WO2023199084A2 - Blocs de données délimitées et définies par des symboles pour écrire des histoires de données riches destinées à être utilisées avec une intelligence artificielle - Google Patents

Blocs de données délimitées et définies par des symboles pour écrire des histoires de données riches destinées à être utilisées avec une intelligence artificielle Download PDF

Info

Publication number
WO2023199084A2
WO2023199084A2 PCT/IB2022/000824 IB2022000824W WO2023199084A2 WO 2023199084 A2 WO2023199084 A2 WO 2023199084A2 IB 2022000824 W IB2022000824 W IB 2022000824W WO 2023199084 A2 WO2023199084 A2 WO 2023199084A2
Authority
WO
WIPO (PCT)
Prior art keywords
data
symbolic
blocks
anatomic
data blocks
Prior art date
Application number
PCT/IB2022/000824
Other languages
English (en)
Other versions
WO2023199084A3 (fr
Inventor
Matthew A. MOLENDA
Original Assignee
Mofaip, Llc
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Mofaip, Llc filed Critical Mofaip, Llc
Priority to PCT/IB2023/000535 priority Critical patent/WO2024023584A2/fr
Publication of WO2023199084A2 publication Critical patent/WO2023199084A2/fr
Publication of WO2023199084A3 publication Critical patent/WO2023199084A3/fr

Links

Classifications

    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
    • G16H10/00—ICT specially adapted for the handling or processing of patient-related medical or healthcare data
    • G16H10/60—ICT specially adapted for the handling or processing of patient-related medical or healthcare data for patient-specific data, e.g. for electronic patient records

