Abstract: Mechanisms are provided for implementing a patient complexity classification (PCC) computing system. The PCC computing system receives medical image study data for a patient that comprises one or more medical image data structures and one or more corresponding medical image metadata data structures. A natural language processing engine of the PCC computing system performs natural language processing on the medical image metadata data structure to extract features indicative of at least one of patient or medical image characteristics. A complexity classifier of the PCC computing system evaluates the extracted features to determine a patient complexity indicating a complexity of a medical condition of the patient. Routing logic associated with the PCC computing system routes the one or more medical image data structures and one or more corresponding medical image metadata data structures to one or more downstream patient evaluation computing systems based on the determined patient complexity.

Abstract: Methods and systems for structuring, storing and displaying time series data in a user interface. One system includes processors executing instructions to determine, from time series data from a first sensor, a first subset of time series data for the first batch from the first start time and the first end time, determine, from the time series data from the first sensor, a second subset of time series data for the second batch from the second start time and the second end time, generate a time series user interface comprising a chart, the chart including a first plot for the first subset of time series data and a second plot for the second subset of time series data, the first plot being aligned to the second plot, and cause presentation of the time series user interface.

Abstract: Provided is a lab clearinghouse device configured to provide centralized management of medical tests across a plurality of medical providers, a plurality of lab payers, and a plurality of laboratories. The lab clearinghouse device is configured to communicate with a plurality of medical laboratories, medical providers, and lab payers to efficiently and effectively order and manage medical testing.

Abstract: The embodiments of the present disclosure provide systems and methods for managing inventory placement, comprising a memory storing instructions and at least one processor configured to execute the instructions. The processor may be configured to receive an identifier of a product for inventory placement, and determine, based on historical shipment data stored in a database, a region with the highest customer demand for the product. The processor may predict, using a machine learning algorithm, a product tag associated with the product based on at least a temperature associated with the region with the highest customer demand for the product. The processor may further modify the database to assign the product tag to the product identifier, and assign the product for placement in a fulfillment center. The fulfillment center may be associated with a fulfillment center tag corresponding to the product tag assigned to the product.

Abstract: A system for programming a monitoring device includes a computing device configured to obtain a user datum of a plurality of user datums from a monitoring device, calculate a signal profile as a function of the user datum, identify a scan frequency correlated to the signal profile, wherein identifying further comprises receiving a frequency training set relating at least a first element of a vigor element to at least a first frequency requirement and using a frequency machine learning process, wherein the frequency machine learning process is configured using the signal training set, generate a device scheme as a function of the scan frequency, and program the monitoring device as a function of the device scheme.

Abstract: Dental images are processed according to a first machine learning model to determine teeth labels. The teeth labels and image are processed using a second machine learning model to label anatomy. The anatomy labels, teeth labels, and image are processed using a third machine learning model to obtain feature measurements, such as pocket depth and clinical attachment level. The feature measurements, labels, and image may be input to a fourth machine learning model to obtain a diagnosis for a periodontal condition. Machine learning models may further be used to reorient, decontaminate, and restore the image prior to processing. A machine learning model may be trained with images and randomly generated masks in order to perform inpainting of dental images with missing information.

Abstract: Clinical activity network applications are described. Some embodiments provide an Operational Intelligence Platform (“OIP”) that is configured to generate a clinical activity network that uniformly represents actions and data updates occurring in a source healthcare system. Each clinical activity instance associates an action (or activity) with a context, a time, a user, and a patient or other subject. By analyzing collections of clinical activity instances, applications can identify patterns, processes, and relationships within the healthcare setting that would not necessarily be detectable when analyzing data from a single source, such as the source customer database.

Abstract: The present disclosure relates to methods and systems for obtaining a nutraceutical composition. The method comprises the use of genome-scale metabolic networks of microorganisms to identify nutraceutical compositions for one or more health conditions.

Abstract: The specimen analyzer includes: an input unit which receives an input of subject attribute information; an analysis unit which performs measurement of a specimen collected from a subject, and which performs analysis of the specimen based on a measurement result and the subject attribute information received by the input unit; and a controller which causes the analysis unit to be incapable of analyzing the specimen unless the input of the subject attribute information is performed with the input unit.

