WO2022180786A1 - 画像処理装置、画像処理方法及び記憶媒体 - Google Patents
画像処理装置、画像処理方法及び記憶媒体 Download PDFInfo
- Publication number
- WO2022180786A1 WO2022180786A1 PCT/JP2021/007346 JP2021007346W WO2022180786A1 WO 2022180786 A1 WO2022180786 A1 WO 2022180786A1 JP 2021007346 W JP2021007346 W JP 2021007346W WO 2022180786 A1 WO2022180786 A1 WO 2022180786A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- image
- mask
- image processing
- processing apparatus
- granularity
- Prior art date
Links
- 238000012545 processing Methods 0.000 title claims abstract description 111
- 238000003672 processing method Methods 0.000 title claims description 8
- 235000019580 granularity Nutrition 0.000 claims abstract description 78
- 238000003384 imaging method Methods 0.000 claims abstract description 16
- 238000005070 sampling Methods 0.000 claims description 4
- 238000012360 testing method Methods 0.000 claims description 2
- 238000001574 biopsy Methods 0.000 description 111
- 238000000034 method Methods 0.000 description 29
- 239000002245 particle Substances 0.000 description 24
- 238000001839 endoscopy Methods 0.000 description 17
- 230000015654 memory Effects 0.000 description 15
- 238000010586 diagram Methods 0.000 description 12
- 230000003902 lesion Effects 0.000 description 12
- 238000012986 modification Methods 0.000 description 10
- 230000004048 modification Effects 0.000 description 10
- 238000000605 extraction Methods 0.000 description 5
- 238000007689 inspection Methods 0.000 description 5
- 208000023514 Barrett esophagus Diseases 0.000 description 4
- 208000023665 Barrett oesophagus Diseases 0.000 description 4
- 210000001072 colon Anatomy 0.000 description 4
- 201000010099 disease Diseases 0.000 description 4
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 206010028980 Neoplasm Diseases 0.000 description 3
- 238000013528 artificial neural network Methods 0.000 description 3
- 238000004891 communication Methods 0.000 description 3
- 238000001514 detection method Methods 0.000 description 3
- 238000012549 training Methods 0.000 description 3
- 206010025323 Lymphomas Diseases 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 230000002183 duodenal effect Effects 0.000 description 2
- 210000003238 esophagus Anatomy 0.000 description 2
- 208000027866 inflammatory disease Diseases 0.000 description 2
- 210000002429 large intestine Anatomy 0.000 description 2
- 230000001613 neoplastic effect Effects 0.000 description 2
- 210000000056 organ Anatomy 0.000 description 2
- 210000000813 small intestine Anatomy 0.000 description 2
- 210000002784 stomach Anatomy 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 206010009944 Colon cancer Diseases 0.000 description 1
- 241001391944 Commicarpus scandens Species 0.000 description 1
- 208000011231 Crohn disease Diseases 0.000 description 1
- 206010061825 Duodenal neoplasm Diseases 0.000 description 1
- 208000000289 Esophageal Achalasia Diseases 0.000 description 1
- 208000000461 Esophageal Neoplasms Diseases 0.000 description 1
- 208000000624 Esophageal and Gastric Varices Diseases 0.000 description 1
- 206010061968 Gastric neoplasm Diseases 0.000 description 1
- 208000007882 Gastritis Diseases 0.000 description 1
- 201000003741 Gastrointestinal carcinoma Diseases 0.000 description 1
- 208000034991 Hiatal Hernia Diseases 0.000 description 1
- 206010020028 Hiatus hernia Diseases 0.000 description 1
- 206010030136 Oesophageal achalasia Diseases 0.000 description 1
- 206010030155 Oesophageal carcinoma Diseases 0.000 description 1
- 206010030216 Oesophagitis Diseases 0.000 description 1
- 208000009565 Pharyngeal Neoplasms Diseases 0.000 description 1
- 206010034811 Pharyngeal cancer Diseases 0.000 description 1
- 206010039580 Scar Diseases 0.000 description 1
- 206010054184 Small intestine carcinoma Diseases 0.000 description 1
- 208000005718 Stomach Neoplasms Diseases 0.000 description 1
- 208000007107 Stomach Ulcer Diseases 0.000 description 1
- 206010056091 Varices oesophageal Diseases 0.000 description 1
- 201000000621 achalasia Diseases 0.000 description 1
- 201000011510 cancer Diseases 0.000 description 1
- 239000002775 capsule Substances 0.000 description 1
- 206010009887 colitis Diseases 0.000 description 1
- 208000029742 colonic neoplasm Diseases 0.000 description 1
- 238000003745 diagnosis Methods 0.000 description 1
- 208000000718 duodenal ulcer Diseases 0.000 description 1
- 206010013864 duodenitis Diseases 0.000 description 1
- 210000001198 duodenum Anatomy 0.000 description 1
- 201000000312 duodenum cancer Diseases 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000002708 enhancing effect Effects 0.000 description 1
- 201000004101 esophageal cancer Diseases 0.000 description 1
- 208000024170 esophageal varices Diseases 0.000 description 1
- 201000010120 esophageal varix Diseases 0.000 description 1
- 208000006881 esophagitis Diseases 0.000 description 1
- 230000006870 function Effects 0.000 description 1
- 206010017758 gastric cancer Diseases 0.000 description 1
- 230000002496 gastric effect Effects 0.000 description 1
- 201000005917 gastric ulcer Diseases 0.000 description 1
- 238000011478 gradient descent method Methods 0.000 description 1
- 208000014617 hemorrhoid Diseases 0.000 description 1
- 201000002313 intestinal cancer Diseases 0.000 description 1
- 230000000968 intestinal effect Effects 0.000 description 1
- 201000008267 intestinal tuberculosis Diseases 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 210000002569 neuron Anatomy 0.000 description 1
- 208000003154 papilloma Diseases 0.000 description 1
- 208000014081 polyp of colon Diseases 0.000 description 1
- 208000015768 polyposis Diseases 0.000 description 1
- 238000013179 statistical model Methods 0.000 description 1
- 201000011549 stomach cancer Diseases 0.000 description 1
- 208000023984 stomach polyp Diseases 0.000 description 1
- 210000000115 thoracic cavity Anatomy 0.000 description 1
- 208000018408 tumor of duodenum Diseases 0.000 description 1
- 208000019553 vascular disease Diseases 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
- 230000003936 working memory Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/25—Determination of region of interest [ROI] or a volume of interest [VOI]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/0002—Inspection of images, e.g. flaw detection
- G06T7/0012—Biomedical image inspection
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B1/00—Instruments for performing medical examinations of the interior of cavities or tubes of the body by visual or photographical inspection, e.g. endoscopes; Illuminating arrangements therefor
- A61B1/00002—Operational features of endoscopes
- A61B1/00004—Operational features of endoscopes characterised by electronic signal processing
- A61B1/00009—Operational features of endoscopes characterised by electronic signal processing of image signals during a use of endoscope
- A61B1/000094—Operational features of endoscopes characterised by electronic signal processing of image signals during a use of endoscope extracting biological structures
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B1/00—Instruments for performing medical examinations of the interior of cavities or tubes of the body by visual or photographical inspection, e.g. endoscopes; Illuminating arrangements therefor
