MLchartDataset catalogue

Patent · US12027250B2 · B2 · US

Medical information processing apparatus and information processing method

(11) Publication number
US12027250B2
(21) Application number
16/768,081
(22) Filing date
2018-10-02
(30) Priority date
2017-12-05
(43) Publication date
2024-07-02
(45) Date of grant
2024-07-02
(51) IPC
A61B 1/00; A61B 5/02; A61B 90/00; A61M 16/01; B25J 9/16; G05B 13/02; G16H 15/00; G16H 20/40; G16H 30/20; G16H 30/40; G16H 40/67; G16H 50/20; G16H 50/70; G16H 70/20
(52) CPC
  • G16H Healthcare informatics, i.e. information and communication technology [ICT] specially adapted for the handling or processing of medical or healthcare data: 20/40, 15/00, 30/20, 30/40, 40/67, 50/20, 50/70, 70/20
  • A61B Diagnosis; surgery; identification: 1/00009, 34/10, 5/02042, 90/37
  • A61M Devices for introducing media into, or onto, the body; devices for transducing body media or for taking media from the body; devices for producing or ending sleep or stupor {}: 16/01
  • B25J Manipulators; chambers provided with manipulation devices: 9/1697
  • G05B Control or regulating systems in general; functional elements of such systems; monitoring or testing arrangements for such systems or elements: 13/0265, 2219/45117
(73) Assignee
Sony Olympus Medical Solutions Inc
(72) Inventors
Mitsuaki Shiraga
(54) Title
Medical information processing apparatus and information processing method
(57) Abstract

The present disclosure provides a medical information processing apparatus that comprises: a selection unit that selects, from among candidate data including surgery data acquired during surgery, target data corresponding to selection conditions that include a condition relating to a patient attribute, a condition relating to a surgical-procedure type, and a condition relating to a disease type; an extraction unit that detects a feature from the selected target data and extracts, from the target data, feature data corresponding to the detected feature; and an editing processing unit that edits the extracted feature data, wherein the extraction unit extracts at least a medical image as the feature data.

Full text
View on Google Patents

Claims (20)

