Medical record guided computer-aided detection

Abstract

A framework for medical record guided computer aided detection. One or more properties of a subject of interest previously reported in one or more prior electronic medical records are extracted and used to cut a region of interest of a segmented body region in newly acquired medical image data of a patient. An object of interest may then be detected within the clipped region of interest.

Description

Technical Field

[0001] The present disclosure relates generally to data processing, and more particularly to a framework for medical record guided computer-aided detection. Background Art

[0002] In clinical settings, it is common for medical providers to have access to the patient's prior electronic medical records. Electronic medical records (EMRs) often contain information about previously discovered nodules, pathological findings, with or without detailed descriptions of the corresponding properties (e.g., size, location in the lungs, density). During lung cancer screening or follow-up, radiologists need to access previously detected nodules and examine their development to decide on medical strategies.

[0003] As CT images have been used by radiologists to identify nodules, computer-aided diagnosis (CAD) algorithms have emerged to facilitate the review process. For example, a lung CAD system can act as a secondary reader or a parallel primary reader to assist radiologists in finding nodules in CT images. However, in both working modes, the CAD system needs to scan the entire CT image volume and complete the following steps: data preprocessing, lung segmentation, region proposal, post-classification, and post-filtering. Most of these steps may involve deep learning neural networks.

[0004] The current CAD-assisted workflow requires the user to run the lung CAD system again on newly acquired CT images from the same patient. On average, it takes the lung CAD system about 3 to 4 minutes to process a standard chest CT series. During this time, the user must wait for the process to complete, which causes unnecessary delays in the reading workflow. In the previous EMR, the previously discovered nodules have been described by text, which typically includes information such as location in the CT image, nodule diameter, density, and other characteristics of the nodules. Ignoring such information provided in the previous EMR is disadvantageous.

[0005] Another disadvantage of ignoring the prior EMR is that the findings provided by CAD may not be consistent between the newly acquired CT images and the previously acquired CT images. During the new CT scanning process, changes in CT acquisition parameters and body orientation may result in inconsistent output findings compared to the previous scan. If a nodule reported in the prior EMR is gone, the radiologist needs to manually reference the reported location for additional review. This is not only time-consuming, but also negatively affects the radiologist's confidence in the CAD system. Summary of the invention

[0006] This paper describes a framework for computer-aided detection. One or more properties of an object of interest previously reported in one or more prior electronic medical records are extracted and used to crop a region of interest (ROI) of a segmented body region in newly acquired medical image data of a patient. The object of interest can then be detected within the cropped region of interest. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] A more complete appreciation of the present disclosure and its many attendant aspects will be readily obtained by reference to the following detailed description when considered in conjunction with the accompanying drawings as the present disclosure and its many attendant aspects become better understood.

[0008] Figure 1 A block diagram illustrating an exemplary system is shown;

[0009] Figure 2 An exemplary method of computer-aided detection is shown; and

[0010] Figure 3 An exemplary method for computer-aided detection of lung nodules is shown. DETAILED DESCRIPTION

[0011] In the following description, numerous specific details are set forth, such as examples of specific components, devices, methods, etc., in order to provide a thorough understanding of the implementation of the present framework. However, it will be clear to those skilled in the art that these specific details need not be adopted to practice the implementation of the present framework. In other instances, in order to avoid unnecessarily obscuring the implementation of the present framework, well-known materials or methods are not described in detail. Although the present framework is susceptible to various modifications and alternative forms, specific embodiments thereof are shown by way of example in the accompanying drawings and will be described in detail herein. However, it should be understood that there is no intention to limit the invention to the particular form disclosed, but on the contrary, it is intended to cover all modifications, equivalents, and alternatives that fall within the spirit and scope of the present invention. In addition, for ease of understanding, certain method steps are described as separate steps; however, these separately described steps should not be interpreted as necessarily order-dependent in their execution. Independent of the use of grammatical terms, individuals having male and female identities are also included in the terminology. "