- A61B1/00002—Operational features of endoscopes
- A61B1/00004—Operational features of endoscopes characterised by electronic signal processing
- A61B1/00009—Operational features of endoscopes characterised by electronic signal processing of image signals during a use of endoscope
- A61B1/000096—Operational features of endoscopes characterised by electronic signal processing of image signals during a use of endoscope using artificial intelligence
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/11—Region-based segmentation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/50—Extraction of image or video features by performing operations within image blocks; by using histograms, e.g. histogram of oriented gradients [HoG]; by summing image-intensity values; Projection analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/52—Scale-space analysis, e.g. wavelet analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/74—Image or video pattern matching; Proximity measures in feature spaces
- G06V10/761—Proximity, similarity or dissimilarity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/762—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using clustering, e.g. of similar faces in social networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
- G06V10/771—Feature selection, e.g. selecting representative features from a multi-dimensional feature space
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/77—Processing image or video features in feature spaces; using data integration or data reduction, e.g. principal component analysis [PCA] or independent component analysis [ICA] or self-organising maps [SOM]; Blind source separation
- G06V10/80—Fusion, i.e. combining data from various sources at the sensor level, preprocessing level, feature extraction level or classification level
- G06V10/806—Fusion, i.e. combining data from various sources at the sensor level, preprocessing level, feature extraction level or classification level of extracted features
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/70—Arrangements for image or video recognition or understanding using pattern recognition or machine learning
- G06V10/82—Arrangements for image or video recognition or understanding using pattern recognition or machine learning using neural networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10068—Endoscopic image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V2201/00—Indexing scheme relating to image or video recognition or understanding
- G06V2201/03—Recognition of patterns in medical or anatomical images
Definitions
- the present disclosure relates to the technical field of image processing apparatuses, image processing methods, and storage media that process images acquired in endoscopy.
- Patent Literature 1 discloses a method of learning a learning model that outputs information about a lesion site included in captured image data when captured image data generated by an imaging device is input.
- Non-Patent Document 1 discloses Feature Pyramid Networks, which is a neural network architecture that performs inference using multi-scale feature maps.
- Non-Patent Document 2 discloses British guidelines for taking biopsies.
- an area of interest such as a lesion site
- detection is difficult depending on the target area of interest (for example, flat lesions, etc.).
- the position of the lesion site is specified in units of pixels, it may not be possible to detect the spot of interest that should be detected.
- an object of the present disclosure is to provide an image processing apparatus, an image processing method, and a storage medium that are capable of suitably identifying an attention point such as a lesion site from an image captured in an endoscopy. That is.
- Acquisition means for acquiring a photographed image obtained by photographing an object to be inspected by an imaging unit provided in the endoscope; a mask image generation means for generating a plurality of mask images showing candidate areas of attention points for the photographed image with different granularities; an identifying means for identifying the attention point based on the plurality of mask images; It is an image processing apparatus having
- the computer Acquiring a photographed image of an object to be inspected by an imaging unit provided in the endoscope, Generating a plurality of mask images showing candidate areas of attention points for the captured image with different granularities, Identifying the spot of interest based on the plurality of mask images; It is an image processing method.
- One aspect of the storage medium is Acquiring a photographed image of an object to be inspected by an imaging unit provided in the endoscope, Generating a plurality of mask images showing candidate regions of attention points for the captured image with different granularities,
- a storage medium storing a program for causing a computer to execute processing for identifying the attention point based on the plurality of mask images.
- FIG. 1 shows a schematic configuration of an endoscopy system
- 2 shows the hardware configuration of an image processing apparatus
- FIG. 10 is a diagram showing an outline of processing related to extraction and display of a biopsy region
- 1 is a functional block diagram of an image processing device
- FIG. (A) is a diagram showing an outline of processing for selecting a mask image as a biopsy site-specifying image.
- (B) is a diagram showing an outline of processing for generating a biopsy site-specifying image that integrates mask images.
- FIG. 11 is a diagram showing an outline of processing of a granularity adjusting unit
- 4 shows a display example of a display screen displayed by a display device in an endoscopy.
- FIG. 4 is an example of a flowchart showing an outline of processing executed by the image processing apparatus during endoscopy in the first embodiment
- FIG. 11 is a block diagram of an image processing apparatus according to a second embodiment
- FIG. It is an example of a flowchart executed by an image processing apparatus in the second embodiment.
- FIG. 1 shows a schematic configuration of an endoscopy system 100.
- the endoscopy system 100 provides biopsy site candidates (sites to be subjected to biological tissue sampling inspection) to an examiner such as a doctor who performs an examination or treatment using an endoscope.
- the system mainly includes an image processing device 1 , a display device 2 , and an endoscope 3 connected to the image processing device 1 .
- a biopsy site is an example of a "point of interest.”
- the image processing device 1 acquires images (also referred to as “captured images Ia”) captured by the endoscope 3 in time series from the endoscope 3 and displays a screen based on the captured images Ia on the display device 2 .
- the captured images Ia are images captured at predetermined time intervals during at least one of the process of inserting the endoscope 3 into the subject and the process of removing the endoscope 3 .
- the image processing apparatus 1 analyzes the captured image Ia to identify the biopsy site from the captured image Ia, and causes the display device 2 to display information about the identified biopsy site.
