In April of this year, American information technology and computing giant IBM (NYSE:IBM) announced that it would be applying the data analytics power of its Watson computing platform to the field of personal health by establishing IBM Watson Health. This cloud computing platform seeks to provide all organizations with a focus on health and wellness solutions, including insurers, physicians and researchers, with a tool for for drawing meaningful data from the one million gigabytes of health-related data that IBM estimates each person will generate in a lifetime. IBM Watson Health was bolstered by a couple of strategic acquisitions in the first half of this year, including Cleveland-based healthcare cloud provider Explorys and its massive clinical database as well as integrated population health management software developer Phytel of Dallas, TX.
Just a few months after establishing this health and wellness cloud platform, IBM has secured its first Foundational Life Sciences partner for Watson Health in Teva Pharmaceutical Industries (NYSE:TEVA) of Petah Tikva, Israel. Statements from senior Teva officials indicate that the world’s largest manufacturer of generic pharmaceuticals will use the personal health cloud to connect more directly with physicians and patients and enable individualized treatment optimization options. It’s hoped that the Big Data tools available through IBM’s Watson analytics will be powerful enough to help Teva and others come up with more effective treatments for the millions of people worldwide suffering from chronic health conditions such as migraine, asthma or neurodegenerative diseases.
More effective data analysis tools for the healthcare industry would be a major boon to insurers, drugmakers and many other companies in the field. For example, CVS Health Corp. (NYSE:CVS) is using data analytics and digital tools as part of the medication adherence program it’s developing to ensure that patients take prescribed medications. Some have argued that better data analysis tools in hospital settings would cut administrative costs, improve care coordination and reduce insurance fraud. IBM had already partnered with insurance provider WellPoint, now Anthem Inc. (NYSE:ANTX), to utilize Watson’s data analysis skills to provide doctors with treatment suggestions based on medical records and research databases. It’s pretty clear that there’s certainly a valuable market out there for IBM as it further develops the Watson Health cloud.
IBM’s Healthcare Analytics Patents: From Patient Medical Record Privacy to Remote Patient Diagnosis
A quick survey of IBM’s patents recently issued by the U.S. Patent and Trademark Office turned up nearly a dozen patents issued in 2015 related to patient health, which is not surprising given IBM’s stature as the world’s top earner of patents for more than two decades. Data analysis of medical records and the new health discoveries that could be drawn by them are at the forefront of U.S. Patent No. 9104786, which is titled Iterative Refinement of Cohorts Using Visual Exploration and Data Analytics. The method for defining a cohort comprised of multiple data objects protected here involves obtaining an initial cohort seeding, storing the cohort seeding in a memory, visualizing the cohort using a selected view to present a current cohort, reducing the cohort using visual filters, expanding the cohort using selected analytics which use computational techniques to identify additional objects for the cohort based on similarity input parameters and determining if the cohort should be further modified. This innovation provides an interactive system for visualizing patient cohort groups for the better analysis of medical and diagnostic histories of patients from similar demographics, supporting healthcare discovery. Those who might be a little worried by the idea of health companies being able to view very personal information might be appeased slightly by the privacy invention featured within U.S. Patent No. 9064033, entitled Intelligent Decision Support for Consent Management. This patent discloses a computer program product containing computer-readable program code configured to receive an access request for a record, identify information associated with the record and access request including the access request’s purpose, compute an importance measurement and a sensitivity measurement for the access request, compute a threshold value reflecting the probability of values drawn from declined requests and granted requests, compute a relevance measurement for the access request and compute a consent recommendation based on the previous access request measurements which includes a decision to grant or deny access. This system is designed to prevent the abuse of patient medical record data in cloud computing environments that could compromise patient privacy.
