MOLECULAR QUANTUM AND LOGIC PROCESS OF CONSCIOUSNESS—VITAMIN D BIG-DATA IN COVID-19—A CASE FOR INCORPORATING MACHINE LEARNING IN MEDICINE

10.5281/zenodo.10435649

ABSTRACT
There are unique relations between statistics and logic operating in neural networks in the human brain. Like 
machine learning, big data and artificial intelligence (AI) techniques could emulate human brain functions rapidly and assist healthcare decisions, especially in emergencies like COVID-19. We explore published positive 
modulatory effects of vitamin D on the human immune system from randomized control clinical trials (RCTs) and meta-analyses taken as an example. The analysis confirmed a robust negative correlation between serum 25(OH)D concentration with increased susceptibility to symptomatic SARS-CoV-2, complications, hospitalization, and deaths. Instead of using available such data in mid-2020, regulators relied upon RCTs from big pharma. Utilizing advanced Machine Learning paradigms using cleaned data would have expedited proper decision-making, better guidance on patient management, and approval of cost-effective early therapies like vitamin D and calcifediol, reducing the cost of care and millions of hospitalizations and deaths from SARS-CoV-2. Using Catuskoti logic in AI, broader than Boolean logic would have enhanced unbiased decision-making and developed data-driven algorithms to control outbreaks. Scientific evidence existed in mid-2020 that vitamin D and calcifediol rapidly boost the immune system, thus preventing SARS-CoV-2-related complications and deaths. However, statistical misconceptions, lack of broader vision, and failure to use big data analysis and machine learning approaches prevented using generic agents as prophylactic and adjunct therapies. We present the argument for hypothesis generation in conjunction with innovation in machine learning, using Catuskoti-based XOR and XNOR circuits as a solution to expedite the categorization of vulnerability, developing practical algorithms, and rapid approvals of repurposed drugs for future pandemics. Such data-driven analyses through machine learning programs minimize conflicts of interest and expedite decision-making. In the future, big data can be analyzed using desktop computers, facilitating prompt and proper decision-making and expedited drug approvals, especially for generics. This approach will better manage future epidemics and pandemics—cost-effective early therapeutic interventions, preventing complications, hospitalizations, and deaths, and reducing healthcare burden and cost.