Definitions

  • This invention relates to medical systems, and more particularly, to using symbolic delimiters and symbolic definitions to automatically categorize and collate data blocks to write rich stories about a file or area of a medical record that can be analyzed and retrieved by a search engine, artificial intelligence, or machine learning.
  • the present invention includes a method and system for data collation, retrieval, organization, analysis, and summarization of health data by utilizing order-agnostic symbol delimited data linked to standardized symbols, such as emojis or unicode characters.
  • a file name and metadata building method uses a data-blocking engine with low-character-count, symbolic delimiters automatically applied to each data field.
  • the data blocks are orderless and structureless, with no header requirements. Similar to physical construction blocks, the digital data blocks can be constructed, built upon, deconstructed, rearranged, or modified.
  • the data blocks build digital foundations on which artificial intelligence (Al) and/or machine learning can gather, collate, modify, and serve language agnostic, database agnostic, and platform agnostic data for individual patients or for populations, such as in a research search engine that retrieves automatically de-identified datasets of data tagged as “OK” to use in research, among other tags, solving privacy concerns at the same time.
  • Artificial intelligence Al
  • machine learning can gather, collate, modify, and serve language agnostic, database agnostic, and platform agnostic data for individual patients or for populations, such as in a research search engine that retrieves automatically de-identified datasets of data tagged as “OK” to use in research, among other tags, solving privacy concerns at the same time.
  • a research search engine (a “reSearch engine”) is also included in this invention, which for individual patients can retrieve specific health data from a plurality of databases, file formats, and bookmarks; and for populations can retrieve automatically de-identified datasets of data tagged as “OK” to use in research among other tags, solving privacy concerns at the same time.
  • Photos and other attachments can automatically be added to these categorizations and defined anatomic sites, and the metadata can be combined from the encounter, the anatomic site, and the additional metadata.
  • some cameras use geographic GPS coordinates in (Exchangeable Image File Format (EXIF)) a photo’s metadata to document the location of a photo.
  • EXIF Exchangeable Image File Format
  • the defined anatomic sites and the encounter data blocks are automatically combined with the geographic GPS data from the photos, and the data blocks can further be used to triage wartime injuries to the most appropriate medical outpost. It is contemplated that burns, percentage body surface area, gunshot wounds, chemical injuries are examples of critical data points that need to be communicated quickly, perhaps initially even over radio communication with low data bandwidth methods.
  • injuries are documented and communicated in any coded, linguistic, or symbolic language and automatically triaged based on anatomic sites of involvement, body surface area, injury type, injury intensity, injury count, injury distribution, and geographic GPS location. It is contemplated that the anatomic site encoding system is even customizable and encryptable, so even if intercepted, the coded communications would need a decoder to make sense of it.
  • the present invention applies a vocabulary builder and a site naming sequence configuration through artificial neural networks to break down anatomic site descriptions into data blocks including site name, laterality, prefixes, suffixes, enhanced modifiers to describe direction, custom descriptions and triangulations, automatic relationship descriptions with magnitude modifiers, code sequences, translations, synonyms, groupings, symbolic references, crossmappings, and other metadata.
  • these are some of the “anatomic components” or data blocks of the anatomic site name, and the present invention can arrange these in customized ways based on user preference or language.
  • Language inputs can also be verbal or spoken and translated into a standardized anatomic site name and visualization.
  • Automatic language modification also applies language specific considerations, such as changing laterality endings for masculine vs. feminine terms (izquierdo vs izquierda); or to display considerations for languages that read from right-to- left, such as Hebrew and Arabic.
  • This combined with applying neural networks to data block translations of the user interface, symptoms, morphologies, durations, numbers/alphabets, descriptions, diagnoses, diagnoses extensions, tags, visualizations, legends, and all other components of the software engine allow of the present invention for automatic, enhanced translation of the entire medical encounter, with or without anatomy visualizations and images.
  • the anatomic site, standardized anatomy codes, names, and symbols, patient data, diagnosis data, encounter data, tags, and other data can be used to generate language-agnostic file naming, grouping, and exporting function with optional universal symbolic low-character-count delimiters to automatically write a language agnostic story about files, documentation blocks, bookmarks within patient charts, labeled specimens, photos, attachments, links, and other metadata.
  • This symbolic delimited and defined data can be truncated in a file name, encrypted into static or evolving QR codes (or other codes with or without encryption), stored in exported file metadata, exported to a database or file wrapper (such as a Digital Imaging and Communications in Medicine (DICOM) wrapper), filtered, searched, de-identified, encoded, and tagged.
  • data blocks are combined and separated with meaningful delimiters, such as symbolic delimiters like universally translatable emojis, into an order-independent, structureless, meaningful story.
  • the invention writes a “novel” about the file using data blocks that does not have to fit into an EHR or other defined data structure (since the invention is database independent and agnostic, and data structure independent), and include that “novel” in the filenames, file metadata, or both, or include that novel as bookmarks within other records, thus automatically creating a filter and data target point.
  • a reSearch engine used for research on individual patients or populations, can retrieve identifiable or scrubbed (de-identified) health data based on the data blocks.
  • Al collates these data blocks to put together a history or timeline for an individual patient related to a specific anatomic region of interest, by using a single anatomic site or category, or a group of anatomic sites or categories, or other non-anatomy data blocks. Still other benefits and advantages of the invention will become apparent to those skilled in the art to which it pertains upon a reading and understanding of the following detailed specification.