- the display device 2 is a display or the like that performs a predetermined display based on a display signal supplied from the image processing device 1 .
- the endoscope 3 mainly includes an operation unit 36 for an examiner to perform predetermined input, a flexible shaft 37 inserted into an organ to be imaged by an examinee, and an ultra-compact imaging device. It has a tip portion 38 containing an imaging unit such as an element, and a connection portion 39 for connecting to the image processing apparatus 1 .
- endoscopes targeted in the present disclosure include, for example, pharyngeal endoscopes, bronchoscopes, upper gastrointestinal endoscopes, duodenal endoscopes, small intestine endoscopes, colonoscopes, capsule endoscopes, thoracic scopes, laparoscopes, cystoscopes, choledoscopes, arthroscopes, spinal endoscopes, angioscopes, epidural endoscopes, and the like.
- the image processing device 1 mainly includes a processor 11 , a memory 12 , an interface 13 , an input section 14 , a light source section 15 and a sound output section 16 . Each of these elements is connected via a data bus 19 .
- the processor 11 executes a predetermined process by executing a program or the like stored in the memory 12.
- the processor 11 is a processor such as a CPU (Central Processing Unit), a GPU (Graphics Processing Unit), or a TPU (Tensor Processing Unit).
- Processor 11 may be composed of a plurality of processors.
- Processor 11 is an example of a computer.
- the memory 12 is composed of various volatile memories used as working memory, such as RAM (Random Access Memory) and ROM (Read Only Memory), and non-volatile memory for storing information necessary for processing of the image processing apparatus 1. be done.
- the memory 12 may include an external storage device such as a hard disk connected to or built into the image processing apparatus 1, or may include a storage medium such as a detachable flash memory.
- the memory 12 stores a program for the image processing apparatus 1 to execute each process in this embodiment.
- the memory 12 also stores mask image generation model information D1.
- the mask image generation model information D1 is information about a model (also called "mask image generation model") for generating a mask image, which is an image representing a biopsy site in the captured image Ia.
- the mask image generation model information D1 includes, for example, parameters necessary for constructing the mask image generation model.
- the mask image generation model is any machine learning model or statistical model. The model is trained to output multiple mask images shown. Note that when the mask image generation model is configured by a neural network, the mask image generation model information D1 includes, for example, the layer structure, the neuron structure of each layer, the number and size of filters in each layer, and the weight of each element of each filter. contains various parameters of In this embodiment, as an example, the mask image generation model is Feature Pyramid Networks. Details of the mask image generation model will be described later with reference to FIG.
- the mask image generation model information D1 may be stored in an external device capable of wired or wireless data communication with the image processing apparatus 1 instead of the memory 12 .
- This external device may be one or more server devices.
- the interface 13 performs an interface operation between the image processing device 1 and an external device.
- interface 13 provides display information “Ib” generated by processor 11 to display device 2 .
- the interface 13 also supplies the endoscope 3 with light or the like generated by the light source unit 15 .
- the interface 13 also supplies the processor 11 with an electrical signal indicating the captured image Ia supplied from the endoscope 3 .
- the interface 13 may be a communication interface such as a network adapter for performing wired or wireless communication with an external device, and may be a hardware interface conforming to USB (Universal Serial Bus), SATA (Serial AT Attachment), or the like. may
- the input unit 14 generates an input signal based on the operation by the inspector.
- the input unit 14 is, for example, a button, touch panel, remote controller, voice input device, or the like. As will be described later, the input unit 14 receives an input (external input) designating the granularity of the biopsy site to be displayed.
- the light source unit 15 generates light to be supplied to the distal end portion 38 of the endoscope 3 .
- the light source unit 15 may also incorporate a pump or the like for sending out water or air to be supplied to the endoscope 3 .
- the sound output unit 16 outputs sound under the control of the processor 11 .
- FIG. 1 A specific description will be given of processing related to identification and display of a biopsy site by the image processing device 1.
- FIG. 3 is a diagram showing an outline of processing related to extraction and display of a biopsy region.
- the image processing apparatus 1 generates a plurality of mask images "Im1" to "ImN" (N is 2 integers greater than or equal to ).
- the mask images Im1 to ImN are images showing the reliability of biopsy sites for each grid.
- the grid is a virtual area obtained by dividing the photographed image Ia into a lattice, and is a rectangular area corresponding to a plurality of pixels of the photographed image Ia.
- the mask images Im1 to ImN may be images showing a heat map (reliability map) of the reliability on the captured image Ia, and are binary images showing whether or not each grid is a biopsy site. There may be.
- FIG. 3 shows, as an example, mask images Im1 to ImN, which are binary images in which the pixels to be biopsied are white and the other pixels are black.
- the mask images Im1 to ImN have different granularities (resolutions) representing reliability of biopsy sites, and have grid numbers corresponding to the granularities.
- the mask image Im1 is an image having a grid with the number of vertical and horizontal grids of 4 ⁇ 4
- the mask image Im2 is an image having a grid with the number of vertical and horizontal grids of 8 ⁇ 8. It's becoming
- the image processing apparatus 1 selects or merges the generated N mask images Im1 to ImN with different granularities based on a predetermined algorithm to obtain an image for specifying a biopsy site ("biopsy site (also referred to as "specific image Ie").
- biopsy site also referred to as "specific image Ie”
- the image processing apparatus 1 selects the mask image Im1 as the biopsy site identification image Ie.
- the image processing apparatus 1 presents the specified biopsy site according to the granularity specified by the user through the input unit 14 or the like.
- a granularity of 6 ⁇ 6 is specified, and the image processing device 1 converts the biopsy site-specific image Ie (here, a 4 ⁇ 4 mask image) into a 6 ⁇ 6 mask image, Based on the mask image after conversion, the specified biopsy site is highlighted on the captured image Ia with the specified granularity.
- the image processing apparatus 1 can suitably present the candidate of the biopsy site to the user.
- the image processing apparatus 1 may display an image representing the specified biopsy site separately from the captured image Ia, instead of superimposing the specified biopsy site on the captured image Ia. This display example will be described with reference to FIG.
- the mask image generation model has a network architecture based on Feature Pyramid Networks.
- the mask image generation model performs a convolution operation on the captured image Ia, which is the input image, to generate a multi-scale feature map (that is, a tensor obtained by convolution), and biopsy each feature map.
