Abstract Most processes in an IVF laboratory are performed manually by technicians and embryologists, such as tracking and prepare dishes, pipetting of gametes and embryos from one dish to another, loading freezing straws, and operating micromanipulators for ICSI or biopsy. In most centres doctors manage hormonal stimulation based on their own learned experience or at most of their team. This results in significant variability in pregnancy outcomes from centre to centre and even operator to operator. For example, centres can be compared based on egg donor cycles, which are all young and fertile patients selected in similar ways in most centres. But even In egg donor cycles, embryo aneuploidy rates vary 20% to 60% (1) while pregnancy rates fluctuate from 10% to 80% depending on the clinic (2). Furthermore, although large amounts of data are generated from >2M procedures of IVF a year worldwide, most of it is siloed in each centre in a myriad of EMRs or worse, spreadsheets, impossible to query globally. The lack of a massive dataset to query and the absence of a constant baseline in pregnancy outcomes makes introducing new innovations harder to determine if they work or not. Automation should allow constant results with similar results clinic to clinic, especially desirable for networks to offer a consistent service across centres. By reducing the reliance on highly skilled manual labour, which is facing a global scarcity, automation has the potential to offer more people affordable treatment. Moreover, the utilization of automation generates vast amounts of data, with machines sharing information to inform medical decisions based on millions of standardized treatments, surpassing the insights of individual physicians. This data-driven approach also establishes a stable foundation for innovation by minimizing human variables. Advancements in Affordable Robotics, coupled with high precision and miniaturization, alongside the evolution of software, are transforming robots into versatile tools akin to smartphones. This transformation encompasses various fields, including microfluidics, miniaturized imaging systems, Artificial Intelligence, and the Internet of Things (IoT). Some automation has already occurred, such time-lapse incubators with imaging allowing for AI classification of embryos, or witness systems. Although historical lack of government funding and current trends of consolidation of clinics into networks, may slow innovation, numerous startup companies are actively developing automation solutions for IVF laboratory operations and patient management. These solutions range from tools for remote patient monitoring to automating processes such as egg denudation, sperm analysis, oocyte and embryo vitrification and warming, as well as procedures like automated ICSI and in vitro fertilization. Furthermore, artificial intelligence is being harnessed for tasks like embryo selection and hormonal stimulation, with groundbreaking concepts like artificial uterus also under exploration. Nonetheless, the objective of automation should be to enhance the abilities of the embryologist, not to reduce their need. Once the process has been standardized, it should pave the way for the creation of novel and more interesting tasks. This empowers the embryologist to engage in higher complexity or more high touch tasks, rather than merely replacing labor.
Search from 250 M+ research papersSearch from 250 M+ research papers
Sep 18, 2024
Elizabeth Chloe Romanis + 2
Open Access