  • FIG. 1 is an omnidirectional data model that illustrates the capabilities of the data block engine.
  • FIG. 2 is a screenshot showing map synchronization and pin-level data containing symbolic tags and categorizations.
  • FIG. 3 is the screenshot of FIG. 2 with portions translated to Chinese and added NovelFileTM.
  • FIG. 4 is a screenshot showing options to customize the anatomic site name sequence configuration and the data blocks within a file name builder.
  • FIG. 5 is a screenshot of a modal view of a thumbnail image and its accompanying symbol delimited and symbol defined file name.
  • FIG. 6 is a screenshot of a coordinated anatomy data in correspondence with a color- coded legend; and symbolic definitions for the anatomic site group.
  • FIG. 7 is a legend of representative symbolic searches within a reSearch engine.
  • FIG. 8 is a legend of representative application examples to re-create pins on regions of interest
  • FIG. 9 is a screenshot of an artificial intelligence collated patient history in an anatomic region of interest made possible by data blocks.
  • FIG. 1 depicts an omnidirectional data model 330 where uncoordinated 332 and coordinated 334 anatomy data 336, non-anatomy data 338, geographic data 342, tags 344, and data buckets 340 containing photos, attachments, and links enable the capabilities 350 of the data block engine 352.
  • an uncoordinated data 332 example would be a linguistic description like “right ear.”
  • Coordinated data 334 examples would be the position of a pin on a visual anatomic map, or selection of an anatomic region of interest on an image that has an anatomic map.
  • a membership and category 346 example would be the “right ear” belonging to the “head and neck” in a hierarchical relationship, and the “auditory system” in a functional system.
  • Data buckets 340 overlap with anatomy data 336, tags 344, non-anatomy data 338, and communicate with geographic 342 and linguistic language 348, and contain data such as photographs, attachments such as pdf reports, and links such as hyperlinks to a specific bookmark in a medical record.
  • a geographic coordinated data 342 example includes the geographic GPS coordinates of a photograph captured by a GPS enabled phone with a camera that stores the GPS data in the photograph.
  • tags 344 are used to supply data block tags to the data block engine 352, with an example being “OK” tag symbolized by the “OK emoji” signifying that the patient has approved to allow their data to be used in research
  • the data block engine 352 has capabilities 350 in record generation, retrieval, deidentification, translation to any coded, linguistic or symbolic language, sequencing timed data into timelines, form generation, visualization of healthcare data (such as re-creating a point on an anatomic map to show the location of a disease, or selecting a region of interest on an anatomic map to search the research engine), tracking of records and health data including monitoring of populations, and filtering and scrubbing structureless, orderless health data to deliver the required results.
  • the data blocks flow in all directions, making it omnidirectional. It is contemplated that all of the data are connected to a neural network that can be used by Al and/or machine learning.
  • FIG. 2 depicts a screen shot with dynamic patient and encounter synchronization of data to map.
  • Real-time synchronized map data 60 includes text like patient name, symbolic translations like for male sex as a symbolic definition, and symbolic delimiters like the birthday cake emoji to represent date of birth and the calendar emoji to represent encounter date. It is contemplated that some symbols such as those for patient sex stand on their own as symbolic definitions, and do not need additional data, while others like the birthday cake emoji would typically be tied to other data like a date, and thus the date of birth would be delimited by the birthday cake emoji in an order independent and structureless string of data blocks.
  • the cursor 14 is shown over the right (superior) paramedian forehead with a corresponding color-coded legend 12 to the bottom left of the figure.
  • two pins are dropped, with a brown representing cryosurgery procedure to a diagnosis of an inflamed seborrheic keratosis, and a red A” representing a shave biopsy procedure on the diagnosis of a neoplasm of uncertain behavior of skin (2F72.Y), with the representing the precise point of the biopsy procedure.
  • the OR code 62 that contains dynamic anatomic address (a multidimensional and reproducible anatomic site that is trackable at different time points) information is partially redacted, but it contains patient data, re-creation data, anatomic site data, and other data blocks.
  • This embodiment also depicts the icons for changing procedures, diagnoses, list memberships, and pins to distribution segments, along with a dropdown that shows the pin preview and description of the pin should the user wish to automatically change multiple components of the dynamic anatomic address but keep the selected anatomic site and site descriptions constant.
  • the user could easily change from “. A” representing a shave biopsy procedure on the diagnosis of a neoplasm of uncertain behavior of skin (2f72.y) to a diagnosis of melanoma, thus dynamically changing some of the data blocks related to this pin and the other data attached to the pin, such as the file name diagnosis.
  • the diagnosis component and other components like the pin description change dynamically in the dynamic anatomic address, but the anatomic site information and visualization of the location remains static.
  • the user is a computer, and for example, the morphology data blocks can be detected from the images and added automatically. It is contemplated that other features could be added into different parts of the application from automatic image detection, such as skin type and skin tone (not shown here). It is contemplated that skin subtyping and sub-toning can also be done (not shown here), such as a customized skin type and tone that has a much more granular scale and based on dynamic anatomic address average calculations and confidence intervals, and other features such as age.
  • sun exposed skin on the face may be more darkly pigmented in older patients who spent a lot of time in the sun, and the skin on their inner arm or buttock where there is typically less sun exposure would provide a more accurate skin type and tone measurement for the overall patient, and is therefore weighted more heavily in the skin subtype and subtone calculations from the different skin tone and skin type data blocks based on anatomic locations.
