The epidemic reflects the direction of the new era of medicine - the layout of AI has become a top priority

One of the more commonly used words in the English- language world is "pandemic", a term that, to some extent, reflects the cruel nature of a public health disease because of its percesive infectious nature, whether it's the cost of Wuhan's closure or the 2 million people infected in the United States.
how humans can stop the spread of the virus or disease, which only a few countries, such as China, can do, but the answer is obvious - run faster than the virus.
testing 9 million people in Qingdao in five days is the best proof of this answer.
but testing may be more testing the supply chain, and the pharmaceutical industry could not have given a matching answer in terms of treatment or prevention.
1 as a drug, no vaccine in the past could be developed in months or even a year or two, counting the fastest recorded four years of mumps vaccine, the average time for vaccine development is ten years.
can you imagine a decade before humans can use the new crown vaccine? The fact is that the country's new crown vaccine is infinitely close to the market, without skipping any routine research and development steps.
And other therapeutic drugs, such as the neutral antibody that cures Trump, are moving forward at an exhilarating pace, and these unprecedented achievements, in addition to advances in pharmaceutical technology, are also a credit to AI artificial intelligence, which we often hear about.
AI application in the field of biomedicine In the rapid response to the new crown outbreak, AI is mainly involved in the following work: the use of AI to analyze the potential efficacy of the original compound library for the virus, similar to the chip process is facing the limits, humans almost to some extent, exhausted the compound molecules of the drug possible, but wan Ai can never be well-named examples tell us that traditional methods can never quickly and accurately grasp the real therapeutic effects of these molecules, but AI in its limited time, to achieve the most likely work in the sea needle, the most likely results of research and development.
In addition to the analysis of the original library, AI can also independently design antibody molecules that may target the virus, Wuhan outbreak of the critical moment, has been cured patients will donate plasma with antibodies to help fight the epidemic.
the virus that helped Mr. Trump recover within 3-5 days, is this artificial antibody.
Humans have mastered the manufacturing techniques that don't require extracting antibodies from recovering patients, but how to design and screen the most effective antibodies is a huge project -- an AI that can process huge amounts of data instantaneously and evolve to learn.
but these are just the tip of the iceberg for AI's use in medicine, and AI can make a difference almost entirely in the pharmaceutical sector.
Figure 2 Drug discovery and design from new molecular design to new bio-target identification, including target-based, ideographic and multi-target drug discovery, drug use and biomarker identification, AI plays an important role in drug target identification and validation.
the potential of AI, especially during drug trials, to reduce the time it takes for drugs to be approved and marketed.
this can save a lot of money, which means lower drug costs for patients and more treatment options.
For example, drug researchers can identify and validate new cancer drug targets using longitudinal electronic medical records (EMR records), next-generation sequencing, and other "histological data" to create representative models for individual patients.
Figure 3. Biomedical and clinical data processing in the medical system for macular degeneration and diabetic macular edema developed by AI to date, the most developed use of AI should be to read, group and interpret large amounts of text data using algorithms.
it provides a more efficient way to validate or discard assumptions by examining large amounts of data in a growing number of research publications.
this application can save a lot of time for researchers in the life sciences.
addition, many clinical studies still rely on paper records, in the form of patients keeping a paper diary of when they took the drug, other medications taken, and adverse reactions that occurred.
can be collected and interpreted by AI, whether it's handwritten notes, test results, environmental factors, and imaging scans.
advantages of this approach to AI, faster data research and cross-references, and the ability to combine and extract data into available formats for analysis.
Figure 4. Compared to existing computer analysis programs (centre), the new artificial intelligence (right) can more accurately locate tumors (Photo: Google) A Cognizant study shows that about 80 per cent of clinical trials fail to meet the group schedule, and that one-third of all terminated Phase III clinical trials are terminated due to group difficulties.
Rare diseases and personalized medications combine AI with information about human scans, patient biology and analysis to detect diseases such as cancer in a variety of applications, and even predict the health problems they may face based on the patient's genetic information.
ibm Watson, which is used in oncology, is an example of how each patient's medical information and medical history can be used to recommend personalized treatment options.
based on individual test results, responses to past medications, and historical patient data on drug responses, AI was able to develop personalized drug treatment options.
in the pharmaceutical industry is to find participants for trials, in addition to helping to understand clinical trial data.
using advanced predictive analysis, AI can analyze genetic information to identify suitable populations of patients participating in the trial and determine the optimal sample size.
AI technology can read free-form text and unstructured data, such as doctor's notes and ingestion documents, that patients enter into clinical trial applications.
: 86% of clinical trials fail to recruit enough patients.
this leads to slower research and delays patients' access to life-saving drugs.
artificial intelligence application, which predicts treatment outcomes, saves more time and money by matching drug interventions to individual patients, reducing previous work involving trial and error.
Machine learning models can predict a patient's response to possible drug treatment by reasoning about potential relationships between factors that may affect the outcome, such as the body's ability to absorb compounds, their distribution in the body, and the body's metabolic capacity.
in addition to medical diagnosis, the development of biomarkers is also an important task in the process of drug discovery and development.
, for example, use predictive biomarkers to identify potential responders to molecular target therapy before conducting drug testing in humans.
AI uses biomarker models "trained" by large data sets in the process.
drug reuse is expected to be one of the most direct areas where AI-based technologies can bring significant value to budget-facing pharmaceutical companies.