- Location inference (in this embodiment, generation of mask images with different granularities) is performed.
- the number of mask images output by the mask image generation model and the granularity (resolution) of each mask image are preset in the learning stage.
- Such a mask image generation model is learned in advance based on the training data, and the learned parameters are stored in the mask image generation model information D1.
- the training data are a plurality of sets of captured images as input images and a plurality of mask images as correct data.
- the error (loss) between the output of the mask image generation model and the correct data when the input image is input is minimized by using the gradient descent method, the error backpropagation method, etc. parameters are determined.
- the mask image generation model is not limited to Feature Pyramid Networks, and may be any learning model that outputs mask images with different granularities for the input image.
- images are generated by resizing the captured image Ia, which is an input image, to a plurality of resolutions, and inference (in this embodiment, mask image generation) is performed on the generated images.
- inference in this embodiment, mask image generation
- FIG. 4 is an example of functional blocks of the image processing apparatus 1.
- the processor 11 of the image processing apparatus 1 functionally includes a photographed image acquisition unit 31, a mask image generation unit 32, an identification unit 33, a granularity adjustment unit 34, and a display control unit 35. and have In FIG. 4, the blocks that exchange data are connected by solid lines, but the combinations of blocks that exchange data are not limited to those shown in FIG. The same applies to other functional block diagrams to be described later.
- the captured image acquisition unit 31 acquires captured images Ia captured by the endoscope 3 via the interface 13 at predetermined intervals. Then, the captured image acquisition unit 31 supplies the acquired captured image Ia to the mask image generation unit 32, the granularity adjustment unit 34, and the display control unit 35, respectively.
- the mask image generation unit 32 acquires the captured image Ia supplied from the captured image acquisition unit 31, and inputs the captured image Ia to the mask image generation model configured based on the mask image generation model information D1 to generate the mask image. Im1 to ImN are generated.
- the identifying unit 33 identifies the biopsy site based on the mask images Im1 to ImN generated by the mask image generating unit 32.
- the specifying unit 33 generates a biopsy site specifying image Ie based on the mask images Im1 to ImN. Then, the identification unit 33 identifies the biopsy site based on the biopsy site identification image Ie.
- the specifying unit 33 may select one mask image from the mask images Im1 to ImN as the biopsy site specifying image Ie, and integrate the mask images Im1 to ImN or part thereof to specify the biopsy site. An image Ie may be generated. A specific example of the processing of the specifying unit 33 will be described later.
- the granularity adjustment unit 34 adjusts the biopsy site identification image Ie supplied from the identification unit 33 so as to have the granularity specified by the user (that is, the number of vertical and horizontal grids) (also referred to as “particle size adjustment image Iec”). call it).
- the granularity adjustment unit 34 acquires, for example, information designating the granularity from the input unit 14, and generates the granularity adjusted image Iec based on the granularity indicated by the acquired information.
- the granularity setting information specified in advance by the user or the default granularity setting information may be stored in advance in the memory 12 or the like.
- the grain size adjustment unit 34 generates the grain size adjusted image Iec by referring to the setting information described above from the memory 12 or the like.
- the generation of the granularity-adjusted image Iec by the granularity adjustment unit 34 will be supplementarily explained with reference to FIG. Note that the processing by the granularity adjustment unit 34 is processing performed when the granularity used for display is specified or set by the user, and is not an essential configuration.
- the display control unit 35 When the granularity used for display is specified or set by the user, the display control unit 35 generates the display information Ib based on the granularity-adjusted image Iec generated by the granularity adjustment unit 34, and displays the generated display information Ib on the display device. 2 to cause the display device 2 to display information about the biopsy region. On the other hand, if the granularity used for display is not specified or set by the user, the display control unit 35 generates display information Ib based on the biopsy site identification image Ie, and supplies the generated display information Ib to the display device 2. , the display device 2 is caused to display information about the biopsy region. A display example displayed on the display device 2 by the display control unit 35 will be described later with reference to FIG. 7 .
- each component of the captured image acquisition unit 31, the mask image generation unit 32, the identification unit 33, the granularity adjustment unit 34, and the display control unit 35 can be realized by the processor 11 executing a program, for example. Further, each component may be realized by recording necessary programs in an arbitrary nonvolatile storage medium and installing them as necessary. Note that at least part of each of these components may be realized by any combination of hardware, firmware, and software, without being limited to being implemented by program software. Also, at least part of each of these components may be implemented using a user-programmable integrated circuit, such as an FPGA (Field-Programmable Gate Array) or a microcontroller. In this case, this integrated circuit may be used to implement a program composed of the above components.
- FPGA Field-Programmable Gate Array
- each component may be configured by an ASSP (Application Specific Standard Produce), an ASIC (Application Specific Integrated Circuit), or a quantum processor (quantum computer control chip).
- ASSP Application Specific Standard Produce
- ASIC Application Specific Integrated Circuit
- quantum processor quantum computer control chip
- FIG. 5A is a diagram showing an overview of the process of selecting one mask image from the mask images Im1 to ImN as the biopsy site identification image Ie.
- FIG. 5B is a diagram showing an outline of processing for generating a biopsy site identification image Ie by integrating the mask images Im1 to ImN.
- the identification unit 33 identifies the maximum score for each mask image when the mask image represents the reliability score of the biopsy site for each grid, and The mask image with the highest score is selected as the biopsy site identification image Ie.
- the reliability is an index representing the high possibility that the biopsy site exists, and the higher the reliability, the higher the possibility that the biopsy site exists.
- the specifying unit 33 can select the mask image with the highest reliability as the biopsy site specifying image Ie.
- the specifying unit 33 calculates the degree of similarity for each pair of the mask images Im1 to ImN, and selects the biopsy site-identifying image Ie based on the degree of similarity. do.
- the identifying unit 33 may calculate the similarity using any image similarity index such as MSE (Mean Squared Error) or SSIM (Structural Similarity).
- the identifying unit 33 identifies the outlier mask image based on the degree of similarity, and excludes the identified mask image from targets to be selected as the biopsy site identifying image Ie.