  • each photo and attachment are joined into a collated bucket for this pin and dynamic anatomic address which has a symbolic definition, and are tagged and notated with other symbolic tags, definitions, and delimiters. It is contemplated that a symbolic definition can also serve as a symbolic delimiter, with the “tag” emoji being an example of this.
  • the “tag” emoji symbolically delimits other symbolic defined tags, such as a magnifying glass emoji to tag a photo as a closeup, or an “OK” emoji tagging the image as approved for use in research. Thumbnails are shown for each photo and attachment in the currently joined bucket and clicking on it will bring up a modal to view the thumbnails as navigable and editable images in larger size (shown in FIG. 3 ).
  • Symbolic emoji tags are shown for brevity, but these tags are also translated with linguistic parallels in any coded or linguistic language using a neural network. Photo notes and link descriptions are only translated based on user preference. Links are also categorized with symbolic delimiters and tagged with language agnostic symbolic and translatable text tags.
  • FIG. 3 depicts the screenshot of FIG. 2 with portions translated to Chinese.
  • This embodiment has been simultaneously and real-time translated to Chinese 244 in all areas except select manual inputs still in English 240.
  • This embodiment also shows a symbolically delimited and symbolically defined NovelFileTM filename 249 for the image 247 in bucket belonging to this pin at this dynamic anatomic address.
  • the NovelFileTM filename 249 tells a language agnostic story about the pin by using symbols, and order does not matter because of the symbolic delimitation.
  • the symbolic story telling file name can also be included with the metadata of the exported or saved file, based on user preference.
  • the exemplar photo 247 in this embodiment belongs to the dynamic anatomic address and pin and is editable and able to be marked up.
  • the same photos or attachments or links or forms can simultaneously belong to other dynamic anatomic addresses, such as the other * pin (obscured in this screenshot by the translated photo modal 248) and exist in multiple buckets and dynamic anatomic addresses simultaneously.
  • Some useful examples of multimedia have multiple dynamic anatomic addresses are illustrated by the photo in this figure.
  • On the map there is a shave biopsy to rule out melanoma (. A, red); and an inflamed seborrheic keratosis (*, brown) treated with cryosurgery right above it. the recent surgical scar above that can also have details automatically pulled about that surgery based on its data blocks derived from its dynamic anatomic address (not shown) and other data blocks, into a shadow chart or timeline view, for example.
  • This photo 247 can belong to all three dynamic anatomic addresses and have differentiated, and the same data blocks related to the anatomic locations, in the ongoing example simultaneously, because it is relevant to each. It is contemplated that timeline views, data collation, anatomic region filtering, communication, translation, and future documentation into the correct buckets are all achieved through the dynamic anatomic addressing and data blocking models applied by the described software engines and artificial neural networks working together. Also shown in this embodiment is a translated dropdown that allows for conversion of dynamic anatomic address component of a pin to be converted to a translated distribution segment that is visually represented on the map (not shown), and has other differences such as diagnosis, but maintain its pin position through an invisible anchor pin, and bucket contents.
  • FIG. 4 is a screenshot showing options to customize the shown data with options to automatically and dynamically order, categorize, and show data 270 like coded translations, symbolic categorizations of anatomy and health data, and symbolic delimiters and symbolic definitions related to data blocks 274.
  • the naming sequence can individually toggle data blocks 272 related to anatomic site name components, code translations and options, optional separators, and symbolic categorizations and symbolic definitions such as emoji groups in this exemplar. As shown in FIG.
  • the images and attachments can be customized 274 to a desired configuration as well.
  • the images and attachment of present invention uses symbolic delimiters (such as a birthday cake emoji for date of birth) and definitions or categorizations (such as patient sex) automatically, allowing for automatic storytelling, data aggregation, and data filtering through filenames, file metadata, bookmarks, links, file wrappers, and other digital repositories for data and metadata.
  • anatomic site and dynamic anatomic address component data blocks can be represented simultaneously in multiple ways in the story, including but not limited to emoji groups as a symbolic category or definition, code strings, linguistic description and categories, separated laterality and site name components, test ID, pin ID, pin coordinates, pin angles and deviations, pin relationships, pin level of hierarchy, pin organ system, pin anatomy system, site relationships, site segmentations, site level of hierarchy, site organ system, site anatomy system, and other representations.
  • FIG. 5 This embodiment shows an example file name 246 for a photo 247 that is named with symbolic delimiters, health data including anatomic site data, and symbolic definitions.
  • unique symbolic characters emojis and unicode characters
  • file names may have size limitations (such as 256 characters), the data blocks may be truncated in the filename and simultaneously placed into metadata fields for the file which have larger data storage capabilities.
  • data blocks can document changes and histories over time, such as when some data blocks change at different time points.
  • Each symbolic delimited and symbolic defined component of the exemplar file name 246 is broken down in a table 245. It is contemplated that even more data blocks can be stored within the file metadata or within file wrappers, such as DICOM wrappers.
  • symbolic delimiters and symbolic definitions can optionally use standard character-based delimiters and text abbreviations or descriptions of the symbols for legacy systems that do not support all the unicode and emoji symbols (not shown), so they may be used in parallel with legacy datasets.
  • platform agnosticism single- or low-character count categorizations that are stored in the filename and/or file metadata, or within sections or progress notes or reports, allowing this concept to work regardless of the electronic health record system or database in use
  • order agnosticism the order of the data blocks does not matter
  • (3) allowing for structureless and orderless data (3) no database is required
  • language agnosticism standardized symbolic delimiters and symbolic definitions confer meaning regardless of language, and construct a human readable story just with the data blocks