  1. A medical information processing system, comprising: training circuitry configured to: train a deep learning model on data corresponding to selection conditions that include conditions relating to patient attributes including at least one attribute different from a subject of a medical image, conditions relating to surgical-procedure types, and conditions relating to disease types; and processing circuitry configured to: select, from candidate data including surgery data acquired during surgery, target data corresponding to the selection conditions that include conditions relating to patient attributes including at least one attribute different from a subject of a medical image, conditions relating to surgical-procedure types, and conditions relating to disease types; detect a feature from the selected target data; extract feature data corresponding to the detected feature using the deep learning model, from the target data, the feature data including at least the medical image; edit the extracted feature data to generate protocol data for a case related surgical procedure based on previous operations and patients having similar attributes to a current patient; provide intraoperative navigation based on the protocol data; and monitor the surgery relative to the case related surgical procedure, wherein on condition that the surgery deviates from the case related surgical procedure to a modified surgical procedure, the processing circuitry is configured to extract feature data corresponding to the modified related surgical procedure, generate updated protocol data for the modified related surgical procedure, provide intraoperative navigation based on the updated protocol data, and feedback the updated protocol data to the training circuitry.
  2. The medical information processing system according to claim 1, wherein the target data includes a medical captured image captured by an imaging device during surgery, and the medical image is the medical captured image.
  3. The medical information processing system according to claim 1, wherein the target data includes a medical illustration image, and the medical image is a medical illustration image.
  4. The medical information processing system according to claim 1, wherein the target data includes non-image data, and the processing circuitry is configured to detect a feature from the non-image data.
  5. The medical information processing system according to claim 1, wherein the target data includes sound collected by a sound collection device during surgery, and the processing circuitry also extracts sound as the feature data.
  6. The medical information processing system according to claim 1, wherein the target data includes text which is input by an operation to an operating device, and the processing circuitry also extracts text as the feature data.
  7. The medical information processing system according to claim 1, wherein the processing circuitry is configured to detect a state-related feature and/or an action-related feature from the target data.
  8. The medical information processing system according to claim 7, wherein the processing circuitry is configured to detect, as the state-related feature, a point in time in the target data when a predetermined state is detected from the target data, and detect, as the action-related feature, a point in time in the target data when a predetermined action is detected from the target data.
  9. The medical information processing system according to claim 1, wherein the processing circuitry is configured to generate the protocol data for the surgical procedure from a plurality of cases.
  10. The medical information processing system according to claim 1, wherein the processing circuitry is configured to cause a display to display the medical image corresponding to the surgery data acquired during surgery.
  11. The medical information processing system according to claim 10, wherein the processing circuitry is configured to: select the target data by setting the selection condition based on the surgery data acquired during surgery, extract medical images by detecting features from the selected target data, and cause the display to display the medical image corresponding to the surgery data, among the extracted medical images.
  12. The medical information processing system according to claim 11, wherein the processing circuitry is configured to issue a notification regarding the grounds for extracting the medical images displayed on the display.
  13. The medical information processing system according to claim 11, wherein the medical image corresponding to the surgery data and displayed on the display correspond to a navigation image during surgery.
  14. The medical information processing system according to claim 1, wherein the processing circuitry is configured to: control an arm capable of carrying a medical instrument, specify a surgery progress status based on the feature data corresponding to the surgery data acquired during surgery, and control the arm such that the medical instrument corresponding to the specified progress status is carried to a predetermined position corresponding to the specified progress status.
  15. The medical information processing apparatus system according to claim 1, wherein the processing circuitry is configured to provide the interoperative navigation based on the updated protocol data includes displaying a medical image of a next scene that corresponds to the modified surgical procedure.
  16. An information processing method executed by a medical information processing system, the method comprising: training a deep learning model on data corresponding to selection conditions that include a condition relating to patient attributes, wherein the patient attributes include at least one attribute independent of a subject, a condition relating to a surgical-procedure type, and a condition relating to a disease type; selecting, from among candidate data including surgery data acquired during surgery, target data corresponding to selection conditions that include the condition relating to patient attributes, wherein the patient attributes include at least one attribute independent of a subject of a medical image, a condition relating to a surgical-procedure type, and a condition relating to a disease type; detecting a feature from the selected target data and extracting, from the target data, feature data corresponding to the detected feature, the feature data including at least the medical image; editing the extracted feature data using the deep learning model to generate protocol data for a case related surgical procedure based on previous operations and patients having similar attributes to a current patient; providing intraoperative navigation based on the protocol data; monitoring the surgery relative to the case related surgical procedure; and generating a modified surgical procedure, including extracting feature data corresponding to the modified surgical procedure deviating from the case related surgical procedure, generating updated protocol data for the modified surgical procedure, providing intraoperative navigation based on the updated protocol data, and feeding back the updated protocol data to the training.
  17. The method as claimed in claim 16, wherein providing the interoperative navigation based on the updated protocol data includes displaying a medical image of a next scene that corresponds to the modified surgical procedure.
  18. The method according to claim 16, further comprising: controlling an arm capable of carrying a medical instrument, specifying a surgery progress status based on the feature data corresponding to the surgery data acquired during surgery, and controlling the arm such that the medical instrument corresponding to the specified progress status is carried to a predetermined position corresponding to the specified progress status.
  19. The method according to claim 16, wherein the target data includes non-image data, and further comprising detecting a feature from the non-image data.
  20. The method according to claim 16, wherein the target data includes sound collected by a sound collection device during surgery, and further comprising extracting sound as the feature data.

Description

The present disclosure relates to a medical information processing apparatus and an information processing method.