[0012] Unless otherwise stated, it will be clear from the following discussion that it will be appreciated that terms such as "segmenting," "generating," "registering," "determining," "aligning," "locating," "processing," "computing," "selecting," "estimating," "detecting," "tracking," or the like may refer to the actions and processes of a computer system or similar electronic computing device that manipulates and transforms data represented as physical (e.g., electronic) quantities within registers and memories of a computer system into other data similarly represented as physical quantities within computer system memories or registers or other such information storage, transmission, or display devices. Embodiments of the methods described herein may be implemented using computer software. If written in a programming language that conforms to a recognized standard, the sequence of instructions designed to implement the methods may be compiled so as to be executed on a variety of hardware platforms and for interfacing with a variety of operating systems. In addition, the implementation of the present framework is not described with reference to any particular programming language. It will be appreciated that a variety of programming languages ​​may be used.

[0013] For the sake of brevity, an image corresponding to an object (e.g., a tissue, an organ, a tumor, etc. of a subject (e.g., a patient, etc.) or a portion thereof (e.g., a region of interest (ROI) in an image) may be referred to as an image of the object or a portion thereof (e.g., ROI) or an image of the object or a portion thereof (e.g., ROI) or the object itself. For example, an ROI of an image corresponding to the lungs or the heart may be described as the ROI including the lungs or the heart. As another example, a chest or an image including the chest may be referred to as a chest image or simply as the chest. For the sake of brevity, a portion of an image corresponding to an object being processed (e.g., extracted, segmented) may be described as the object being processed. For example, extracting a portion of the image corresponding to the lungs from the rest of the image may be described as the lungs being extracted.

[0014] In a typical clinical workflow, if the radiologist is not assisted by a computer-aided detection (CAD) algorithm, the radiologist either reviews the newly acquired CT images from scratch or waits for the CAD system to complete the detection process from scratch on the newly acquired CT images. Running CAD directly on the entire CT image series is time consuming. Additionally, there is a chance that the CAD system may miss findings previously reported in the prior EMR.

[0015] EMR has been used to improve CAD algorithms for pulmonary nodule detection. For example, nodule information described in EMR can be used to provide additional training labels or weakly supervised labels for nodule detection. These are evidences showing that EMR does contain information indicating the location of nodules and the nature of nodules in CT images. Text extracted from medical reports can be useful for pulmonary nodule detection.

[0016] Registration between the prior CT image and the current CT image can be performed to align the two CT images. However, this requires knowledge of the world coordinates of the findings in the prior CT image. However, in the EMR, radiologists do not report the location of the findings in world coordinates. An alternative is to run CAD on the prior CT image. However, this is also very time consuming and may miss findings previously reported by the reader in the prior EMR.

[0017] A framework for medical record-guided computer-aided detection is described herein. According to one aspect, the framework facilitates a disease (e.g., lung cancer) follow-up reading workflow by leveraging information provided by a prior EMR. Radiologists can benefit from improved efficiency by being able to quickly locate nodules or other objects of interest described in a prior EMR without having to wait for the software to scan the entire newly acquired CT image volume. CAD algorithms can avoid missing findings previously reported in an EMR, even if the newly acquired CT image is significantly different from the CT image acquired during the previous scan, for example, in terms of acquisition parameters.

[0018] Therefore, the present framework increases efficiency by running a local search for objects of interest rather than a global search. By running the object detection algorithm on a locally cropped ROI, a lot of time is saved. At the same time, instead of performing a global general search for all types of nodules, specific findings as described in the report are detected, which is less likely to miss reported findings. The present framework can be designed to search for findings that match the objects of interest described in the EMR, and the cropped ROI limits the search range to improve efficiency. These and other exemplary advantages and features will be described in more detail in the following description.

[0019] Figure 1 1 is a block diagram illustrating an exemplary system 100. System 100 includes a computer system 101 for implementing a framework as described herein. In some implementations, computer system 101 operates as a standalone device. In other implementations, computer system 101 can be connected (e.g., using a network) to other machines, such as medical imaging device 102 and workstation 103. In a networked deployment, computer system 101 can operate in the capacity of a server (e.g., as a client user machine in a server-client user network environment, or as a peer machine in a peer-to-peer (or distributed) network environment).