- the specifying unit 33 excludes, for example, a mask image whose average value of similarity to other mask images is smaller than a predetermined value from targets to be selected as the biopsy site specifying image Ie. Then, the specifying unit 33 selects, for example, the mask image with the lowest granularity (that is, the lowest resolution) from the remaining mask image group as the biopsy site specifying image Ie.
- the identification unit 33 identifies the mask image that is an outlier based on the similarity of each pair of the mask images Im1 to ImN.
- the mask images Im1 to ImN may be clustered based on , and the biopsy site-specifying image Ie may be selected from the mask image group with the largest number of clusters.
- the specifying unit 33 may perform clustering using any clustering method such as the x-means method. Then, the specifying unit 33 selects the mask image with the lowest granularity (that is, the lowest resolution) from the mask image group of the cluster with the largest number as the biopsy site specifying image Ie.
- the specifying unit 33 resizes the mask images Im1 to ImN to have a common granularity, and averages (or adds) the resized mask images Im1 to ImN for each corresponding grid.
- a specific image Ie is generated.
- the specifying unit 33 converts the mask images Im1 to ImN into mask images having a granularity (number of grids) of 8 ⁇ 8, and then integrates the converted mask images into an 8 ⁇ 8 grid.
- a biopsy site-specific image Ie with a granularity of .
- the specifying unit 33 may set weights “w1” to “wN” for each of the mask images Im1 to ImN, and perform weighted averaging using these weights to generate the biopsy site specifying image Ie. good.
- the weights w1 to wN in this case may be predetermined values, or may be set to values corresponding to the maximum reliability score for each grid of each mask image.
- the specifying unit 33 may select the mask image to be integrated from the mask images Im1 to ImN based on the degree of similarity between the mask images Im1 to ImN.
- the identification unit 33 identifies mask images that are outliers based on the degree of similarity, and integrates a group of mask images excluding the identified mask images to generate the biopsy site identification image Ie.
- the identifying unit 33 identifies a mask image group that has the largest number of clusters by the above-described clustering method, and generates a biopsy site identification image Ie by integrating the identified mask image group.
- the specifying unit 33 can preferably exclude mask images that include errors due to disturbances or the like, and generate the biopsy site specifying image Ie that accurately represents the biopsy site candidate region.
- FIG. 6 is a diagram showing an overview of the processing of the granularity adjustment unit 34.
- the particle size adjustment unit 34 converts the biopsy site identification image Ie having a particle size of 4 ⁇ 4 into a particle size adjustment image Iec of 6 ⁇ 6, which is the particle size specified by the user.
- the granularity adjustment unit 34 performs the conversion processing shown in FIG. 6 using an arbitrary image resizing method.
- the particle size adjustment unit 34 defines a two-dimensional coordinate system having a common value range with one of the four corners as the origin for each mask image, and each grid in the biopsy site identification image Ie and the particle size adjustment image Iec.
- a coordinate value eg, center coordinate value
- the particle size adjustment unit 34 sets the grid value (score) of the particle size adjusted image Iec to the grid value of the biopsy site identification image Ie that is closest to the coordinate values corresponding to the grid.
- the particle size adjustment unit 34 may determine the value of each grid of the particle size adjusted image Iec from the values of the grids of the plurality of biopsy site specific images Ie by performing a known weighted average process based on the coordinate values. .
- FIG. 7 shows a display example of a display screen displayed by the display device 2 during an endoscopy.
- the display control unit 35 of the image processing device 1 transmits display information Ib generated based on the captured image Ia acquired by the captured image acquisition unit 31 and the granularity adjusted image Iec generated by the granularity adjustment unit 34 to the display device 2. , the display screen shown in FIG.
- the display control unit 35 of the image processing device 1 displays the latest captured image 70, the biopsy site map 71, and the particle size display area 72 on the display screen.
- the display control unit 35 displays a moving image based on the latest captured image Ia acquired by the captured image acquiring unit 31 as the latest captured image 70 .
- the display control unit 35 also displays a biopsy site map 71 based on the particle size adjustment image Iec acquired from the particle size adjustment unit 34 .
- the display control unit 35 may display the particle size-adjusted image Iec as it is as the biopsy site map 71 , or may display an image obtained by subjecting the particle size-adjusted image Iec to predetermined processing as the biopsy site map 71 . You may In the latter example, for example, the display control unit 35 may perform a process of converting the granularity-adjusted image Iec into a binary image or an image with a predetermined gradation of 3 or more.
- the display control unit 35 may display a display corresponding to the biopsy site map 71 so as to be superimposed on the latest captured image 70 .
- the display control unit 35 may display a heat map based on the particle size-adjusted image Iec on the biopsy site map 71, and the range in which the value in the particle size-adjusted image Iec is equal to or greater than a predetermined threshold is May be highlighted.
- the display control unit 35 may display an image based on the granularity-adjusted image Iec.
- the display control unit 35 displays the current granularity setting (here, 16 ⁇ 16) in the granularity display area 72, and also displays a change button 73 for instructing to change the granularity setting. Then, when the display control unit 35 detects that the change button 73 has been selected, the input unit 14 receives an input specifying the granularity to be set. In this case, the display control unit 35 may receive designation of granularity to be set based on the voice input signal acquired by the input unit 14 . In this case, the display control unit 35 may determine whether or not to change the granularity setting and the granularity after the change based on the voice input signal without providing the change button 73 .
- the display control unit 35 may determine whether or not to change the granularity setting and the granularity after the change based on the voice input signal without providing the change button 73 .
- the image processing apparatus 1 can suitably present regions that are candidates for biopsy sites to the examiner, and support efficient and effective biopsy. can.
- the image processing apparatus 1 can preferably present biopsy site candidates to the examiner according to the granularity specified by the user.
- FIG. 8 is an example of a flow chart showing an overview of the process executed by the image processing apparatus 1 during an endoscopy in the first embodiment.
- the captured image acquisition unit 31 of the image processing device 1 acquires the captured image Ia (step S11).
- the captured image acquisition unit 31 of the image processing apparatus 1 receives the captured image Ia from the endoscope 3 via the interface 13 .
- the mask image generation unit 32 of the image processing apparatus 1 generates N mask images with different granularities from the photographed image Ia acquired in step S11, using a mask image generation model configured based on the mask image generation model information D1. Im1 to ImN are generated (step S12). Based on the N mask images Im1 to ImN, the specifying unit 33 generates a biopsy site specifying image Ie, which is a mask image representing the biopsy site region (step S13).