  • modifiable as additional information becomes available, additional data blocks can be added to existing records without actually altering the integrity of the record.
  • the diagnosis might have been neoplasm of uncertain behavior (2F72.Y) at the time of the photo; but after biopsy it was determined to be basal cell carcinoma, nodular type (2C32-XH2CR0); (6) data collation, aggregation and de-identification: select data blocks can be searched for, collated and/or aggregated, and automatically de-identified on an as-needed basis simply by removing the health data blocks that contain Protected Health Information (PHI).
  • PHI Protected Health Information
  • HIPAA Health Insurance Portability and Accountability Act
  • formulaic searches through health data blocks can enable filtered, de-idenlified, aggregated research data that has been preapproved for use in research (i.e., tagged as “OK” to indicate patient approval;.
  • Granular tagging and data block application of dynamic anatomic addresses, or individual photos or tiles, can address patient privacy concerns automatically by only allowing appropriately tagged content into research search results.
  • Standardized and symbol-delimited health data blocks create a foundation for a research search engine (a “reSearch engine”) for healthcare research data.
  • the included formulaic search example searches, aggregates, and delivers de-identified photos symbol for "camera”) for all male patients with basal cell carcinoma (2C32) on the nose between ages 30-40 (calculated as age at time of encounter, by encounter date minus birthdate) by searching and modifying the health metadata blocks that have been tagged as “OK to use in resear It is contemplated that the order of the formulaic query does not matter. It is further contemplated that the files are automatically scrubbed of identifiable patient information and delivered to the researcher with relabeled data blocks file naming protocols. Standardized and symbol-delimited and symbol defined data blocks represent a new frontier in medicine and research and are the key to innumerable new clinical and research capabilities.
  • Health data block labeling and tagging creates foundations for Al-assisted collation, retrieval, organization, and summarization of health data and records.
  • multiple diagnoses and treatments can be linked simultaneously to a dynamic anatomic address and different anatomic regions and dynamic anatomic addresses through their data block components.
  • machine learning and Al generate a history about a region of interest by using health data blocks including blocks from dynamic anatomic addresses, diagnoses, treatments, and dates in an area of interest.
  • the health data blocks can be combined, deconstructed, and built upon, and exist simultaneously in unlimited dynamic anatomic addresses and in unlimited block data structures, including a block chain.
  • a context aware, automatically generated history includes direct links to relevant photos or other imaging (like X-rays, ultrasounds, etc.), documents, forms, reports, and data collated from and made possible by the data blocks.
  • FIG 6. depicts a screenshot illustrating a selected anatomic site 18 at the left dorsal proximal interphalangeal joint of little finger, coordinated anatomy data 16 is shown in correspondence with a color-coded legend 12; and symbolic definitions 19 for the anatomic site group 20 are shown for each dimension, level and layer in an English embodiment at the end of each anatomic site description. It is contemplated that the symbolic delimited and symbolic defined emoji categorization captures all of anatomic site data in a granular way simultaneously regardless of language or code set through a cross-mapping data set, neural networks, and data block engine.
  • FIGs. 7 and 8 depicts exemplar emoji searches 65 and application examples 66 with their unicode backup. While emoji characters may render differently on different systems, they are backed up by unicode and confer the same meaning nearly universally. Some emojis may even display differently in different countries based on inherent emoji localization features.
  • the birthday cake emoji may display differently in Japan than in the US; and the birthday cake emoji is closer to a global universal symbol for “date of birth” than “DOB.”
  • “fecha de nacimiento” means “date of birth” or “birthday” and may be abbreviated as FDN in a record system, thus it is contemplated that these different data headers in different languages create the need for manual data cross-mapping in multi-national research an health data.
  • FIG. 9 depicts an exemplar single Al, context aware, automatically generated history includes direct links to relevant photos or other imaging (like X-rays, ultrasounds, etc.), documents, forms, reports, and health data.
  • the underlined sections 64 shown in the present embodiment represent hyperlinks directly to the relevant notes, reports, photos, and other health record information.
  • Al and machine learning generated single patient history in English collated, organized, and presented from data blocks in the selected anatomic site. It is contemplated that data are organized in a timeline based on the selected anatomic site 52, and account for regional anatomic sites as well (with information delivered for the left cheek, which is automatically included in the specialty context and the anatomy context).
  • automatic cross-links are generated to procedure summaries, photos, results, prescriptions, and other health data associated with the dynamic anatomic addresses and data blocks, represented as blue hyperlinks in the figure. It is further contemplated that progression, transformation, recurrence, growth, resolution, and other changes can be tracked, documented, and analyzed automatically because of the platform that includes the dynamic anatomic addressing method and the data blocking engine. It is further contemplated that symbolic delimitations and symbolic definitions and symbolic categorization in all systems form artificial neural networks that enable language agnostic, order agnostic, platform agnostic, modifiable, and targetable results, collation, and aggregation.
  • an example of a more global result, with deep anatomic sites includes applying the dynamic anatomic address method and data blocks, and the data block engine of this invention, to answer global questions that affect superficial, deep, or systems- based anatomy.
  • a standardized distribution tracking output that automatically segments the lungs in layered or three-dimensional space tracks how a respiratory virus affects different areas of the lungs with fibrosis, inflammation, hemorrhage, and other morphology features identified on medical imaging or biopsies.
  • the inflammatory profile is linked to different dynamic anatomic address areas of the lung where fluid samples were taken, providing dynamic collated answers to questions like: Is the inflammatory response different in the lower lung versus the upper lung?