For many biopharmaceutical companies, re-using previously known drugs or late-stage candidates in new therapeutic areas is an ideal strategy because the application has a lower risk of unintended toxicity or side effects in human trials and may reduce research and development spending.
drug compliance and dosage are a huge problem for pharmaceutical companies, ensuring that voluntary participants in clinical studies comply with drug research programs.
patients in the drug study do not follow the trial rules, they must be removed from the study, which may affect the results of the drug study.
important factor in successful drug trials is to ensure that participants take the required dose of the study drug at a specified time.
that's why it's so important to make sure you're drug-dependent.
using remote monitoring and test results evaluation algorithms, AI can select patients who are highly dependent and suitable for participation in the study from all patients.
leap-up production process using existing data analysis techniques based on multiple analysis and predictive analysis is the most important application scenario in process development.
AI offers many opportunities for process improvement in the development and production processes.
AI can perform quality control, reduce design time, reduce material waste, increase production reuse, perform predictive maintenance, and more.
AIS are used in a number of ways to increase productivity, produce faster, and waste less.
, for example, CNC (Computer CNC) can be used to complete process data entry or management that typically relies on human intervention.
AI machine learning algorithms not only ensure very precise execution of tasks, but also analyze processes to find areas where tasks can be simplified, reducing material waste, speeding production, and meeting key quality attributes (CQA) more consistently.
We all know that process development is crucial in drug production, whether for protein antibody drugs, CAR-T cell therapy, vaccines, etc., and that a stable and effective process is a prerequisite for ensuring the production of qualified drugs.
design of the process is time-and-effort and is often repeatedly validated between retracts.
use of advanced data analysis can accelerate the process of research and development, saving resources.
For example, experimental design (DOE) can obtain maximum information with minimal experimental volume, and further analysis of experimental data can establish design space to increase understanding and further optimization of the process, thereby realizing regulatory requirements (QbD).
In biopharmaceutical process development, the core role of DOE is becoming more obvious and gradually widely used (e.g. cell strain screening, media formulation, upstream and downstream operating conditions setting, drug drying, sterile filling, etc.).
addition, in the production process, real-time data acquisition, analysis capabilities, become the key to affect cost, quality, and even batch pass rate, but also to affect almost all biopharmaceutical enterprises and CMO key factors.
strict requirements for the quality of medicines make the production process more and more standardized.
the use of modern equipment and instruments, such as Sedtoris' PAT process analysis technology, makes real-time release detection (RTRT) possible.
pharmaceutical production process is often the pursuit of golden batch: a repeatable process that consistently optimizes yield and quality for real-time release of medicines.
multivariate analysis technology (MVDA) can help create the trajectory of gold batches and monitor production processes in real time, real-time diagnostics, real-time forecasting and forecasting, and real-time control.
use of data-driven technologies has resulted in significant cost savings and significant benefits for drug manufacturers.
leading the way in this regard, including manufacturers such as Sedtoris, which specializes in production process development, already have their own process analysis systems.
in their super-factories around the world, these systems are operating with high precision and efficiency that were previously unsoable.
example, every company involved in the development of the new crown vaccine has promised to produce hundreds of millions of vaccines a year, a figure previously unthinkable.
practitioners know that a single pollution, it may take more than a month to carry out a comprehensive treatment, whether the batch is qualified or not is only after the completion of the production of the testing link can be known.
and the probability of these problems occurring in the super-factory has been minimized. Outsiders and forerunners in
say AI, mostly to Internet technology companies, including AlphaGo, which defeated Google, but since 2017 international pharmaceutical giants have deployed their own AI systems to improve the efficiency of new drugs such as Novatis, Roche, Pfizer, Johnson and Johnson.
on September 25, 2020, it was announced that Sedulis had joined the ranks of DFKI shareholders in the German AI research centre, and that Sedulis had become the first life sciences company in its shareholder group.
" German Center for Artificial Intelligence Research (DFKI) was founded in 1988 and focuses on major industries of artificial intelligence, including big data analytics, knowledge management, picture processing and understanding and natural language processing, human-computer interaction, and robotics.
the transformation from research to practical application, a large number of industrial achievements have been formed in the past 30 years.
created 2,500 new jobs in the IT industry, created more than 70 spin-off companies and is the world's largest non-profit AI research organization, with shareholders in the world's top 10 technology companies, including Google, Intel, Microsoft, BMW, SAP and Airbus.
, Seddris had established a Joint Research Laboratory (SAIL) with DFKI to develop cutting-edge tools and artificial intelligence applications for the production of advanced drugs.
Joint Research Laboratory (SAIL) is currently working on how to apply deep learning algorithms and methods to image recognition of cells and organisms, analysis and modeling of biological systems, and simulation and optimization of biopharmaceutical production processes.
ISAIL is expanding further, including a special "wet lab" that combines and tests new ARTIFICIAL processes directly with cellular and molecular biology experiments.
as an independent laboratory, SAIL and protected data space for Sedulis' partners and customers.
Conclusion AI application has been more and more deeply into all areas of biomedicine, in the face of complex systems and complex problems, AI is the best solution at this stage, we are pleased to see that the biomedicine field such as Sedulis, the mainstay of the field, is actively embracing emerging technologies, and strive to match their own layout, such a big game, to biomedicine next stage of development set the tone and direction.