Abstract: A directive based access system and method manage access permissions in systems. In one embodiment, the directive based access system and method may be used to orchestrate effective secure access control and communications in multi-cloud distributed systems. In one implementation, the directive based access system and method may include a lineage traceability enforcement engine that uses a lineage traceability. The directive based access system and method may also be implemented using other mechanisms such as blockchain based Hyperledger based system.

Abstract: A computer-aided design (CAD) system enables physical articles to be customized via printing or embroidering and enables digital content to be customized and electronically shared. A CAD user interface may be generated that includes an image of a model of an article of manufacture and a customizable template. The customizable template may include user customizable design areas. One or more defined rules associated with respective customizable areas may be accessed. In response to a user selection of a default content item and a corresponding rule, content items may be automatically used to populate other template design areas and/or change a color of one or content items. Manufacturing instructions corresponding to the user customizations may be transmitted to a printing system using a file that includes location, rotation, and/or scale data.

Abstract: Artificial intelligence is introduced into document review to identify content suggestions from input to generate suggested annotations for the reviewed document. An approach is provided for receiving an electronic document that contains original content from an original electronic document for review and electronic mark-ups provided by a first user. One or more electronic mark-ups that represent content suggestions proposed by the first user are identified from the electronic document. For each electronic mark-up of the one or more electronic mark-ups identified a document portion of the original content that corresponds to the electronic mark-up is identified, and an annotation is generated for the electronic mark-up comprising the electronic mark-up and a first user ID for the first user and associating the annotation to the document portion identified.

Abstract: A computer-implemented method is provided to provide analysis of claim information. The system may receive claim data from a plurality of entities and, for one or more claim items in the received claim data, determine a format of the claim item. The system may convert the claim item from the determined format into a standard format. The system may receive a selection of one or more providers, and determine one or more claim items associated with the provider in the selection. The system may generate a user interface, the user interface comprising a visual representation of one or more attributes of the one or more claim items.

Abstract: A dynamic decision forest may be used to guide clinical or other decision queries such that multiple decision trees are followed in sequence and jumps occur between decision trees based on previously received query responses and the system's determination of what useful data has not been gathered. For example, rather than storing data concerning all possible paths between various decision trees, a system using the dynamic decision forest may determine at the time of use that a jump should occur from one decision tree to another decision tree despite the system not being pre-populated with data representative of a relationship between the source and destination of the jump. A clinical note may be created based on the path followed through the dynamic decision forest, and iterative verification may be used in order to improve the likelihood that the correct path is being followed.

Abstract: A medical device includes a temperature sensor configured to deliver a temperature signal and an optical sensor configured to deliver an optical signal. The medical device also includes a microcontroller configured to receive the temperature signal and the optical signal. The microcontroller is configured to calculate, in real-time, a body temperature, a pulse rate, a respiratory rate, and a blood oxygen concentration based on the received temperature signal and the received optical signal. The medical device also includes a display configured to display the body temperature, the pulse rate, the respiratory rate, and the blood oxygen concentration of a patient.

Abstract: A detection device which, when an arrhythmia occurs, can prompt rapid taking of a countermeasure against the arrhythmia, and an alarm system having the detection device are provided. A portable detection device 10 has: a measuring section which can measure an electrocardiogram in a state where the section is applied to the body surface of the user; an analyzing section which analyzes whether the electrocardiogram contains an abnormal waveform indicating an arrhythmia or not; and a transmitting section which, if the analyzing section detects the abnormal waveform, transmits a detection signal indicating that the abnormal waveform is detected, to a communication terminal 20 around the user.

Abstract: By extracting clinical data of any format from respective different sources, a data repository normalized to a generic format is created. A medical domain specific language may be used to interact with the data repository for identifying cohorts and gaps in care for the respective cohorts. Any rules for finding gaps in care are converted into the medical domain specific language for determining gaps. This standardization in both the data repository and rule application may allow for a true cost and time to value solution accessible to many different medical practices.

Abstract: A method, a system, and a computer program product are provided. A machine learning model is generated to process adverse event information and produce multiple corresponding medical codes associated with the adverse event information, wherein the multiple medical codes are semantically and hierarchically related in a medical taxonomy. The machine learning model includes multiple parallel output layers, each of which is associated with a corresponding medical code. The machine learning model is trained with training data elements, each of which includes adverse event information mapped to respective multiple medical codes, wherein results from each of the output layers adjusts the machine learning model. After completing the training, information pertaining to an adverse event is applied to the machine learning model to determine the corresponding multiple medical codes within the medical taxonomy.