[0020] In one implementation, the computer system 101 includes a processor device or central processing unit (CPU) 104, which is coupled to one or more non-transitory computer-readable media 105 (e.g., computer storage or memory devices), a display device 108 (e.g., a monitor), and various input devices 110 (e.g., a mouse, touchpad, or keyboard) via an input-output interface 121. The computer system 101 may further include support circuits such as caches, power supplies, clock circuits, and communication buses. Various other peripheral devices such as additional data storage devices and printing devices may also be connected to the computer system 101.

[0021] The present technology can be implemented in various forms of hardware, software, firmware, special processors or combinations thereof, either as part of microinstruction code or as part of an application or software product or a combination thereof, which are executed via an operating system. In some implementations, the technology described herein is implemented as a computer-readable program code tangibly embodied in one or more non-transitory computer-readable media 105. In particular, the present technology can be implemented by a processing module 107. The non-transitory computer-readable medium 105 may include a random access memory (RAM), a read-only memory (ROM), a magnetic floppy disk, a flash memory and other types of memory or a combination thereof. The computer-readable program code is executed by a processor device 104 to process data provided by, for example, a medical imaging device 102 and a database 106. As such, the computer system 101 is a general-purpose computer system that becomes a special-purpose computer system when the computer-readable program code is executed. The computer-readable program code is not intended to be limited to any particular programming language and its implementation. It will be appreciated that a variety of programming languages ​​and their encodings can be used to implement the teachings of the disclosure contained herein. The same or different computer-readable medium 105 may be used to store a database 106 including, but not limited to, image datasets, knowledge bases, individual subject data, electronic medical records (EMRs), diagnostic reports (or files) of subjects, or combinations thereof.

[0022] The medical imaging device 102 acquires medical image data 132. Such medical image data 132 may be processed by the processing module 107. The medical imaging device 102 may be a radiological scanner and / or appropriate peripheral devices (e.g., a keyboard and a display device) for acquiring, collecting, and / or storing such medical image data 132. The medical imaging device 102 may acquire medical image data 132 from a subject or patient using techniques such as high-resolution computed tomography (HRCT), magnetic resonance (MR) imaging, computed tomography (CT), spiral CT, X-ray, angiography, positron emission tomography (PET), fluoroscopy, ultrasound, single photon emission computed tomography (SPECT), or combinations thereof. Other types of imaging modalities may also be useful.

[0023] Medical image data 132 is data representing a medical image (or more than one medical image in some examples). For example, medical image data 132 may include an array or list of pixel or voxel values. When processed by suitable image viewing software, medical image data 132 results in the presentation of the medical image (or multiple medical images) it represents. An image file containing medical image data 132 may further include one or more attributes and attribute values. One or more attribute values ​​are separate and distinct from the medical image data 132 and instead include a text string indicating the content of the medical image data 132. In some examples, such attribute values ​​may be referred to as metadata for an image file. In some examples, the portion of an image file storing attributes and attribute values ​​may be referred to as a header of an image file, and the attributes and attribute values ​​may be referred to as header data for an image file.

[0024] A specific example of an image file is a Digital Imaging and Communications in Medicine (DICOM) file. A DICOM file stores medical image data as pixel data in a specified data element, and further stores one or more attributes as one or more other data elements, each of the one or more attributes having an attribute value that includes a text string indicating the content of the medical image data. An example of such a DICOM attribute is "Study Description", whose attribute value is a text string describing the study of which the medical imaging data is part (e.g., 'NERU0^HEAD', where the medical imaging data has a patient's head region) and thereby indicates the content of the medical imaging data. There are also other examples of such DICOM attributes, such as "Series Description", "Examined Body Part", and other examples.

[0025] The workstation 103 may include a computer and appropriate peripherals such as a keyboard and a display device, and may operate in conjunction with the overall system 100. For example, the workstation 103 may communicate with the medical imaging device 102 so that the medical image data 132 may be presented or displayed at the workstation 103. The workstation 103 may communicate directly with the computer system 101 to display processed data and / or output results 144. The workstation 103 may include a graphical user interface to receive user input via an input device (e.g., a keyboard, a mouse, a touch screen, a voice or video recognition interface, etc.) to manipulate the visualization and / or processing of data.