- the granularity adjustment unit 34 generates a granularity-adjusted image Iec based on the granularity specified by the user (step S14).
- the display control unit 35 causes the display device 2 to display the captured image Ia acquired in step S11 and the information about the biopsy site (step S15).
- the display control unit 35 causes the display device 2 to display the particle size-adjusted image Iec or an image based on the particle size-adjusted image Iec as the information about the biopsy site.
- the display control unit 35 can suitably present the presence and position of the specified biopsy site with the granularity specified by the user.
- the user may view the biopsy site identification image Ie output to the display device 2 and input to change the particle size on the spot.
- the display control unit 35 displays the biopsy site identification image Ie on the display device 2 after step S13, and the particle size adjustment unit 34 detects an input to change the particle size by the input unit 14, in step S14, A granularity-adjusted image Iec is generated based on the granularity specified by the user.
- the display control unit 35 causes the display device 2 to display the captured image Ia acquired in step S11 and information about the biopsy site.
- the image processing apparatus 1 determines whether or not the endoscopy has ended (step S16). For example, the image processing apparatus 1 determines that the endoscopy has ended when a predetermined input or the like to the input unit 14 or the operation unit 36 is detected. When the image processing apparatus 1 determines that the endoscopy has ended (step S16; Yes), the processing of the flowchart ends. On the other hand, when the image processing apparatus 1 determines that the endoscopy has not ended (step S16; No), the process returns to step S11. Then, the image processing apparatus 1 executes the processes of steps S11 to S16 on the photographed image Ia newly generated by the endoscope 3. FIG.
- the image processing apparatus 1 divides the photographed image Ia into grids of a plurality of patterns, and performs biopsy based on the mask images Im1 to ImN, which are the detection results of the biopsy site for each grid pattern. Identify the location.
- the biopsy site can be identified with appropriate accuracy even for a site such as a flat lesion that is difficult for even a doctor to determine, and can be suitably presented to the examiner as a candidate for the biopsy site. Therefore, in this case, an efficient and effective biopsy can be assisted.
- the image processing apparatus 1 can preferably present the biopsy region to the examiner according to the granularity specified by the user, thereby enhancing the user's convenience.
- the image processing apparatus 1 may process, after the examination, a video composed of the photographed image Ia generated during the endoscopy.
- the image processing apparatus 1 when an image to be processed is specified based on a user's input from the input unit 14 or the like at an arbitrary timing after the examination, the image processing apparatus 1 performs time-series photographed images constituting the image.
- the processing of the flowchart of FIG. 8 is sequentially performed for Ia. If the image processing apparatus 1 determines in step S16 that the target video has ended, it ends the processing of the flowchart.
- the process of the flowchart is performed on the photographed image Ia of .
- the image processing apparatus 1 may output information about the presence or absence of the biopsy site.
- the display control unit 35 functions as an output control unit that controls display and sound output, controls the display of the display device 2 based on the above-described embodiment, and controls the biopsy site identification image Ie or the granularity Based on the adjusted image Iec, the presence/absence of the biopsy site is determined, and the sound output unit 16 is controlled.
- the display control unit 35 for example, in the biopsy site identification image Ie or the particle size adjustment image Iec, if there is a grid value representing the reliability of the presence of the biopsy site is greater than or equal to a predetermined threshold, It determines that there is a biopsy site, and instructs the sound output unit 16 to output a predetermined sound or voice.
- the above-described threshold value is stored in the memory 12 or the like in advance, for example.
- the image processing apparatus 1 may display information regarding the presence or absence of the biopsy site instead of notifying the information regarding the presence or absence of the biopsy site by sound. In this case, for example, when the image processing apparatus 1 determines that a biopsy site exists, the image processing apparatus 1 notifies the user of the presence of the biopsy site by highlighting the latest captured image 70 with a border effect. good.
- the image processing apparatus 1 can suitably notify the examiner of the existence or non-existence of a biopsy site, and suitably suppress overlooking of the biopsy site. Then, according to this modified example, the image processing apparatus 1 may output at least one of the information regarding the presence or absence of the biopsy site and the information regarding the position in the captured image Ia where the target site exists.
- the target to be detected by the mask image generation model is not limited to the biopsy site, and may be any target site that the inspector needs to pay attention to.
- Such points of interest include a lesion site, an inflamed site, a surgical scar or other cut site, a fold or projection site, and the tip portion 38 of the endoscope 3 in the lumen. It may be a place where it is easy to contact (easy to break) on the wall surface of the wall.
- the disease condition of the lesion site to be detected is exemplified as (a) to (f) below.
- the mask image generation model is learned so that when the photographed image Ia is input, it outputs a mask image showing predetermined points of interest with different granularities.
- the mask image generation unit 32 of the image processing apparatus 1 generates mask images Im1 to ImN related to the spot of interest from the captured image Ia using such a mask image generation model, and the specifying unit 33 generates the mask images Im1 to ImN A mask image is generated by selecting from or integrating these.
- the granularity adjustment unit 34 changes the granularity of the mask image to the granularity specified by the user, and the display control unit 35 displays the target location based on the mask image generated by the granularity adjustment unit 34.
- the image processing apparatus 1 can suitably present the point of interest to the inspector. Also, in this modification, an effect equivalent to the effect in extracting a biopsy site described in the first embodiment before this modification is achieved.
- FIG. 9 is a block diagram of an image processing device 1X according to the second embodiment.
- the image processing device 1X includes acquisition means 31X, mask image generation means 32X, and identification means 33X.
- the image processing device 1X may be composed of a plurality of devices.
- the acquisition means 31X acquires a photographed image of the inspection object photographed by the photographing unit provided in the endoscope.
- the acquisition unit 31X can be the captured image acquisition unit 31 in the first embodiment (including modifications; the same applies hereinafter).
- the acquisition unit 31X may immediately acquire the captured image generated by the imaging unit, or may acquire, at a predetermined timing, the captured image generated by the imaging unit and stored in the storage device in advance.
- the mask image generation means 32X generates a plurality of mask images that indicate candidate areas of attention points for the photographed image with different granularities.
- the mask image generator 32X can be, for example, the mask image generator 32 in the first embodiment.