Both high-tech and low-tech methods for treating medical conditions are addressed by the technology protected through U.S. Patent No. 8968197, titled Directing a User to a Medical Resource. The computer program product for directing a user to a medical resource claimed here contains code executable by a processor to determine a user’s medical condition based on physiological sensor data and correlate the medical condition to a medical resource by identifying medical resources, determining the product provided by the medical resource, rejecting resources that cannot ameliorate the user’s medical condition and issuing directions to the user for the nearest medical resource; medical resources could be a nearby medical facility, a nutritional vendor or a park bench where a person who is experiencing excessive heart rate can sit and rest. This innovation would enable a patient wearing this physiological sensor system who is, for example, experiencing a cardiac event to find directions to a hospital or merely instruct the patient to sit down if paramedics aren’t needed.
The analysis of medical images is another important digital tool in the healthcare cloud and IBM has improved its IP holdings in that area with U.S. Patent No. 8949427, entitled Administering Medical Digital Images with Intelligent Analytic Execution of Workflows. The method of administering medical digital images claimed here involves receiving a medical image business object representing a transaction for a service request made by a healthcare provider, selecting a medical image business object comprised of a medical image and metadata and processing the image with the use of clinical workflows. This innovation is intended to improve the sharing and analysis of medical images across a multi-facility enterprise. Medical imaging is also an important aspect of the invention protected by U.S. Patent No. 8977959, issued under the title Visualization of Medical Conditions in a Virtual Universe. It discloses a computer-implemented method for virtual diagnosis which involves receiving patient imaging data in a memory medium, generating a set of avatars corresponding to the human patient emulating a medical condition of the patient, providing an interface that lets the patient choose the avatar most closely matching the patient’s physical state, receiving a treatment plan from a remote medical practitioner in a virtual clinic to address the patient’s medical condition, simulating a patient outcome based on the treatment plan and adjusting the appearance of the avatars based on the outcome simulation. This invention is designed to establish a remote patient treatment system which seeks to provide medical diagnosis and prognosis information to geographically isolated patients.
IBM is also leveraging the natural language processing aspects of its Watson platform to help doctors deliver the right diagnoses in clinical settings, as is reflected within U.S. Patent No. 9002773, entitled Decision-Support Application and System for Problem Solving Using a Question-Answering System. The method protected here involves receiving an initial query and problem case information in a decision support system which contains instructions that are processed to automatically extract relevant information to generate a natural language query, generating answers to the natural language query, calculating numerical values of evidence dimensions for each answer, calculating corresponding confidence values for each answer, outputting the query, answers and confidence values to a decision-maker and determining whether previous answers and confidence values have changed based on new domain knowledge. This innovation seeks to bring more structure to the process of medical diagnosis, which sometimes involves retrieving patient data scattered across different computer systems.
Patent Applications of Note: Natural Language Access of Medical Records and Prescribing Exercise Routes
Natural language processing systems for medical applications are also at the center of U.S. Patent Application No. 20150142701, filed under the title Evidence Based Medical Record. This patent application would protect a method of providing a confidence-estimation-based inference by receiving a query concerning a patient from a user, accessing an electronic health record (EHR) for the patient, querying the user for a second natural language component regarding the patient, receiving the second component in response to the query, calculating a first probability density function using the first component and a second probability density function with the second component, combining the probability density functions using a Gaussian mixture model, calculating at least one conditional probability table using the Gaussian mixture model and providing the confidence-estimation-based inference based on the conditional property table. This system enables doctors and medical professionals to access EHRs using natural language techniques for easier retrieval.
