6.19 Evidence-based evidence-based medicine is a macro-level sequence operation, which is a sequence of special combinations with small probabilities, that is, we need to pay more attention to the matching of sequences, i.e., information searching. Because in essence, medical methods are to seek the matching of sequences at different levels so that the overall score is the highest and the efficacy is the best, experience is the basis for our Bayesian formula calculation, but experience may be wrong, which requires more scale experiments to verify. Of course, this is a validation of the large-scale hierarchy, and our selective expression for unique individuals may be a unique pathway, i.e., individualized treatment. In this process, we need various indicators/fixed points to determine the progress of a particular pathway.

We need to precisely target a specific type of population, even down to the individual level, which is a kind of operation on the sequence. In general, large-scale clinical trials are the calculation of a small number of sequences/effects at the macro level, and we hope to be able to determine the unique type of individual through the measurement of certain indicators, so as to have a more accurate grasp of the sequence, so as to formulate a plan to avoid various side effects to the greatest extent. Because in essence, a control experiment is a matching operation on a sequence, and the granularity of a general experiment is too large, and we need good luck to make the matching operation a relatively definite fixed point. Of course, in fact, our body has many special sequences of such target formulas, which can be regarded as special patterns. Large-scale randomized double-blind controlled trials are the most reliable and, of course, the least efficient.

Meta-analysis, the selection of a certain index of multiple experimental results for adjustment analysis, can extract a certain pattern, can obtain a certain correlation between sequences.

The sampling method is the confidence in the law of large numbers.

Correlation needs to be processed at the lower levels, otherwise there will be a positive correlation between the number of people who drowned and the rise in ice cream sales, and we need to dig into the underlying relationships.

The macro-level relationship of preventive medicine discovery is the hidden Markov sequence of specific clinical observations, that is, relatively high-dimensional changes, because diseases are essentially the result of the interaction between people and the environment, and the incidence of diseases at the macro level is reflected in the occurrence of diseases in microscopic individuals, and the latter contributes to the statistics of the former, which is a coupling relationship. This reflects the similarity of different levels, and the laws of the crowd are similar to the laws within the body.

The relationship between sequences can be expressed by regression equations, and a certain correlation can be formed between the patterns formed by a particular sequence, just like axiomatic system knowledge, that is, the sequence of a particular pattern formed by coupling at different levels can always form a relatively definite correlation with other specific patterns, such as formulas. Today's medicine is statistically approaching this system of axioms. Because there are many possible variables for different specific operation objects, which can be expressed in the form of sequences, we can only find the fixed point relationship between specific sequences through statistics, and the specific experiment is to process the specific sequences and discover their possible effects through certain indicators.

Cohort studies are methods of studying a single variable of sequences, which can only derive interactions between sequences on a small scale. Dealing with (not) - expressing (not) this is all the possibilities, so we can derive certain relations (p-values, etc.). However, clinical experiments are the expansion of low-level variables of this sequence to the network level of the organism, so in order to obtain a better relationship, it is necessary to select more macroscopic indicators and relationships to find a more certain linear relationship at the statistical level. For various conditions, the experimental factor is a part of the sequence, in which the level of its expression will have an impact on the expression of the overall sequence, which we can quantify in the form of a matrix, and the antagonism or synergy between different variables can be analyzed through orthogonal experiments. The relationship of a sequence can only be expressed in the form of probability. Based on this, new sequence relationships can then be derived to guide new drug treatments.

The coupling within the sequence and the coupling between the sequences, which is the specific sequence operation that can be associated with macroscopic disease occurrence. The specific etiology is a fixed point probability that is raised to a threshold. The operation of sequences is an abstraction of concrete disease development.

Essentially, illness is an expression of a relationship, so there is a certain probability that it will be able to heal itself. It may be possible to make the internal sequences of the body methylated like epigenetic through the treatment of various drugs, so that the probability of its expression changes, which may be helpful for the cure of diseases. Just as the input energy causes the body to be in a state of highly active disordered combination (such as atoms), which can finally anneal to form a relatively orderly result.

7.30.2016

Evidence-based medicine is our ideal to be able to find specific therapies that can have an impact on the health of the body at the macro level, through multiple levels of evidence, i.e., information, and is a bridge between basic research and clinical treatment. We can apply this idea to traverse from the relations discovered by the central law (low-dimensional) to the various treatments of the organism (high-dimensional), which can be expressed as macroscopic Bayesian probability calculations, which are large-scale calculations that can discover certain patterns (fixed points). The establishment of its theoretical system is a macroscopic classification mechanism, through the comparison of the group level (large-scale random double-blind experiment) to sort out a certain sequence, and then through the pattern recognition algorithm to sort out the possible patterns, and then through experiments to verify their correctness, as if the neural network can be used to continuously optimize. We believe that the doctor's diagnosis and treatment thinking is also a similar idea, first of all, the collection of various information, and then the matching experience to make a certain judgment (represented by sequence matching), the key in this process is to find the characteristic changes of the fixed point, so as to be able to make accurate judgments with a relatively high probability. Then you can take the appropriate approach to treatment, which is actually a sequential operation.

Therefore, we need to collapse this network level into a certain path, that is, select specific questions to do specific exploration, that is, tumors, diabetes, etc. Then, according to the efficacy of various evidence, a certain evaluation is made, and then a certain combination of therapies is considered to achieve specific practice. We are now trying to amplify the impact of clinical evidence-based practice to the molecular level, so as to integrate various operations and provide certain guidance for clinical practice.

To achieve this step, it is necessary to systematically conduct a comprehensive evaluation, i.e., based on large-scale information calculations, and then in a specific pathway, i.e., the obvious relationship between sequence parameters, such as the effect of A on B, which will have a probabilistic impact on a more macroscopic C, such as the effect of antiarrhythmic drugs on mortality after myocardial infarction (Furberg 1983).

The direction of multivariate influence, the pooled effect-size method, and the processing through various measures of statistical significance are high-dimensional operations, such as Fisher's pooled P-values. The result is a reduction in bias (like wave interference), which is a computation of a data structure such as a sequence. We intend to use meta-analyses to quantify the effects of various molecular level mechanisms on the health of the body. We need to measure a variety of metrics. The final effect synthesis can be seen not only as a Bayesian operation, but also as the path collapse of the network, like a cocktail therapy.

Sequence operation: The study object, the intervention, the control, and the outcome are all sequences, and their coupling is a kind of matching operation, and then the state of the sequence can be changed (the expression probability of the Markov sequence), like a Turing machine.

The range should be in a certain segment to have a more obvious probability connection, and the shorter the relevant segment of the sequence, the greater the correlation. This corresponds to various effect sizes such as proportion, variance. It can be continuously classified by fast Fourier transforms, and finally refined to a certain 1/0 sequence of basic objects. Then, like calculus, a high-dimensional structure is formed through traversal. Therefore, the comparison of the series needs can be derived from the measurement of certain indicators.

The exact problem is actually a kind of network path collapse, which is highly probable. For example, the effect of certain control measures on a particular type of disease is manifested in the different distribution of its results. Through the coupling of multiple mechanisms, we can make this evidence-based evidence-based medicine continue to expand the boundaries of human cognition as much as mathematics based on previous research.

The sequence formed by multiple variables is calculated as a new variable, which requires us to integrate the various effects of the previous variables, which requires us to adopt certain observation indicators. Then there's the hypothesis testing.