Landscapes

  • Health & Medical Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Epidemiology (AREA)
  • General Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • Primary Health Care (AREA)
  • Public Health (AREA)
  • Medical Treatment And Welfare Office Work (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

La présente invention comprend un procédé de collation, de récupération, d'organisation, d'analyse et de récapitulation de données relatif à des données de santé en faisant appel à des données délimitées par des symboles et définies par des symboles agnostiques d'ordres liées à des symboles normalisés, tels que des émojis ou des caractères unicode. Un procédé de construction de nom de fichier et de métadonnées fait appel à un moteur de blocs de données comprenant un faible nombre de caractères, des délimiteurs symboliques et des définitions symboliques appliqués automatiquement à chaque champ de données. Les blocs de données sont sans ordre et sans structure, et sans aucune exigence d'en-tête. De manière similaire à des blocs de construction physiques, les blocs de données numériques peuvent être construits, bâtis, déconstruits, réarrangés ou modifiés. Les blocs de données construisent des fondations numériques sur lesquelles l'intelligence artificielle et/ou l'apprentissage automatique peuvent rassembler, collationner, modifier et desservir des données agnostiques de langage, des données agnostiques de bases de données et des données agnostiques de plateformes pour des patients individuels ou pour des populations, telles que dans un moteur de recherche lié à la recherche qui récupère des ensembles de données automatiquement anonymisés de données étiquetées en tant que « OK » à utiliser dans la recherche parmi d'autres étiquettes, résolvant ainsi en même temps les problèmes de confidentialité.
PCT/IB2022/000824 2021-12-10 2022-12-12 Blocs de données délimitées et définies par des symboles pour écrire des histoires de données riches destinées à être utilisées avec une intelligence artificielle WO2023199084A2 (fr)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/IB2023/000535 WO2024023584A2 (fr) 2022-07-26 2023-07-25 Systèmes et procédés utilisant des modèles et des cartes de langage et de vision multidimensionnels pour catégoriser, décrire, coordonner et suivre des données d'anatomie et de santé