In surgery, various data generated by various surgery-related devices (hereinafter may be referred to as “surgery data” collectively or to denote individual data items) is recorded on a recording medium. Under these circumstances, techniques for efficiently managing surgery data have been developed. Techniques for efficiently managing surgery data include the technique disclosed in Patent Literature 1 below, for example.

Furthermore, techniques for analyzing endoscope images obtained by an endoscope have been developed. The technique disclosed in Patent Literature 2 below, for example, may be cited as “a technique for establishing categories corresponding to pathologic diagnoses by using results obtained by extracting cell nuclei contained in an endoscope image to measure features and using results obtained by performing texture analysis on the whole endoscope image”.

As mentioned earlier, surgery data is generated by various devices in surgery, and the surgery data is recorded on a recording medium. For example, in a case where a health worker such as a physician is attempting to make secondary use of surgery data stored on a recording medium, the health worker needs to perform an enormous amount of work in which the health worker “reproduces all the surgery data and uses editing software or the like to crop a snapshot” and “compiles surgery data and medical illustration images such as autopsy diagrams to create a workflow for a case surgery protocol”, and so forth.

Citations (28)

  • US20160278870A1
  • US20030216836A1
  • JP3104051U
  • JP2006043209A
  • WO2006077798A1
  • US20100198402A1
  • US20110046476A1
  • US20090088634A1
  • US20170345155A1
  • US20120183191A1
  • JP2011167301A
  • US20140287393A1
  • US20140343586A1
  • US20140185888A1
  • JP2016042982A
  • JP2016154810A
  • WO2016200887A1
  • US20160381256A1
  • JP2017047022A
  • WO2017083768A1
  • US20170143284A1
  • US20170151027A1
  • US20190046813A1
  • US20180065248A1
  • US20180137244A1
  • US20180233222A1
  • US9788907B1
  • US20190133693A1
Record as JSON
{
  "publication_number": "US12027250B2",
  "country": "US",
  "kind": "B2",
  "title": "Medical information processing apparatus and information processing method",
  "abstract": "The present disclosure provides a medical information processing apparatus that comprises: a selection unit that selects, from among candidate data including surgery data acquired during surgery, target data corresponding to selection conditions that include a condition relating to a patient attribute, a condition relating to a surgical-procedure type, and a condition relating to a disease type; an extraction unit that detects a feature from the selected target data and extracts, from the target data, feature data corresponding to the detected feature; and an editing processing unit that edits the extracted feature data, wherein the extraction unit extracts at least a medical image as the feature data.",
  "claims": [
    "1. A medical information processing system, comprising: training circuitry configured to: train a deep learning model on data corresponding to selection conditions that include conditions relating to patient attributes including at least one attribute different from a subject of a medical image, conditions relating to surgical-procedure types, and conditions relating to disease types; and processing circuitry configured to: select, from candidate data including surgery data acquired during surgery, target data corresponding to the selection conditions that include conditions relating to patient attributes including at least one attribute different from a subject of a medical image, conditions relating to surgical-procedure types, and conditions relating to disease types; detect a feature from the selected target data; extract feature data corresponding to the detected feature using the deep learning model, from the target data, the feature data including at least the medical image; edit the extracted feature data to generate protocol data for a case related surgical procedure based on previous operations and patients having similar attributes to a current patient; provide intraoperative navigation based on the protocol data; and monitor the surgery relative to the case related surgical procedure, wherein on condition that the surgery deviates from the case related surgical procedure to a modified surgical procedure, the processing circuitry is configured to extract feature data corresponding to the modified related surgical procedure, generate updated protocol data for the modified related surgical procedure, provide intraoperative navigation based on the updated protocol data, and feedback the updated protocol data to the training circuitry.",
    "2. The medical information processing system according to claim 1, wherein the target data includes a medical captured image captured by an imaging device during surgery, and the medical image is the medical captured image.",