[0026] It will be further understood that because some of the constituent system components and method steps depicted in the accompanying drawings can be implemented in software, the actual connections between the system components (or process steps) may vary depending on the manner in which the present framework is programmed. Given the teachings provided herein, one of ordinary skill in the relevant art will be able to envision these and similar implementations or configurations of the present framework.

[0027] Figure 2 An exemplary method 200 for computer-aided detection is shown. It should be understood that the steps of method 200 may be performed in the order shown or in a different order. Additional, different, or fewer steps may also be provided. In addition, method 200 may be used Figure 1 The system 100 may be implemented by a different system or a combination thereof.

[0028] At 202, the processing module 107 receives one or more prior electronic medical records (EMRs) of the patient and newly acquired medical image data. The one or more prior EMRs may be retrieved from, for example, the database 106. The one or more prior EMRs contain information or findings entered during a prior visit of the patient. The findings may describe physical and psychological events of the patient investigated by a healthcare provider during a physical examination, for example, performed by the healthcare provider's senses and / or a medical device (e.g., a diagnostic or medical imaging device). The findings may describe one or more objects of interest, such as lung nodules, lesions, or other abnormalities. The findings may indicate, for example, the properties (e.g., location, size, type, density) of such objects of interest. The newly acquired medical image data may be acquired by the medical imaging device 102 during the patient's current visit (i.e., after one or more visits that generated the prior EMRs). For example, the newly acquired medical image data may be a series of CT images acquired at various angles of the patient's body. Other types of medical image data may also be acquired.

[0029] At 204, the processing module 107 segments the body region in the newly acquired medical image data to generate a segmented body region. In some implementations, the segmented body region includes different segments. For example, different lobes in the lungs are segmented and identified based on the newly acquired medical image data. In some implementations, the body region is segmented using attributes with attribute values, and the attribute values ​​are stored in an image file (e.g., a DICOM file) containing the newly acquired medical image data. For example, see U.S. patent application No. 17 / 806,159, filed on June 9, 2022, "Determining a Body Region represented by Medical Imaging Data," which is incorporated herein by reference. One or more text strings associated with the attribute values ​​can be input into a machine learning model that is trained to output a body region based on the input of one or more such text strings. Determining a body region based on a text string of an image file may be less resource intensive and therefore more efficient than, for example, determining a body region by extracting and analyzing the medical image data itself.

[0030] Alternatively or in combination thereof, the body region is segmented by sampling voxels from newly acquired medical image data around a single point of interest such that at least one voxel is skipped between two sampled voxels and applying a trained classifier to the sampled voxels to identify the type of organ at the single point of interest. For example, see U.S. patent application Ser. No. 18 / 312,822, filed May 5, 2023, attorney docket No. 2022P05356US, "Method for identifying a type of organ in a volumetric medical image," which is incorporated herein by reference. A single point of interest is a point in the medical image data for which it is desired to identify the type of organ corresponding to the point. A user can select a single point of interest.

[0031] At 206, the processing module 107 extracts one or more properties of at least one object of interest previously reported in a prior EMR. In some implementations, the processing module 107 applies one or more natural language processing (NLP) techniques to identify and extract properties (e.g., size, location, density) associated with an object of interest (e.g., a lung nodule) previously reported in a prior EMR. The properties extracted from the prior EMR may include, but are not limited to, the location, density, size, or diameter of the object of interest. Other properties may also be extracted. The location of the object of interest may be described with reference to segmented body segments rather than using coordinates. For example, the EMR may include the following description of the location of a lung nodule in plain language: "lateral segment of the middle lobe of the right lung."

[0032] NLP techniques for extracting the properties of an object of interest include, for example, artificial intelligence or machine learning techniques (e.g., deep neural networks). In some implementations, properties are extracted from a patient data model (PDM), which is generated from a prior EMR. PDM is an abstract model that organizes the elements of patient data and standardizes how they are related to each other and how they are related to the properties of real-world entities. PDM is the central foundation on which various clinical workflows store and exchange data. All data should be described according to a common PDM (cPDM) to exchange across different systems. PDM makes data searchable. For example, "a patient's pulmonary nodule location" can be extracted by querying using a medical system nomenclature (SNOMED) code (e.g., 786838002 for pulmonary nodules and 246267002 for location) as a search keyword.