- the specifying means 33X specifies a spot of interest based on a plurality of mask images.
- the specifying unit 33X can be, for example, the specifying unit 33 in the first embodiment.
- FIG. 10 is an example of a flowchart showing a processing procedure in the second embodiment.
- the acquisition unit 31X acquires a photographed image of an inspection object photographed by the photographing unit provided in the endoscope (step S21).
- the mask image generating means 32X generates a plurality of mask images indicating candidate areas of attention points for the photographed image with different granularities (step S22).
- the identifying means 33X identifies the spot of interest based on the plurality of mask images (step S23).
- the image processing apparatus 1X can preferably identify a point of interest in a photographed image of an inspection target based on a plurality of mask images showing candidate areas of the point of interest with different granularities. can.
- the plurality of mask images are images representing the reliability of the spot of interest for each grid corresponding to the plurality of pixels of the photographed image, and the number of the grids determined according to the granularity is the plurality of mask images. 1.
- the image processing apparatus according to appendix 1 or 2, wherein the mask image generating means generates a plurality of feature maps having different scales based on the captured image, and generates the plurality of mask images from the plurality of feature maps.
- Appendix 4 4. The image processing apparatus according to any one of appendices 1 to 3, wherein the identifying means identifies the attention point based on one mask image selected from the plurality of mask images.
- Appendix 5 4.
- the image processing apparatus according to any one of appendices 1 to 3, wherein the identifying means identifies the attention point based on a mask image obtained by integrating the plurality of mask images. [Appendix 6] 6.
- the image processing apparatus according to appendix 4 or 5, wherein the specifying unit calculates a degree of similarity between the plurality of mask images, and determines a mask image to be selected or integrated based on the degree of similarity.
- Appendix 7 7.
- the image processing apparatus according to any one of appendices 1 to 6, further comprising output control means for outputting information regarding the attention point specified by the specifying means.
- Appendix 8 8.
- Appendix 9 further comprising granularity adjusting means for converting the image specifying the attention point into a granularity adjusted image representing the attention point with a specified granularity; 9.
- the image processing device according to appendix 7 or 8, wherein the output control means causes a display device to display the granularity-adjusted image or an image based on the granularity-adjusted image.
- Appendix 10 10. The image processing apparatus according to any one of appendices 1 to 9, wherein the attention point is a point to be subjected to a biological tissue sampling test.
- a storage medium storing a program for causing a computer to execute processing for identifying the attention point based on the plurality of mask images.
- Reference Signs List 1 1X image processing device 2 display device 3 endoscope 11 processor 12 memory 13 interface 14 input unit 15 light source unit 16 sound output unit 100 endoscopy system
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Medical Informatics (AREA)
- General Health & Medical Sciences (AREA)
- Evolutionary Computation (AREA)
- Artificial Intelligence (AREA)
- Databases & Information Systems (AREA)
- Computing Systems (AREA)
- Software Systems (AREA)
- Life Sciences & Earth Sciences (AREA)
- Surgery (AREA)
- Radiology & Medical Imaging (AREA)
- Nuclear Medicine, Radiotherapy & Molecular Imaging (AREA)
- Optics & Photonics (AREA)
- Biophysics (AREA)
