Neuroimaging database systems and methods
Inventors
Bradshaw, Vincent • Henderson, Theodore • Faherty, Jennifer • Bitto, Donald • Villegas-Mauter, Nikki
Assignees
Interested in licensing this patent?
MTEC can help explore whether this patent might be available for licensing for your application.
Publication Number
US-8838565-B2
Publication Date
2014-09-16
Expiration Date
Abstract
Systems for and methods of utilizing a neuroimaging database are presented. The systems and methods include techniques for analyzing the pathophysiological basis of a chronic brain disease and/or the effectiveness of a treatment for a chronic brain disease, obtaining data for research of a chronic brain disease, searching for chronic brain disease symptoms identified in a clinical patient, searching a database by comparing the brain scan images of patients with suspected indications of chronic brain disease with other patients in the database to identify sets of patients with similar indications in their brain scan images, displaying brain scan information regarding a person, and using image pattern matching to analyze the pathophysiological basis of a chronic brain disease and/or the effectiveness of a proposed or previously administered treatment for a chronic brain disease.
Core Innovation
The invention is a neuroimaging database system that stores electronic patient records and uses computer processors to query subsets of the records based on diagnostic and treatment or symptom information for chronic brain disease. Each patient record includes one or more digital representations of a baseline functional brain scan image and one or more digital representations of a concentration functional brain scan image, together with digital representations of medical history, neuropsychiatric assessment, demographic information, and a set of clinical symptoms, and optionally one or more diagnostic findings by a radiologist or physician.
The system identifies a treatment set of patients diagnosed with the chronic brain disease and having received the treatment and a non-treatment set of patients diagnosed with the chronic brain disease but not receiving the treatment. For the treatment set and the non-treatment set, the system outputs statistical comparisons of each subset to a normative set of electronic records corresponding to patients that were not diagnosed with the chronic brain disease and did not receive the treatment, and the results are used so that information regarding the pathophysiological basis of the chronic brain disease or the effectiveness of the treatment may be analyzed.
The invention further supports identifying common characteristics by finding one or more features present in at least a percentage of a disease subset and outputting, for each feature, the percentage of normative database records in which the one or more features are present. A further aspect is the use of a perfusion pattern index file that comprises a digital representation of statistical deviations of brain perfusion levels in a plurality of regions from values derived from normative databases of patients with no indications of chronic brain disease, with storage, query-based matching, determination of an average PPI file, and comparison of each patient's PPI file to the average PPI file.
Another aspect provides symptom-based retrieval and simultaneous display of baseline and concentration scan information for a person and for at least one brain scan image of the candidate subset patients whose records match the symptoms.
Claims Coverage
The partial set includes five independent claims that cover five inventive features: treatment-set versus non-treatment-set statistical comparison to a normative set, identification of common disease-associated features compared against a normative database, symptom-based ranking and diagnostic output, symptom-based simultaneous display of brain scan information, and perfusion pattern index creation, matching, averaging, and patient-by-patient comparison.
Treatment set and non-treatment set statistical comparison to a normative set
Receive a first input and execute a first query to identify a first subset corresponding to a treatment set of patients diagnosed with the chronic brain disease and having received the treatment; receive a second input and execute a second query to identify a second subset corresponding to patients diagnosed with the chronic brain disease that did not receive the treatment; output a first statistical comparison of the first subset to a normative set corresponding to patients not diagnosed with the chronic brain disease and not receiving the treatment; output a second statistical comparison of the second subset to the normative set; analyze information regarding the pathophysiological basis of the chronic brain disease or the effectiveness of the treatment based on the first statistical comparison and second statistical comparison.
Disease subset feature identification compared to normative feature percentages
Receive an input, convert the input into a first query, and execute the first query to identify a disease subset corresponding to patients whose records include a diagnosis of the chronic brain disease; identify and output one or more features present in at least a percentage of the disease subset; for each feature, execute a second query to identify a subset of a normative database corresponding to patients not diagnosed with the chronic brain disease and whose records include at least one of the features; output, for each feature, the percentage of normative database records in which the one or more features are present, including presentation of the one or more features and the percentage of disease and normative records in which the one or more features are present.
Symptom-based candidate subset ranking and possible diagnosis output
Receive an input reflecting a set of symptoms of a clinical patient, convert the input into a first query, and execute the first query to identify a candidate subset of patient database records corresponding to patients whose records include the set of symptoms; rank the candidate subset according to relevancy to the set of symptoms to produce a ranking; output in user viewable form at least a portion of the ranking whereby a possible diagnosis of the clinical patient is output.
Symptom-based candidate subset selection with simultaneous display of baseline and concentration scans
Receive an input reflecting a set of symptoms of the person, convert the input into a first query, and execute the first query to identify a candidate subset of patient database records corresponding to patients whose records include the set of symptoms; cause the simultaneous display of at least one baseline functional brain scan image and at least one concentration functional scan image, and at least one brain scan image of the candidate subset of the patient database records.
Perfusion pattern index matching and average-based comparison
Create a perfusion pattern index file for a first patient by determining digital levels of statistical deviation in brain perfusion levels between the brain scans of the first patient and normative brain scans of patients with no indications of chronic brain disease; store the PPI file in the plurality of records; execute a query to identify a subset of records corresponding to patients whose database records indicate a PPI file that matches the first patient's PPI file based on one or more search parameters; determine an average PPI file for the subset; output a comparison of the PPI file of each patient in the set to the average PPI file.
Across the independent claims, the core coverage combines electronic patient records with functional brain scan representations, symptom inputs and query-based retrieval, and normative database comparisons. The claims further cover treatment-set versus non-treatment-set statistical comparisons, disease feature identification with disease-versus-normative percentages, and PPI-based matching with determination of an average PPI file and patient-by-patient comparison.
Stated Advantages
Enables analysis of information regarding the pathophysiological basis of the chronic brain disease or the effectiveness of the treatment based on statistical comparisons against a normative set.
Identifies one or more features associated with a chronic brain disease and outputs feature presence percentages in both disease and normative databases for user presentation.
Provides symptom-based ranking of candidate records and outputs a possible diagnosis.
Provides brain scan information by causing simultaneous display of baseline and concentration scan images for a person and related candidate-subset scan images.
Determines an average PPI file for a matched subset and outputs comparisons of each patient's PPI file to the average PPI file.
Documented Applications
Analyzing chronic brain disease pathophysiological basis and/or the effectiveness of a treatment using statistical comparisons between treatment and non-treatment patient subsets versus a normative set.
Identifying common characteristics associated with a chronic brain disease by detecting features present in disease subsets and comparing feature presence percentages to normative database records.
Locating chronic brain disease symptoms identified in a clinical patient by symptom-based querying, relevancy ranking, and outputting a possible diagnosis.
Providing brain scan information regarding a person with simultaneous display of baseline and concentration scan images and at least one brain scan image of candidate-subset patients whose records match the person’s symptoms.
Locating chronic brain disease perfusion pattern characteristics by creating a PPI file for a first patient, matching PPI files across records using search parameters, determining an average PPI file for the matched subset, and outputting patient-by-patient comparisons to the average PPI file.
Interested in licensing this patent?