[0033] Artificial intelligence (AI) technology can be used to generate PDM. In some implementations, the AI ​​engine decomposes each description in the EMR into what (WHAT) (i.e., abnormal or disease / pathological findings) and where (WHERE) (i.e., abnormal anatomical location). The AI ​​engine may include a neural parser and a named entity recognition module to perform such decomposition. The results can be formatted according to the Fast Healthcare Interoperability Resources (FHIR) standard. Other formats are also useful. The AI ​​engine can assign SNOMED codes to search keywords, allowing PDM to be searchable using SNOMED codes. In addition to SNOMED code searches, semantic searches for similar phrases and sentences using AI can also be enabled without the need for string matching. Alternatively, questions written in natural language for the system can also be asked to search for PDM via questions and answers.

[0034] At 208, the processing module 107 uses the one or more extracted properties of the object of interest to crop a region of interest of the segmented body region. For example, a region of interest (ROI) containing the object of interest is cropped at the extracted position within the segmented body region. The region of interest may contain at least one complete segment of the segmented body region. The region of interest may be defined by a rectangle, a square, or any other geometric shape.

[0035] At 210, the processing module 107 detects at least one object of interest within the cropped region of interest. An object detection algorithm may be used to quickly process small regions of interest. Exemplary object detection algorithms include, but are not limited to, neural-based or machine learning techniques and non-neural-based techniques. The detected objects of interest that match the description in the prior EMR may be shown to the user via, for example, the workstation 103. For example, a non-cancerous solid nodule of 5 mm in size in the left lower lung is described in the EMR. The processing module 107 may detect the nodule within the cropped region of interest and display it to the user.

[0036] The processing module 107 can highlight the detected object of interest in the cropped region of interest. In some implementations, the processing module 107 can highlight the detected object of interest by displaying a text label describing the nodule (e.g., "Type: Solid, Size: 5mm, Malignant: No"). For example, when a user selects a nodule on a medical image via a user interface, the text label can be displayed next to the nodule. In some implementations, the nodule and / or its label is highlighted with relevant text in a prior EMR. For example, the nodule and / or its label is displayed in the same color as the relevant text, while irrelevant text is displayed in a different color. Other types of highlighting can also be applied, such as using the same font or shading.

[0037] Figure 3An exemplary method 300 for computer-aided detection of pulmonary nodules according to one implementation is shown. It should be understood that the steps of method 300 may be performed in the order shown or in a different order. Additional, different, or fewer steps may also be provided. Furthermore, method 300 may be used Figure 1 The system 100 may be implemented by a different system or a combination thereof.

[0038] The processing module 107 receives a newly acquired chest CT image series 302 and one or more prior electronic medical records 306 of a patient. The processing module 107 segments the lung region in the newly acquired medical image data 302 into different lung lobe segments 304. The segmentation can be performed using, for example, a single point of interest selected by a user. Other types of segmentation algorithms are also useful.

[0039] The processing module 107 further extracts properties of the pulmonary nodules previously reported in the prior EMR 306. In some implementations, the processing module 107 applies natural language processing (NLP) techniques to identify and extract medical findings, sizes, and locations associated with the pulmonary nodules previously reported in at least one section 308 of the prior EMR 306.

[0040] At 310, the processing module 107 uses the extracted lung nodule locations to crop a region of interest 312 of the segmented lung region 304. The region of interest can be defined by a rectangle containing at least one complete lung lobe region. The processing module 107 can then detect lung nodules within the cropped region of interest 312.

[0041] At 314 , the detected nodules are highlighted and displayed to the user.

[0042] The following is a list of non-limiting exemplary embodiments disclosed herein:

[0043] Illustrative Embodiment 1. A data processing system comprising: a non-transitory memory device for storing a computer-readable program code; and a processor device in communication with the non-transitory memory device, the processor device operating with the computer-readable program code to perform steps comprising: receiving one or more prior electronic medical records of a patient and newly acquired medical image data, segmenting a body region in the newly acquired medical image data to generate a segmented body region, extracting one or more properties of an object of interest previously reported in the one or more prior electronic medical records, cropping a region of interest of the segmented body region using the one or more properties to generate a cropped region of interest, detecting an object of interest within the cropped region of interest, and highlighting the object of interest in the cropped region of interest.