Systems for helping doctors, especially inexperienced ones, make the proper diagnosis in clinical settings are at the center of U.S. Patent Application No. 20140303986, which is titled Prediction of an Optimal Medical Treatment. The method of determining an optimal treatment disclosed here involves determining a frequency for each healthcare provider indicating how frequently each treats a selected disease, determining an average patient outcome (APO) for treating the disease at each healthcare provider, determining a score for each healthcare provider based on the corresponding frequency and APO, determining which healthcare providers are experts from scores exceeding a threshold level and selecting a treatment proscribed by an identified expert as the optimal treatment. This system lets younger, inexperienced physicians better benefit from the knowledge of expert physicians and pass along that medical knowledge to their patients. An invention for prescribing personal health and wellness exercises is discussed within U.S. Patent Application No. 20140358422, which is titled Prescription-Based Travel Route Recommendation. This claims a computer-implemented method for prescription-based travel route recommendation which involves receiving a prescription of a patient’s medical condition from a healthcare provider including a desired travel route with a start and a destination, a topography of the route, a distance between the start and destination points, a time of day for traversing the route and a travel mode, receiving information corresponding to a patient’s geographical area, identifying a set of possible travel routes, determining at least one travel route for treating the patient’s medical condition and recommending the travel route to the patient. This technology automates the job of identifying an exercise route that meets criteria as prescribed by a medical professional.
Systems for providing patient-specific diagnoses and treatments are also on display within U.S. Patent Application No. 20150235000, titled Developing Health Information Feature Abstractions from Intra-Individual Temporal Variance Heteroskedasticity. The patent application discloses a method to automatically abstract and select an optimal set of variance-related features indicative of an individual outcome in healthcare by generating an abstracted set of candidate variance-related patient features which are temporally heteroskedastic, optimizing each patient feature by identifying a time period in which patient variances and heteroskedasticity are maximized, comparing the optimal abstracted set of variance-related patient features to a historical set of data for a population of patients to create a predictive set of variance-related patient features predicting a target health-related outcome for a population of patients, generating a current patient optimal set of variance-related patient features, comparing the optimal set for the population of patients to the optimal set for the current patient, determining whether the target health-related outcome matches a predefined health-related outcome for the current patient and issuing an alert related to the predefined health-related outcome if it matches the target health-related outcome. This innovation improves the ability of disease self-management programs to determine metrics appropriate to an individual while addressing heteroskedasticity, or the variability of a variable which is unequal across a range of values of a second variable predicting the first one.
Teva’s Issued Patents: Generic Pharmaceuticals for Autoimmune Disease and Kidney Cancer
We would be remiss if we only focused on IBM’s patent portfolio and didn’t survey the holdings of Teva Pharmaceuticals at all. Teva has earned 39 U.S. patent grants thus far in 2015, most of those focused on areas of pharmaceutical compositions and pharmaceutically acceptable salts. We thought we’d take a look at a few of these to get a sense of what areas of medicine in which Teva is looking to improve its position.
A wide spectrum of diseases are the target of the cytokine-inhibiting medication which is
protected by U.S. Patent No. 9127057, entitled Anti-IL-23 Heterodimer Specific Antibodies. This patent protects an isolated or recombinant antibody or antigen-binding fragment that specifically binds to IL-23 comprised of various light chains and heavy chains of amino acids. This innovation provides an antagonist for interleukin-23 (IL-23), overproduction of which has been attributed to a variety of diseases including cancer, infection, inflammatory conditions or graft-versus-host disease (GVHD). Advances against inflammatory diseases of the central nervous system are the focus of U.S. Patent No. 9102620, titled Deuterated N-Ethyl-N-Phenyl-1,2-Dihydro-4-Hydroxy-5-Chloro-1-Methyl-2-Oxoquinoline-3-Carboxamide, Salts and Uses Thereof. This patent discloses a mixture of at least two deuterium-enriched compounds having a specific structure. This innovation provides a treatment for humans suffering from autoimmune disease by administering deuterium-enriched variants of laquinimod without creating toxicity in a patient. Finally, we noted a generic formulation of a renal cell carcinoma treatment originally developed by Pfizer (NYSE:PFE), described within U.S. Patent No. 9067915, entitled Sunitinib and Salts Thereof and Their Polymorphs. The patent protects a crystalline form of sunitinib acetate having a specific powder X-ray diffraction pattern. This invention relates to a new preparation for the multi-kinase inhibitor known as Sutent, prescribed for the treatment of some kidney cancers.