    "3. The medical information processing system according to claim 1, wherein the target data includes a medical illustration image, and the medical image is a medical illustration image.",
    "4. The medical information processing system according to claim 1, wherein the target data includes non-image data, and the processing circuitry is configured to detect a feature from the non-image data.",
    "5. The medical information processing system according to claim 1, wherein the target data includes sound collected by a sound collection device during surgery, and the processing circuitry also extracts sound as the feature data.",
    "6. The medical information processing system according to claim 1, wherein the target data includes text which is input by an operation to an operating device, and the processing circuitry also extracts text as the feature data.",
    "7. The medical information processing system according to claim 1, wherein the processing circuitry is configured to detect a state-related feature and/or an action-related feature from the target data.",
    "8. The medical information processing system according to claim 7, wherein the processing circuitry is configured to detect, as the state-related feature, a point in time in the target data when a predetermined state is detected from the target data, and detect, as the action-related feature, a point in time in the target data when a predetermined action is detected from the target data.",
    "9. The medical information processing system according to claim 1, wherein the processing circuitry is configured to generate the protocol data for the surgical procedure from a plurality of cases.",
    "10. The medical information processing system according to claim 1, wherein the processing circuitry is configured to cause a display to display the medical image corresponding to the surgery data acquired during surgery.",
    "11. The medical information processing system according to claim 10, wherein the processing circuitry is configured to: select the target data by setting the selection condition based on the surgery data acquired during surgery, extract medical images by detecting features from the selected target data, and cause the display to display the medical image corresponding to the surgery data, among the extracted medical images.",
    "12. The medical information processing system according to claim 11, wherein the processing circuitry is configured to issue a notification regarding the grounds for extracting the medical images displayed on the display.",
    "13. The medical information processing system according to claim 11, wherein the medical image corresponding to the surgery data and displayed on the display correspond to a navigation image during surgery.",
    "14. The medical information processing system according to claim 1, wherein the processing circuitry is configured to: control an arm capable of carrying a medical instrument, specify a surgery progress status based on the feature data corresponding to the surgery data acquired during surgery, and control the arm such that the medical instrument corresponding to the specified progress status is carried to a predetermined position corresponding to the specified progress status.",
    "15. The medical information processing apparatus system according to claim 1, wherein the processing circuitry is configured to provide the interoperative navigation based on the updated protocol data includes displaying a medical image of a next scene that corresponds to the modified surgical procedure.",
    "16. An information processing method executed by a medical information processing system, the method comprising: training a deep learning model on data corresponding to selection conditions that include a condition relating to patient attributes, wherein the patient attributes include at least one attribute independent of a subject, a condition relating to a surgical-procedure type, and a condition relating to a disease type; selecting, from among candidate data including surgery data acquired during surgery, target data corresponding to selection conditions that include the condition relating to patient attributes, wherein the patient attributes include at least one attribute independent of a subject of a medical image, a condition relating to a surgical-procedure type, and a condition relating to a disease type; detecting a feature from the selected target data and extracting, from the target data, feature data corresponding to the detected feature, the feature data including at least the medical image; editing the extracted feature data using the deep learning model to generate protocol data for a case related surgical procedure based on previous operations and patients having similar attributes to a current patient; providing intraoperative navigation based on the protocol data; monitoring the surgery relative to the case related surgical procedure; and generating a modified surgical procedure, including extracting feature data corresponding to the modified surgical procedure deviating from the case related surgical procedure, generating updated protocol data for the modified surgical procedure, providing intraoperative navigation based on the updated protocol data, and feeding back the updated protocol data to the training.",