[0044] Illustrative embodiment 2. The data processing system of illustrative embodiment 1, wherein the newly acquired medical image data includes medical image data acquired by high-resolution computed tomography (HRCT), magnetic resonance (MR) imaging, computed tomography (CT), spiral CT, X-ray, angiography, positron emission tomography (PET), fluoroscopy, ultrasound, single photon emission computed tomography (SPECT), or a combination thereof.

[0045] Illustrative Embodiment 3. The data processing system of Illustrative Embodiments 1-2, wherein segmenting the body region includes segmenting different lobes in a lung.

[0046] Illustrative Embodiment 4. The data processing system of Illustrative Embodiments 1-3, wherein segmenting the body region comprises determining the segmented body region using attributes stored in an image file containing the newly acquired medical image data.

[0047] Illustrative Embodiment 5. The data processing system of Illustrative Embodiments 1-4, wherein the one or more properties include a location of the object of interest.

[0048] Illustrative Embodiment 6. The data processing system of Illustrative Embodiments 1-5, wherein the one or more properties include a size, a diameter, or a density of the object of interest.

[0049] Illustrative Embodiment 7. The data processing system of Illustrative Embodiments 1-6, wherein extracting one or more properties of the object of interest previously reported in one or more prior electronic medical records comprises applying one or more natural language processing techniques.

[0050] Illustrative Embodiment 8. The data processing system of Illustrative Embodiments 1-7, wherein highlighting the object of interest in the cropped region of interest includes displaying a text label describing the object of interest.

[0051] Illustrative Embodiment 9. A method includes: receiving one or more prior electronic medical records and newly acquired medical image data of a patient; segmenting a body region in the newly acquired medical image data to generate a segmented body region; extracting one or more properties of an object of interest previously reported in the one or more prior electronic medical records; cropping a region of interest of the segmented body region using the one or more properties to generate a cropped region of interest; and detecting an object of interest within the cropped region of interest.

[0052] Illustrative Embodiment 10. The method of Illustrative Embodiment 9, wherein segmenting the body region comprises segmenting different lobes in a lung.

[0053] Illustrative Example 11. The method of Illustrative Examples 9-10, wherein the object of interest comprises a lung nodule.

[0054] Illustrative Embodiment 12. The method of Illustrative Embodiments 9-11, wherein the one or more properties include a location of the object of interest.

[0055] Illustrative Embodiment 13. The method of Illustrative Embodiments 9-12, wherein the one or more properties include a size, a diameter, or a density of the object of interest.

[0056] Illustrative Embodiment 14. The method of Illustrative Embodiments 9-13, wherein extracting one or more properties of the object of interest previously reported in one or more prior electronic medical records comprises applying one or more natural language processing techniques.

[0057] Illustrative Embodiment 15. The method of Illustrative Embodiments 9-14, further comprising highlighting the object of interest in the cropped region of interest.

[0058] Illustrative Embodiment 16. The method of Illustrative Embodiment 15, wherein highlighting the object of interest in the cropped region of interest comprises displaying a text label describing the object of interest.

[0059] Illustrative Embodiment 17. The method of Illustrative Embodiment 16, wherein highlighting the object of interest in the cropped region of interest further comprises highlighting a text label and related text in a prior electronic medical record using the same color.

[0060] Illustrative embodiment 18. One or more non-transitory computer-readable media embodying a program of instructions executable by a machine to perform steps comprising: segmenting a body region in newly acquired medical image data to generate a segmented body region; extracting one or more properties of an object of interest previously reported in one or more prior electronic medical records; cropping a region of interest of the segmented body region using the one or more properties to generate a cropped region of interest; and detecting an object of interest within the cropped region of interest.

[0061] Illustrative Embodiment 19. The one or more non-transitory computer readable media of Illustrative Embodiment 18, wherein the steps further include highlighting the object of interest in the cropped region of interest.

[0062] Illustrative embodiment 20. The one or more non-transitory computer-readable media of illustrative embodiment 19, wherein highlighting the object of interest in the cropped region of interest includes displaying a text label describing the object of interest.

