Medical Knowledge Computation System
Applying Quantum Mechanics to Knowledge Discovery From Uncertainty
DIFFERENTIAL DIAGNOSIS
BRIDGING MEDICAL BEST PRACTICE (PICO) + EVIDENCE BASED MEDICINE (EBM) + VERY LARGE DATA SCIENCE
PICO Method - Medical Best Practice
Overcoming vagueness in approach by formulating crisper, more answerable questions, particularly by applying the PICO method. It is a popular EBM theme and basically a way of helping physicians formulating a kind of complex query that satisfies basic principles of EBM. It is not in itself the answer although it is intended to reach a more relevant answer through focus on essential elements.
Automated Systematic Review
Systematic Recovery of Best Evidence applies to answering PICO questions, but when topics to be investigated are fairly extensive one is said to perform Systematic Review of the evidence available in the literature and on the Internet. Along with the related idea of meta-analysis, integrating the findings from several randomized clinical trials involving the same or similar drug provides the classic example. This is one reason for retaining the general RCT form for more general application. A Systematic Review includes consideration of external validity or generalizability of the research addressed and its relevance in the immediate context of concern.
Automated Assessment of Quality - Diagnosis
Exhaustive computations ensure that the quality of the evidence extracted meets the highest quality standards. The computations result into automated systematic review. It involves conducting overarching bio-statistical analysis leading into coherence tests that includes Bayesian rules compliance, accuracy, sensitivity, specificity, relative risks, predictive odds etc.
The Bioingine - Ensemble of Machine Learning Algorithms
Hyperbolic Dirac Net Based Deep Learning System - Big Data To Knowledge Discovery



Traditional approaches in AI have been considered weak methods. A recent review of the more modern development of AI argued (a) that humans learn as children essentially by the combination of top-down and bottom-up methods that AI has imitated and (b) that benefit from and even need large amount of input information. 

By “top down” is meant the kind of inference based on knowledge that is prepared in advance, essentially in a form still readily recognizable by humans, and often called or miscalled “Bayesian” inference. “Bottom up” approaches use artificial neural networks that “are computational models, loosely inspired by biological neural networks, consisting of interconnected groups of artificial neurons which process information using a connectionist approach” based on learning weights. Methods used in The BioIngine, by gathering knowledge from structured data and unstructured data mainly reflect “top down” approaches. However, some have significant bottom up character, as well.

Fraud Waste and Abuse Solution
Ensemble of Algorithms - Claims Anomalies Processing - Detection, Prevention and Elimination
SMASH
Detects Commonness of patterns of diagnoses, procedures and other data entries by conducting the data mining with machine learning resulting into a automatic inference net construction in a Top-Down Bayesian kind of inference is the inference method used, although the approach is much broader in The BioIngine. It uses the Hyperbolic Dirac Net (HDN) based on “super-probabilities” or “probability amplitudes” which uses dual-valued hyperbolic-complex probabilities. These are probabilities with an imaginary part proportional to h, rediscovered and applied to probabilistic reasoning in quantum mechanics for particle physics by Nobel Laureate Paul A. M. Dirac, the hidden square root of 1 such that hh = +1.
ALERT
Detects the adjustable similarity distance metric between (a) each record and (b) the rest of the records collectively by conducting the data mining and automated machine learning that is more of the bottom up in approach. Its symmetric comparison of all records with all or most other records is mathematically consistent with the inference approach behind SMASH (because in such symmetric cases the hyperbolic-complex probability reduces to a real number). Nonetheless it has elements of a neural network approach in which implied weights are assigned by a learning process to different aspects of the notion of similarity (and hence dissimilarity).
BILL
Predicts effective contextual costs of items and notes when claim is excessive by conducting data mining and automated machine learning in a bottom-up way, but here the weights relate more specifically to the effective costs of items on a claim, in context. It could also be considered to be a method of solving linear equations when matrix methods such as Cramer’s Rule would “blow up” because of a high degree of departure from additivity.
Population Health - The Big Data into Knowledge Discovery
Applying Quantum Mechanics to Knowledge Discovery From Uncertainty

To provide means to tackle health and the population health system as a physical system like an engine, but in particular in terms of information flow as mutual information and associated probabilities, a systems information-theoretic view* of an engine that  we call an ingine.
To transform Big Data into Probabilistic Knowledge elements make the data practical and useful so it can inform local health outcomesfor populations and individuals, and determinants of it,using the Hyperbolic Dirac Net (HDN) method.
To describe in overview application terms use of these elements in a knowledge representation network  as an inference network , so that detected relationships can be described  and quantified in terms of their predictive power in computation, i.e. in terms of a degree of utility in addressing  determinants, outcomes, equity etc.
To facilitate studies such as Complex Adaptive System to explore and evaluate the transformation necessary to promote population health as a consequence of the various interacting components of the health system.