Abstract: Automatic mapping of semantics in healthcare is provided. Data sets have different semantics (e.g., Gender designated with M and F in one system and Sex designated with 1 or 2 in another system). For semantic interoperability, the semantic links between the semantic systems of different healthcare entities are created (e.g., Gender=Sex and/or 1=F and 2=M) by a processor from statistics of the data itself. The distribution of variables, values, or variables and values, with or without other information and/or logic, is used to create a map from one semantic system to another. Similar distributions of other variable and/or values are likely to be for variables and/or values with the same meaning.

Abstract: The present invention provides a data mining framework for mining high-quality structured clinical information. The data mining framework includes a data miner that mines medical information from a computerized patient record (CPR) based on domain-specific knowledge contained in a knowledge base. The data miner includes components for extracting information from the CPR, combining all available evidence in a principled fashion over time, and drawing inferences from this combination process. The mined medical information is stored in a structured CPR which can be a data warehouse.

Abstract: A method, including receiving a data source selection from a user or software application, the data source including medical information of a plurality of patients, receiving, from the user or software application, a data pattern that is related to a concept to be explored in the data source, querying the data source to find information that approximately matches the data pattern; and receiving the information from the data source, wherein the information includes unstructured data, assigning a classification to individual parts of the information based on the part's relationship to the data pattern, and outputting the classified information to the user or software application.

Abstract: A system and method for lung cancer screening is provided. The system includes a database including structured patient information for a patient population and a domain knowledge base including information about lung cancer; an individual patient record; and a processor for analyzing the patient record with data from the database to determine if a patient has indications of lung cancer. The method includes the steps of inputting patient-specific data into a patient record; performing at least one lung cancer screening procedure on a patient, wherein at least one result from at least one procedure is inputted into the patient record in a structured format; and analyzing the patient record with a domain knowledge base to determine if the patient has indications of lung cancer.

Abstract: Medical related quality of care information is extracted and edited for reporting. Patient records are mined. The mining may include mining unstructured data to create structured information. Measures are derived automatically from the structured information. A user may then edit the measures, data points used to derive the measures, or other quality metric based on expert review. The editing may allow for a better quality report. Tools may be provided to configure reports, allowing generation of new or different reports.

Abstract: A system and method for diagnosis and treatment decisions based on information maximization is disclosed. Utilizing patient information as well as clinical records from other patients can reduce the uncertainty in both diagnosis and treatment options. The information maximization may consider additional data such as risk, cost, and comfort in making a proper medical decision.

Abstract: A system and method for analyzing population-based patient information is provided. The method includes the steps of data mining a plurality of patient records using a domain knowledge base relating to a disease of interest; compiling the mined data into a plurality of structured patient records; inputting at least one patient criteria relating to the disease of interest; and extracting at least one structured patient record matching the at least one patient criteria. The system includes a data miner for mining information from the plurality of patient records using a domain knowledge base relating to a disease of interest and for compiling the mined data into a plurality of structured patient records; an interface for inputting at least one patient criteria relating to the disease of interest; and a processor adapted for extracting at least one of the structured patient records matching the at least one patient criteria.

Abstract: Automated treatment planning is provided with individual specific consideration. One or more prognosis models indicate survivability as a function of patient specific information for a given dose. By determining survivability for a plurality of doses, the biological model represented by survivability as a function of dose is determined from the specific patient. Similarly, the chances of complications or side effects are determined. The chance of survivability and chance of complication are used as or instead of the tumor control probability and normal tissue complications probability, respectively. The desired tumor dosage and tolerance dosage are selected as a function of the patient specific dose distributions. The selected dosages are input to an inverse treatment planning system for establishing radiation treatment parameters.

Abstract: Quality improvement factors in patient care are ranked. Hospital performance is measured, such as a CMS measure. The variables and/or values relative contribution to quality of care is determined using medical records of the hospital. The variables and/or values are ranked according influence of the quality of care result. The ranking is performed by a given medical institution at a desired time rather than based on a broad study. The medical institution may regularly determine variables (e.g., admitting doctor) and/or values (e.g., doctor X) that are relevant to a decreased quality of care. Quality may be regularly improved using a software product.

Abstract: A method is provided for automatically identifying a disease outbreak indicative of a potential bio-terror attack. Patient records are mined from structured and unstructured clinical sources. The patient records are then analyzed by correlating selected patient data contained in the patient records with disease indicia for each of a plurality of diseases. A probability of a disease outbreak is estimated at least in part based on these correlations. Suspicion may also be indicated if anomalous disease clusters are found. If any of the estimated probabilities exceeds a threshold value, a disease outbreak alert is outputted. The disease indicia may be defined by disease progression models, which may be stored in a disease knowledge base.

Abstract: A system and method for screening for coronary heart disease is provided. The method includes the steps of retrieving a test for assessing risk of coronary heart disease, the test including a plurality of data fields relating to coronary risk factors; accessing a database to populate the data fields with information of an individual patient; and calculating a risk assessment of the individual patient developing coronary heart disease. A system includes a first database including a plurality of structured computerized patient records; a second database including a knowledge base relating to coronary heart disease, the second database including at least one test for determining coronary heart disease risk; and a processor for retrieving the at least one test from the second database, populating the at least one test with patient information retrieved from the first database and calculating a risk assessment for at least one patient.

