均值分别为:定期献血者比例44.66%(2233/5000)、献血不良反应发生率0.22%(11/50)、非标量全血采集率0.46%(23/5000)、HBsAg初筛漏检率0.052%(13/25000)、血液采集一针率99.42%(4971/5000)、双份单采血小板采集率86.49%(173/200)、400 mL全血采集率66.50%(133/200)、献血者满意度99.25%(397/400)、带旁路留样系统全血采集袋使用率82.68%(2067/2500)、血液采集职业暴露发生数1例;定期献血者比例和400 mL全血采集率具有较强的正相关(P<0.05);大型血站双份单采血小板采集率、献血不良反应发生率和非标量全血采集率显著低于中、小型血站(P<0.05)。17家血站成分制备过程质量监测指标平均值分别为:血液成分制备血袋破损率0.03%(3/10000)、乳糜血报废率3.05%(61/2000)、溶血报废率0.13%(13/10000)、贴签错误发生数0.06件、血液导管渗漏数8袋、血液穿刺/接驳渗漏数2.76袋、耗材不合格数0.59件;大型血站溶血报废率显著低于中型和小型血站(P<0.05)、大型和中型血站乳糜血报废率显著低于小型血站(P<0.05)。17家血站血液供应过程质量监测指标平均值分别为:血液制剂过期报废率0.023%(23/100000)、血液制剂储存发放破损率0.009%(9/100000)、退血报废率0.106%(53/50000)、用血医院服务满意度99.16%(2479/2500)。不同血站之间血液制剂储存发放破损率与血液成分制备血袋破损率差异具有统计学意义(P<0.05)。在大、中和小型血站之间定期献血者比例、献血不良反应发生率、非标量全血采集率、400 mL全血采集率、双份单采血小板采集率,血液成分制备血袋破损率、乳糜血报废率、溶血报废率、血液制剂过期报废率、血液制剂储存发放破损率和退血报废率的差异具有统计学意义(P均<0.05)。结论山东省血站献血服务、成分制备及血液供应过程质量监测指标体系的建立具有较好的适用性、可行性和有效性,能够客观评价血站采供血过程质量管理水平,推进全省血液管理同质化和标准化发展,为今后开展血站综合评价奠定基础。