[0063] Although the present framework has been described in detail with reference to exemplary embodiments, it will be appreciated by those skilled in the art that various modifications and substitutions may be made thereto without departing from the spirit and scope of the present invention as set forth in the appended claims. For example, within the scope of the present disclosure and the appended claims, elements and / or features of different exemplary embodiments may be combined with each other and / or substituted with each other.

Claims

1. A data processing system, comprising: a non-transitory memory device for storing computer readable program code; as well as a processor device in communication with the non-transitory memory device, the processor device operating with the computer readable program code to perform steps comprising: receiving one or more prior electronic medical records of the patient and the newly acquired medical image data, segmenting a body region in the newly acquired medical image data to generate a segmented body region, extracting one or more properties of the subject of interest previously reported in the one or more prior electronic medical records, cropping a region of interest of the segmented body region using the one or more properties to generate a cropped region of interest, detecting the object of interest within the cropped region of interest, and The object of interest in the cropped region of interest is highlighted.

2. The data processing system according to claim 1, wherein: The newly acquired medical image data includes medical image data acquired by high-resolution computed tomography (HRCT), magnetic resonance (MR) imaging, computed tomography (CT), spiral CT, X-ray, angiography, positron emission tomography (PET), fluoroscopy, ultrasound, single photon emission computed tomography (SPECT) or a combination thereof.

3. The data processing system according to claim 1, wherein: Segmenting the body region includes segmenting different lobes in the lungs.

4. The data processing system according to claim 1, wherein: Segmenting the body region includes determining the segmented body region using attributes stored in an image file containing the newly acquired medical image data.

5. The data processing system according to claim 1, wherein: The one or more properties include a location of the object of interest.

6. The data processing system according to claim 1, wherein: The one or more properties include a size, a diameter or a density of the object of interest.

7. The data processing system according to claim 1, wherein: Extracting the one or more properties of the object of interest previously reported in the one or more prior electronic medical records includes applying one or more natural language processing techniques.

8. The data processing system according to claim 1, wherein: Highlighting the object of interest in the cropped region of interest includes displaying a text label describing the object of interest.

9. A method comprising: receiving one or more prior electronic medical records of the patient and newly acquired medical image data; Segmenting a body region in the newly acquired medical image data to generate a segmented body region; extracting one or more properties of the subject of interest previously reported in the one or more prior electronic medical records; cropping a region of interest of the segmented body region using the one or more properties to generate a cropped region of interest; and The object of interest is detected within the cropped region of interest.

10. The method according to claim 9, wherein: Segmenting the body region includes segmenting different lobes in the lungs.

11. The method according to claim 9, wherein: The object of interest includes a lung nodule.

12. The method according to claim 9, wherein: The one or more properties include a location of the object of interest.

13. The method according to claim 9, wherein: The one or more properties include a size, a diameter or a density of the object of interest.

14. The method according to claim 9, wherein: Extracting one or more properties of the object of interest previously reported in the one or more prior electronic medical records includes applying one or more natural language processing techniques.

15. The method of claim 9, further comprising highlighting the object of interest in the cropped region of interest.

16. The method according to claim 15, wherein: Highlighting the object of interest in the cropped region of interest includes displaying a text label describing the object of interest.

17. The method according to claim 16, wherein: Highlighting the object of interest in the cropped region of interest further includes highlighting the text label and related text in the previous electronic medical record using the same color.

18. One or more non-transitory computer-readable media embodying a program of instructions executable by a machine to perform steps comprising: segmenting a body region in the newly acquired medical image data to generate a segmented body region; extracting one or more properties of the subject of interest previously reported in one or more prior electronic medical records; cropping a region of interest of the segmented body region using the one or more properties to generate a cropped region of interest; and The object of interest is detected within the cropped region of interest.

19. The one or more non-transitory computer-readable media of claim 18, wherein: The steps further include highlighting the object of interest in the cropped region of interest.

20. The one or more non-transitory computer-readable media of claim 19, wherein: Highlighting the object of interest in the cropped region of interest includes displaying a text label describing the object of interest.

Images