    "17. The method as claimed in claim 16, wherein providing the interoperative navigation based on the updated protocol data includes displaying a medical image of a next scene that corresponds to the modified surgical procedure.",
    "18. The method according to claim 16, further comprising: controlling an arm capable of carrying a medical instrument, specifying a surgery progress status based on the feature data corresponding to the surgery data acquired during surgery, and controlling the arm such that the medical instrument corresponding to the specified progress status is carried to a predetermined position corresponding to the specified progress status.",
    "19. The method according to claim 16, wherein the target data includes non-image data, and further comprising detecting a feature from the non-image data.",
    "20. The method according to claim 16, wherein the target data includes sound collected by a sound collection device during surgery, and further comprising extracting sound as the feature data."
  ],
  "description_excerpt": "The present disclosure relates to a medical information processing apparatus and an information processing method.\n\nIn surgery, various data generated by various surgery-related devices (hereinafter may be referred to as “surgery data” collectively or to denote individual data items) is recorded on a recording medium. Under these circumstances, techniques for efficiently managing surgery data have been developed. Techniques for efficiently managing surgery data include the technique disclosed in Patent Literature 1 below, for example.\n\nFurthermore, techniques for analyzing endoscope images obtained by an endoscope have been developed. The technique disclosed in Patent Literature 2 below, for example, may be cited as “a technique for establishing categories corresponding to pathologic diagnoses by using results obtained by extracting cell nuclei contained in an endoscope image to measure features and using results obtained by performing texture analysis on the whole endoscope image”.\n\nAs mentioned earlier, surgery data is generated by various devices in surgery, and the surgery data is recorded on a recording medium. For example, in a case where a health worker such as a physician is attempting to make secondary use of surgery data stored on a recording medium, the health worker needs to perform an enormous amount of work in which the health worker “reproduces all the surgery data and uses editing software or the like to crop a snapshot” and “compiles surgery data and medical illustration images such as autopsy diagrams to create a workflow for a case surgery protocol”, and so forth.",
  "cpc": [
    "G16H 20/40",
    "A61B 1/00009",
    "A61B 34/10",
    "A61B 5/02042",
    "A61B 90/37",
    "A61M 16/01",
    "B25J 9/1697",
    "G05B 13/0265",
    "G05B 2219/45117",
    "G16H 15/00",
    "G16H 30/20",
    "G16H 30/40",
    "G16H 40/67",
    "G16H 50/20",
    "G16H 50/70",
    "G16H 70/20"
  ],
  "ipc": [
    "A61B 1/00",
    "A61B 5/02",
    "A61B 90/00",
    "A61M 16/01",
    "B25J 9/16",
    "G05B 13/02",
    "G16H 15/00",
    "G16H 20/40",
    "G16H 30/20",
    "G16H 30/40",
    "G16H 40/67",
    "G16H 50/20",
    "G16H 50/70",
    "G16H 70/20"
  ],
  "assignees": [
    "Sony Olympus Medical Solutions Inc"
  ],
  "inventors": [
    "Mitsuaki Shiraga"
  ],
  "filing_date": "2018-10-02",
  "publication_date": "2024-07-02",
  "grant_date": "2024-07-02",
  "priority_date": "2017-12-05",
  "application_number": "US-201816768081-A",
  "family_id": "66751426",
  "cited_by_count": 0,
  "citations": [
    "US20160278870A1",
    "US20030216836A1",
    "JP3104051U",
    "JP2006043209A",
    "WO2006077798A1",
    "US20100198402A1",
    "US20110046476A1",
    "US20090088634A1",
    "US20170345155A1",
    "US20120183191A1",
    "JP2011167301A",
    "US20140287393A1",
    "US20140343586A1",
    "US20140185888A1",
    "JP2016042982A",
    "JP2016154810A",
    "WO2016200887A1",
    "US20160381256A1",
    "JP2017047022A",
    "WO2017083768A1",
    "US20170143284A1",
    "US20170151027A1",
    "US20190046813A1",
    "US20180065248A1",
    "US20180137244A1",
    "US20180233222A1",
    "US9788907B1",
    "US20190133693A1"
  ]
}

Record 432 of 8,000 in Patents full text (MLC-0201). Request the full dataset.