Applications Claiming Priority (4)

Application Number Priority Date Filing Date Title
US202163265216P 2021-12-10 2021-12-10
US63/265,216 2021-12-10
US202263267269P 2022-01-28 2022-01-28
US63/267,269 2022-01-28

Publications (2)

Publication Number Publication Date
WO2023199084A2 true WO2023199084A2 (fr) 2023-10-19
WO2023199084A3 WO2023199084A3 (fr) 2024-03-21

Family

ID=88329122

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/IB2022/000824 WO2023199084A2 (fr) 2021-12-10 2022-12-12 Blocs de données délimitées et définies par des symboles pour écrire des histoires de données riches destinées à être utilisées avec une intelligence artificielle

Country Status (1)

Country Link
WO (1) WO2023199084A2 (fr)

Family Cites Families (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20010041992A1 (en) * 2000-03-10 2001-11-15 Medorder, Inc. Method and system for accessing healthcare information using an anatomic user interface
US7286997B2 (en) * 2002-05-07 2007-10-23 Cembex Care Solutions, Llc Internet-based, customizable clinical information system
US20060173858A1 (en) * 2004-12-16 2006-08-03 Scott Cantlin Graphical medical data acquisition system
US20080243550A1 (en) * 2007-04-02 2008-10-02 Yao Robert Y Method and system for organizing, storing, connecting and displaying medical information
US8311848B2 (en) * 2009-10-05 2012-11-13 Muthiah Subash Electronic medical record creation and retrieval system
US20150134361A1 (en) * 2013-11-08 2015-05-14 The Cleveland Clinic Foundation Graphical generation and retrieval of medical records

Also Published As

Publication number Publication date
WO2023199084A3 (fr) 2024-03-21

Similar Documents

Publication Publication Date Title
Wu et al. Comparison of chest radiograph interpretations by artificial intelligence algorithm vs radiology residents
US8311848B2 (en) Electronic medical record creation and retrieval system
Roque et al. A comparison of several key information visualization systems for secondary use of electronic health record content
EP2191399A1 (fr) Système et procédé d'analyse d'enregistrements de données électroniques
CN1839404A (zh) 将医疗信息计算机化和标准化的方法
JP2008506188A (ja) ジェスチャ・ベース報告方法およびシステム
US20170109502A1 (en) System and method for clinical trial candidate matching
KR101850772B1 (ko) 의료용 메타 데이터베이스 관리 방법 및 그 장치
US11630874B2 (en) Method and system for context-sensitive assessment of clinical findings
US20110164820A1 (en) Records Management System and Method
JP2024515534A (ja) 人工知能支援の画像解析のためのシステムおよび方法
JP2010211452A (ja) 症例画像検索装置、方法およびプログラム
Li et al. Digitization of medicine: how radiology can take advantage of the digital revolution
Huang et al. OVQA: A clinically generated visual question answering dataset
US20230368878A1 (en) Systems and methods using multidimensional language and vision models and maps to categorize, describe, coordinate, and track anatomy and health data
US20040167800A1 (en) Methods and systems for searching, displaying, and managing medical teaching cases in a medical teaching case database
WO2024023584A2 (fr) Systèmes et procédés utilisant des modèles et des cartes de langage et de vision multidimensionnels pour catégoriser, décrire, coordonner et suivre des données d'anatomie et de santé
JP7473314B2 (ja) 医療情報管理装置及び医療レポートのメタデータ付加方法
WO2023199084A2 (fr) Blocs de données délimitées et définies par des symboles pour écrire des histoires de données riches destinées à être utilisées avec une intelligence artificielle
JP2004348271A (ja) 治験データ出力装置、治験データ出力方法及び治験データ出力プログラム
Zhou et al. A Generalist Learner for Multifaceted Medical Image Interpretation
Traina et al. Integrating images to patient electronic medical records through content-based retrieval techniques
CN109840275B (zh) 一种医疗搜索语句的处理方法、装置和设备
Wang et al. Deep learning for automating the organization of institutional dermatology image stores
JP5682657B2 (ja) データベースシステム

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 22937332

Country of ref document: EP

Kind code of ref document: A2