Abstract: Systems and methods are provided for improved integration between database systems. Data from multiple disparate computing systems is transmitted via bi-directional communication interfaces. The data is stored in its original form from respective data sources and transformed in stream and/or batch processes into one or more predefined formats. Individual transformations are stored and/or recorded. The transformed and/or integrated data is provided to one or more computing devices via the communication interfaces.

Abstract: A method for capturing patient feedback for a medical treatment includes: receiving user instructions indicating a symptom being treated, a symptom score for the symptom being treated, and an indication of a beginning of a treatment session; receiving patient feedback comprising an updated symptom score for the symptom being treated and/or absence or presence of one or more of a plurality of attributes; storing the received patient feedback and a time in the treatment session at which the respective patient feedback was received; repeating the receiving of patient feedback and storage of data during the treatment session; and receiving a user instruction indicating an ending of the treatment session, wherein an alerting module alerts a user of the computing device for input of patient feedback after a predetermined period of time has elapsed following receipt of patient feedback without a subsequent receipt of patient feedback.

Abstract: Network architectures interface with several real-time streams that control network operation. The network can analyze stream data by storing incoming raw stream data, processing the data, and controlling network operations based thereon. Specific packets, bits, flags, fields, or messages may be targeted, including HL7 ADTs. Data is queued for processing in a FIFO manner to avoid timeline mistakes. The network sets a configuration value based on the data analyzed in this time ordering, and a relational database can be updated in real-time with these values as stream data is incoming. Separate clean-up functions separately manage the database by adjusting the values as they become old. Timeline information is generated and selectively broadcast based on the operations value, potentially with additional information from the targeted data. The network selectively provides this information to users interfaces and displays as dictated by the operating value and information calculated therefrom.

Abstract: Methods, computer systems and computer readable media for receiving data and information from medical devices in a healthcare setting are provided. In embodiments, a graphical indication is provided that indicates an infusion is being administered. Upon an interaction with the user interface being received, a billing link is provided. Once a selection of the billing link is received, encounter specific infusion information is provided. The encounter specific infusion information is provided to a billing service.

Abstract: An article management system collects articles from different sources, processes each article to extract information about the article (such as the topics covered in the article), and stores the information in one or more knowledge graphs. The system also maintains user profiles for each of its users. Each user profile includes a user genome representing the user's interests and preferences. The information in the knowledge graphs can be used to select articles to be displayed to users of the article management system. Articles can be displayed as part of the search results for an associated topic or concept, as part of a personalized feed that is generated based on the data in a user's profile, or as part of a feature that displays multiple articles that cover a given topic from multiple different viewpoints. The system can also provide a platform for users to create and publish research collections.

Abstract: Methods, systems, devices and/or apparatus related to finding and sharing information on the Internet. Specifically, the disclosed methods, systems, devices and/or apparatus relate to utilizing reference materials to find and share topical information on the Internet.

Abstract: The invention provides for a medical imaging system (100, 300, 400) comprising a memory (110) containing machine executable instructions (120) and a processor (106). Execution of the machine executable instructions cause the processor to: receive (200) a magnetic resonance image (300), receive (202) meta data descriptive of the magnetic resonance image, wherein the metadata comprises reference gray scale value data (124) for two or more tissue types; and segment (204) the magnetic resonance image using an image segmentation algorithm (126) that uses the reference gray scale value data.

Abstract: Techniques for providing information regarding healthcare performance in real-time using augmented reality are provided. In one example, a computer-implemented method comprises receiving input data generated in real-time during performance of a healthcare procedure by a user, including video captured of the user during the performance of the healthcare related procedure. The method further includes determining descriptive information characterizing the performance based on the input data, wherein the descriptive information characterizes at least actions performed by the user during the procedure, and determining whether an aspect of one of the actions currently being performed by the user deviates from a defined protocol for the healthcare related procedure based on comparison of the descriptive information with reference descriptive parameters for the healthcare related procedure.

Abstract: The disclosed methods and systems are generally directed to the combined application of predictive analytics with clinical decision analysis to provide optimal recommendations for the implementation of various health care management programs including, for example, prescription medication adherence interventions, and providing recommendations for medication adherence interventions for individual patients, based on predictions of patient adherence and predicted or expected responses to available interventions. Such interventions may be instituted in advance of a potential negative outcome, e.g., non-adherence, to prevent or reduce the occurrence of a negative outcome and realize greater return on investment.