Abstract: A computer-implemented method for privacy-preserving data mining to determine cancer survival rates includes providing a random matrix B agreed to by a plurality of entities, wherein each entity i possesses a data matrix Ai of cancer survival data that is not publicly available, providing a class matrix Di for each of the data matrices Ai, providing a kernel K(Ai, B) by each of said plurality of entities to allow public computation of a full kernel, and computing a binary classifier that incorporates said public full kernel, wherein said classifier is adapted to classify a new data vector according to a sign of said classifier.

Abstract: A method for creating and searching medical ontologies includes providing a semi-structured information source comprising a plurality of articles linked to each other, each article having one or more sections and each article is associated with a concept, creating a directed unlabeled graph representative of the information source, providing a plurality of labels, labeling a subset of edges, and assigning each unlabeled edge an equal probability of being assigned one of the labels. For each node, the probability of each outgoing edge is updated by smoothing each probability by an overall probability distribution of labels over all outgoing edges of each node, and the probability of each incoming edge is updated the same way. A label with a maximum probability is assigned to an edge if said maximum probability is greater than a predetermined threshold to create a labeled graph.

Abstract: The present invention provides systems and methods for automatically ensuring adherence to clinical guidelines during the course of patient treatments. A data source contains patient records, including records for patients being treated; a guidelines knowledge base contains clinical guidelines; and a quality adherence engine is configured to monitor adherence with the clinical guidelines for patients being treated. At least some of the patient records may include information obtained from mining unstructured patient data. The system includes an output component for outputting quality adherence information. The outputted quality adherence information may include reminders, including reminders to take clinical actions in accordance with the clinical guidelines. The outputted quality adherence information may also include warnings that the clinical guidelines have not been observed.

Abstract: The present invention provides a graphical user interface for presentation, exploration and verification of patient information. In various embodiments, a method is provided for browsing mined patient information. The method includes selecting patient information to view, at least some of the patient information being probabilistic, presenting the selected patient information on a screen, the selected patient information including links to related information. The selected patient information may include elements, factoids, and/or conclusions. The selected patient information may include an element linked to unstructured information. For example, an element linked to a note with highlighted information may be presented. Additionally, the unstructured information may include medical images and waveform information.

Abstract: A method for training a classifier for classifying candidate regions in computer aided diagnosis of digital medical images includes providing a training set of images, each image including one or more candidate regions that have been identified as suspicious by a computer aided diagnosis system. Each image has been manually annotated to identify malignant regions. Multiple instance learning is applied to train a classifier to classify suspicious regions in a new image as malignant or benign by identifying those candidate regions that overlap a same identified malignant region, grouping each candidate region that overlaps the same identified malignant region into a same bag, and maximizing a probability P = ? i = 1 N ? p i y i ? ( 1 - p i ) 1 - y i , wherein N is a number of bags, pi is a probability of bag i containing a candidate region that overlaps with an identified malignant region, and yi is a label where a value of 1 indicates malignancy and 0 otherwise.

Abstract: A method for multiple-label data analysis includes: obtaining labeled data points from more than one labeler; building a classifier that maximizes a measure relating the data points, labels on the data points and a predicted output label; and assigning an output label to an input data point by using the classifier.

Abstract: Modeling of prognosis of survivability, side-effect, or both is provided. For example, RILI is predicted using bullae information. The amount, volume or ratio of Bullae, even alone, may indicate the likelihood of complication, such as the likelihood of significant (e.g., stage 3) pneumonitis. As another example, RILI is predicted using uptake values of an imaging agent. Standardized uptake from a functional image (e.g., FDG uptake from a positron emission image), alone or in combination with other features, may indicate the likelihood of side-effect. In another example, survivability, such as two-year survivability, is predicted using blood biomarkers. The characteristics of a patient's blood may be measured and, alone or in combination with other features, may indicate the likelihood of survival. The modeling may be for survivability, side-effect, or both and may use one or more of the blood biomarker, uptake value, and bullae features.

Abstract: A method for predicting survival rates of medical patients includes providing a set D of survival data for a plurality of medical patients, providing a regression model having an associated parameter vector ?, providing an example x0 of a medical patient whose survival probability is to be classified, calculating a parameter vector {circumflex over (?)} that maximizes a log-likelihood function of ? over the set of survival data, l(?|D), wherein the log likelihood l(?|D) is a strictly concave function of ? and is a function of the scalar x?, calculating a weight w0 for example x0, calculating an updated parameter vector ?* that maximizes a function l(?|D?{(y0,x0,w0)}), wherein data points (y0,x0,w0) augment set D, calculating a fair log likelihood ratio ?f from {circumflex over (?)} and ?* using ?f=?(?*|x0)+sign(?({circumflex over (?)}|x0)){l({circumflex over (?)}|D)?l(?*|D)}, and mapping the fair log likelihood ratio ?f to a fair price y0f, wherein said fair price is a probability that class label y0 for exam

Abstract: A method of training a classifier for computer aided detection of digitized medical image, includes providing a plurality of bags, each bag containing a plurality of feature samples of a single region-of-interest in a medical image, where each region-of-interest has been labeled as either malignant or healthy. The training uses candidates that are spatially adjacent to each other, modeled by a “bag”, rather than each candidate by itself. A classifier is trained on the plurality of bags of feature samples, subject to the constraint that at least one point in a convex hull of each bag, corresponding to a feature sample, is correctly classified according to the label of the associated region-of-interest, rather than a large set of discrete constraints where at least one instance in each bag has to be correctly classified.