Abstract: A ranging system includes a terminal device, a ranging apparatus, and an angle sensor disposed in the terminal device or in the ranging apparatus. The ranging apparatus includes a measuring wheel, a rotating shaft disposed on the measuring wheel, and an electronic counter connected to the rotating shaft and configured to obtain distance information, in real time, of a route which the measuring wheel has already traveled. The angle sensor is configured to obtain angle information in real time. The terminal device and the ranging apparatus are configured to transmit data between each other. The terminal device is configured to generate a traveling trajectory diagram of the measuring wheel, in real time, based on the distance information from the electronic counter and the angle information from the angle sensor, and perform optimization processing on the traveling trajectory to generate an optimized trajectory diagram.

Abstract: A measurement device including a hardware processor that: extracts a frame image at a time of deep inspiration from a plurality of frame images of a dynamic image obtained by performing radiation imaging of a dynamic state of a chest of a subject; and performs cardiothoracic ratio measurement to the extracted frame image at the time of deep inspiration.

Abstract: An apparatus for pricing offers of healthcare services includes an application server providing a pricing tool, and a service pricing information database and a cost adjustment information database that are accessible by the application server. The pricing tool is operable to receive an indication of a healthcare service set that is respectively associated with a service detail information record included in the service pricing information database, determine a base physician fee for a primary service of the indicated healthcare service set based on base pricing metrics for the primary service included in the service detail information record, receive an indication specifying a location for performing the primary service, calculate an adjusted physician fee for the primary service based on cost adjustment metrics for a geographic zone that corresponds to the specified location, and provide a recommended price for the indicated healthcare service set based on the calculated adjusted physician fee.

Abstract: Embodiments of the present systems and methods may provide techniques to predict the success or failure of a drug used for disease treatment. For example, a method of determining drug efficacy may include, for a plurality of patients, generating a directed acyclic graph from health related information of each patient comprising nodes representing a medical event of the patient, at least one first edge connecting the first node to an additional node, each additional edge connecting nodes representing two consecutive medical events, the edge having a weight based on a time difference between the two consecutive medical events, capturing a plurality of features from each directed acyclic graph, generating a binary graph classification model on captured features of each directed acyclic graph, determining a probability that a drug or treatment will be effective using the binary graph classification model, and determining a drug to be prescribed to a patient based on the determined probability.

Abstract: Methods and systems are described for a medication container that may be locked using a locking mechanism. In one exemplary method, a request from a user may be received by a medication provider for a medication. The medication may be placed in a medication container with a locking mechanism that is operable to lock and unlock the medication container. The medication container may be locked using the locking mechanism. The locked medication container with the medication within may be provided to the user. Subsequent to providing the locked medication container to the user, a request may be received from the user to access the medication in the medication container. A credential may be sent to the user that is usable, via the locking mechanism, to unlock the medication container. The user may then input the credential into the locking mechanism to unlock the medication container and access the medication therein.

Abstract: The present invention relates to systems and methods for ensuring quality of care services. After care is provided a confirmation is received. This may include a message regarding the care delivery from the provider. The confirmation and message may be conveyed to friends and family of the recipient. Location data from the provider, and feeds from the recipient's location are collected. Any lapses of care and emergencies based upon the location data and feeds are likewise identified. Identifying lapses of care and emergencies includes cross referencing length of time the provider spent at the recipient's location against tasks that were supposed to be performed at the location to determine if there is a discrepancy. Likewise, the time and location of the provider can be cross referenced against the confirmation to determine if there is a discrepancy. In some embodiments the feeds are also analyzed for trigger events.

Abstract: Embodiments in the present disclosure relate generally to computer network architectures for machine learning, artificial intelligence, and risk adjusted performance ranking of healthcare providers. Embodiments of computer network architecture automatically make risk adjusted performance rankings of healthcare service providers and generate and transmit reports of the rankings. Embodiments may generate such rankings either occasionally on demand, or periodically, or as triggered by events such as an update of available data for such forecasts. Embodiments may include a combination of third-party databases to generate the updated forecasts for pending patient clinical episodes, and to drive the forecasting models for the same, including social media data, financial data, socio-economic data, medical data, search engine data, e-commerce site data, and other databases.

Abstract: An information processing apparatus includes an acquisition unit configured to acquire patient's medical information, an extraction unit configured to extract a part of the medical information as material candidates for a reduced medical image, a selection unit configured to assign priorities to a plurality of the extracted material candidates, and, based on the priorities, select materials to be used for generation of the reduced medical image out of a plurality of the material candidates, and a generation unit configured to generate the reduced medical image based on a plurality of the selected materials.

Abstract: Systems, methods, and machine-readable media are provided to facilitate controllable parameter modification of system components based on monitoring indicators of a caliber of process performance. A set of cases is identified from amongst the plurality of cases, the set of cases having similar temporal identifiers and a same type of process performance. A population check-point statistic is generated based on the check-point indicators identified in the set of cases. A first result that indicates whether a first criterion is met is determined based on the population check-point statistic. A second result that indicates whether a second criterion is met is determined based on a check-point indicator. When the first criterion or the second criterion is not met, one or more transmissions are to alert a remote device and to facilitate causing a system component to make a change to a controllable parameter to influence a corresponding check-point.

Abstract: A gamified pharmaceutical packaging, dispensing system, and methodology is disclosed. The innovative gamified smart packaging system and methodology provides a variety of incentives to patients and relevant teams to take their medications, on time, and at the correct dosage. The system comprises monitored pharmaceutical dispensing devices that communicate with a plurality of compute devices, or smartphones. Each time the patient dispenses the medicine from the dispensing device, a signal is transmitted to the compute device. Upon receiving the signal, the compute device records the dispensing event, calculates a specific “score” for that event, and reports the patient's cumulative “score” of successfully dispensing the medicine. If a dispensing event signal is not received in a timely fashion, the system sends a prompt to the patient to take their medication. Accumulating points by properly taking the prescribed medication allows the patient and/or team to be eligible to receive various benefits.

Abstract: Systems, methods, and computer readable storage medium for providing a genericized medical device architecture common to a plurality of medical devices are disclosed. The architecture may comprise at least one diagnostics module associated with at least one of the plurality of medical devices, wherein the at least one diagnostics module is configured to monitor an operational status of the at least one medical device. At least one hardware abstraction layer may be associated with at least one of the plurality of medical devices, and may be configured to provide abstracted access to hardware of the at least one medical device.

Abstract: A method for limiting exposure to infectious disease by controlling access to a controlled access venue to admit individuals who have tested non-infectious for a particular highly infectious disease or have received a vaccination against the same comprises receiving, at a central website operator server, consumer user identification information from consumer users and storing that information in a consumer user database associated with the server. A biometric identifier from the consumer users is transmitted to the consumer user database and each associated with its respective associated consumer user to form an enrollment record. Identity of a presented individual is collected at a medical certification point and an authority record is created in an accessible authority database. At said medical certification point, an antibody test or vaccination is administered and the same is noted in the authority record.

Abstract: A person support apparatus includes a first transceiver adapted to wirelessly communicate with a second transceiver of a headwall interface that is positioned off of the person support apparatus. A communication link is automatically established between the first and second transceivers without requiring a user of the person support apparatus to activate a designated control and without requiring the user to identify the headwall interface. The first transceiver includes a unique identifier assigned to the headwall interface in its messages to the headwall interface. The first transceiver may also automatically transmit a disconnect signal to the headwall interface indicating the termination of the communication link is not accidental. The disconnect signal is sent based on one or more of the following: (1) a brake being off, (2) an A/C power cord being unplugged; and/or (3) a signal strength between the transceivers decreasing.

Abstract: Systems and methods for fully-automated contact-less biometric measurement include receiving a stream of image frames from an internet capable device. Frame data is extracted from the stream of image frames. A predictive load balancer selects a worker server based on a load prediction. An interface delivery server communicates the frame data from the internet capable device to the worker server. A set of patient face images captured by the frame data are extracted. A facial recognition machine learning model determines a patient identity associated with the set of patient face images of the frame data. A shared memory system stores the set of patient face images. The worker server determines biometric measurements based on the set of patient face images in the shared memory system using independent biometric data processing pipelines with shared access to the shared memory system for inter-process communication.

Abstract: The example embodiments may relate to systems, methods, apparatuses, devices, and computer readable media for collecting and presenting information at the patient encounter level. The example embodiments may include processing an encounter detail file associated with treatment a healthcare provider provided to a patient during a patient encounter, generating a linkage that links the encounter detail file with at least one other file associated with the patient encounter, and storing the linkage in a longitudinal patient administration record. The example embodiments may further include processing the longitudinal patient administration record to determine a metric and select an item based on the metric, and generating and communicating an order to a supplier for the selected item to cause delivery of the selected item to a desired location.

Abstract: A method, computer program product, and computing system for obtaining encounter information during a patient encounter; processing the encounter information to detect the execution of a physical event during the patient encounter, thus defining a detected physical event; and deriving information for the detected physical event.

Abstract: A computerized method, system and computer program product may comprise: generating or receiving a search result in response to a user query by a user that searches for an item, the search result depicting the item; and modifying the search result by generating a combination image that depicts user-defined subject matter engaging with the item. The present invention may provide other and/or additional benefits or advantages.

Abstract: Embodiments of the disclosure provide systems and methods for generating a diagnosis report based on a medical image of a patient. The system includes a communication interface configured to receive the medical image acquired by an image acquisition device. The system further includes at least one processor. The at least one processor is configured to detect a medical condition based on the medical image and automatically generate text information describing the medical condition. The at least one processor is further configured to construct the diagnosis report, where the diagnosis report includes at least one image view showing the medical condition and a report view including the text information describing the medical condition. The system also includes a display configured to display the diagnosis report.

Abstract: Systems, methods, and computer readable media related to statistical analysis across surgical videos are disclosed. The methods may include receiving video frames across multiple surgical videos, each surgical video comprising multiple surgical procedures performed by a specific medical professional across differing patients. A set of surgical event-related categories may be accessed and video frames of each surgical video may be analyzed to identify surgical events defined by a subgroup of frames. Each subgroup of frames may be assigned to one of the surgical event-related categories to thereby interrelate subgroups of frames from differing surgical procedures to a common surgical event-related category. Statistics may be derived for each surgical event-related category and then aggregated within each category and displayed with the surgical event-related categories for selection.

Abstract: An improved electronic procurement system is disclosed. The electronic procurement system (“system”) implements features such as real-time adaptive extraction of online data. The system is configured to manage a plurality of page types for webpages and for each page type a plurality of field types for fields in webpages. The system is configured to further generate a computer application that can be integrated into a web browser. For each page type, the computer application is programmed to initially generate a signature for each field type based on minimal user interaction with webpages of the page type. The system is configured to also generate an agent using the signature. The agent is programmed to automatically extract data corresponding to the field types from additional webpages of the page type using the signatures. The agent or an associated background process is programmed to also automatically update the signatures when such data extraction is unsuccessful.

Abstract: Described herein are systems, devices, methods, and media for facilitating emergency communications by an emergency management system. In one embodiment, a method for facilitating emergency communications comprises: identifying a user potentially affected by an emergency; providing an emergency response message to a communication device associated with the user; receiving confirmation of the emergency; in response to receiving confirmation of the emergency, initiating an autonomous communication session with the user through the communication device; extracting emergency information from the autonomous communication session; and providing the emergency information to an emergency service provider (ESP).

Abstract: The method and system determine and report measures of adherence of patients to prescription medication regimens based upon fill gaps in the prescription history of the patients. The measure of adherence may be the proportion of days covered by the medication. The method and system may receive or access information relating to a patient's medical history, including past prescription medication fills. Fills of both the prescription medication and other medications that may substitute for the prescription medication may be considered in determining patient adherence levels. Upon determining the patient's level of adherence, the system and method may generate a report including the adherence information and associated information, such as a home pharmacy location. The report may include an indication of whether new-to-therapy information should be presented to a patient based on patient adherence levels.

Abstract: Mechanisms are provided for implementing a framework to learn multiple drug-adverse drug reaction associations. The mechanisms receive and analyze patient electronic medical record data and adverse drug reaction data to identify co-occurrences of references to drugs with references to adverse drug reactions (ADRs) to thereby generate candidate rules specifying multiple drug-ADR relationships. The mechanisms filter the candidate rules to remove a subset of one or more rules having confounder drugs specified in the subset of one or more candidate rules, and thereby generate a filtered set of candidate rules. The mechanisms further generate a causal model based on the filtered set of candidate rules. The causal model comprises, for each ADR in a set of ADRs, a corresponding set of one or more rules, each rule specifying a combination of drugs having a causal relationship with the ADR.