- Pathology (AREA)
- Signal Processing (AREA)
- Biomedical Technology (AREA)
- Heart & Thoracic Surgery (AREA)
- Molecular Biology (AREA)
- Animal Behavior & Ethology (AREA)
- Public Health (AREA)
- Veterinary Medicine (AREA)
- Quality & Reliability (AREA)
- Endoscopes (AREA)
Abstract
Description
内視鏡に設けられた撮影部により検査対象を撮影した撮影画像を取得する取得手段と、
前記撮影画像に対する注目箇所の候補領域を異なる粒度により示した複数のマスク画像を生成するマスク画像生成手段と、
前記複数のマスク画像に基づき、前記注目箇所を特定する特定手段と、
を有する画像処理装置である。
コンピュータが、
内視鏡に設けられた撮影部により検査対象を撮影した撮影画像を取得し、
前記撮影画像に対する注目箇所の候補領域を異なる粒度により示した複数のマスク画像を生成し、
前記複数のマスク画像に基づき、前記注目箇所を特定する、
画像処理方法である。
内視鏡に設けられた撮影部により検査対象を撮影した撮影画像を取得し、
前記撮影画像に対する注目箇所の候補領域を異なる粒度により示した複数のマスク画像を生成し、
前記複数のマスク画像に基づき、前記注目箇所を特定する処理をコンピュータに実行させるプログラムを格納した記憶媒体である。
図1は、内視鏡検査システム100の概略構成を示す。図1に示すように、内視鏡検査システム100は、内視鏡を利用した検査又は治療を行う医師等の検査者に対して生検箇所(生体組織採取検査の対象となる箇所)の候補を提示するシステムであって、主に、画像処理装置1と、表示装置2と、画像処理装置1に接続された内視鏡スコープ3と、を備える。生検箇所は「注目箇所」の一例である。
画像処理装置1は、内視鏡検査時に生成された撮影画像Iaから構成された映像を、検査後において処理してもよい。
画像処理装置1は、生検箇所の撮影画像Ia中の位置に関する情報を出力する制御に代えて、又は、これに加えて、生検箇所の存否に関する情報を通知する出力を行ってもよい。
マスク画像生成モデルにより検知する対象は、生検箇所に限らず、検査者が注目する必要がある任意の注目箇所であってもよい。このような注目箇所は、病変部位、炎症が生じている箇所、手術痕その他の切り傷が生じている箇所、ひだや突起が生じている箇所、内視鏡スコープ3の先端部38が管腔内の壁面において接触しやすい(閊えやすい)箇所などであってもよい。例えば、病変部位の場合、検出対象となる病変部位の病状は、以下の(a)~(f)ように例示される。
(b)食道:食道ガン、食道炎、食道裂孔ヘルニア、バレット食道、食道静脈瘤、食道アカラシア、食道粘膜下腫瘍、食道良性腫瘍
(c)胃:胃ガン、胃炎、胃潰瘍、胃ポリープ、胃腫瘍
(d)十二指腸:十二指腸ガン、十二指腸潰瘍、十二指腸炎、十二指腸腫瘍、十二指腸リンパ腫
(e)小腸:小腸ガン、小腸腫瘍性疾患、小腸炎症性疾患、小腸血管性疾患
(f)大腸:大腸ガン、大腸腫瘍性疾患、大腸炎症性疾患、大腸ポリープ、大腸ポリポーシス、クローン病、大腸炎、腸結核、痔
図9は、第2実施形態における画像処理装置1Xのブロック図である。画像処理装置1Xは、取得手段31Xと、マスク画像生成手段32Xと、特定手段33Xとを備える。画像処理装置1Xは、複数の装置から構成されてもよい。
内視鏡に設けられた撮影部により検査対象を撮影した撮影画像を取得する取得手段と、
前記撮影画像に対する注目箇所の候補領域を異なる粒度により表した複数のマスク画像を生成するマスク画像生成手段と、
前記複数のマスク画像に基づき、前記注目箇所を特定する特定手段と、
を有する画像処理装置。
[付記2]
前記複数のマスク画像は、前記撮影画像の複数の画素に対応するグリッドごとに前記注目箇所となる信頼度を表した画像であり、前記粒度に応じて定まる前記グリッドの数は前記複数のマスク画像間において異なる、付記1に記載の画像処理装置。
[付記3]
前記マスク画像生成手段は、前記撮影画像に基づきスケールが異なる複数の特徴マップを生成し、当該複数の特徴マップから前記複数のマスク画像を夫々生成する、付記1または2に記載の画像処理装置。
[付記4]
前記特定手段は、前記複数のマスク画像から選択された1のマスク画像に基づき、前記注目箇所を特定する、、付記1~3のいずれか一項に記載の画像処理装置。
[付記5]
前記特定手段は、前記複数のマスク画像が統合されたマスク画像に基づき、前記注目箇所を特定する、付記1~3のいずれか一項に記載の画像処理装置。
[付記6]
前記特定手段は、前記複数のマスク画像間の類似度を算出し、前記類似度に基づき、選択又は統合する対象となるマスク画像を決定する、付記4または5に記載の画像処理装置。
[付記7]
前記特定手段が特定した前記注目箇所に関する情報を出力する出力制御手段をさらに有する、付記1~6のいずれか一項に記載の画像処理装置。
[付記8]
前記出力制御手段は、前記注目箇所の存否に関する情報又は前記注目箇所が存在する前記撮影画像中の位置に関する情報の少なくとも一方を、前記注目箇所に関する情報として出力する、付記7に記載の画像処理装置。
[付記9]
前記注目箇所を特定した画像を、指定された粒度により前記注目箇所を表した画像である粒度調整画像に変換する粒度調整手段をさらに有し、
前記出力制御手段は、前記粒度調整画像又は前記粒度調整画像に基づく画像を表示装置に表示させる、付記7または8に記載の画像処理装置。
[付記10]
前記注目箇所は、生体組織採取検査の対象となる箇所である、付記1~9のいずれか一項に記載の画像処理装置。
[付記11]
コンピュータが、
内視鏡に設けられた撮影部により検査対象を撮影した撮影画像を取得し、
前記撮影画像に対する注目箇所の候補領域を異なる粒度により表した複数のマスク画像を生成し、
前記複数のマスク画像に基づき、前記注目箇所を特定する、
画像処理方法。
[付記12]
内視鏡に設けられた撮影部により検査対象を撮影した撮影画像を取得し、
前記撮影画像に対する注目箇所の候補領域を異なる粒度により表した複数のマスク画像を生成し、
前記複数のマスク画像に基づき、前記注目箇所を特定する処理をコンピュータに実行させるプログラムを格納した記憶媒体。
2 表示装置
3 内視鏡スコープ
11 プロセッサ
12 メモリ
13 インターフェース
14 入力部
15 光源部
16 音出力部
100 内視鏡検査システム
Claims (12)
- 内視鏡に設けられた撮影部により検査対象を撮影した撮影画像を取得する取得手段と、
前記撮影画像に対する注目箇所の候補領域を異なる粒度により表した複数のマスク画像を生成するマスク画像生成手段と、
前記複数のマスク画像に基づき、前記注目箇所を特定する特定手段と、
を有する画像処理装置。 - 前記複数のマスク画像は、前記撮影画像の複数の画素に対応するグリッドごとに前記注目箇所となる信頼度を表した画像であり、前記粒度に応じて定まる前記グリッドの数は前記複数のマスク画像間において異なる、請求項1に記載の画像処理装置。
- 前記マスク画像生成手段は、前記撮影画像に基づきスケールが異なる複数の特徴マップを生成し、当該複数の特徴マップから前記複数のマスク画像を夫々生成する、請求項1または2に記載の画像処理装置。
- 前記特定手段は、前記複数のマスク画像から選択された1のマスク画像に基づき、前記注目箇所を特定する、請求項1~3のいずれか一項に記載の画像処理装置。
- 前記特定手段は、前記複数のマスク画像が統合されたマスク画像に基づき、前記注目箇所を特定する、請求項1~3のいずれか一項に記載の画像処理装置。
- 前記特定手段は、前記複数のマスク画像間の類似度を算出し、前記類似度に基づき、選択又は統合する対象となるマスク画像を決定する、請求項4または5に記載の画像処理装置。
- 前記特定手段が特定した前記注目箇所に関する情報を出力する出力制御手段をさらに有する、請求項1~6のいずれか一項に記載の画像処理装置。
- 前記出力制御手段は、前記注目箇所の存否に関する情報又は前記注目箇所が存在する前記撮影画像中の位置に関する情報の少なくとも一方を、前記注目箇所に関する情報として出力する、請求項7に記載の画像処理装置。
- 前記注目箇所を特定した画像を、指定された粒度により前記注目箇所を表した画像である粒度調整画像に変換する粒度調整手段をさらに有し、
前記出力制御手段は、前記粒度調整画像又は前記粒度調整画像に基づく画像を表示装置に表示させる、請求項7または8に記載の画像処理装置。 - 前記注目箇所は、生体組織採取検査の対象となる箇所である、請求項1~9のいずれか一項に記載の画像処理装置。
- コンピュータが、
内視鏡に設けられた撮影部により検査対象を撮影した撮影画像を取得し、
前記撮影画像に対する注目箇所の候補領域を異なる粒度により表した複数のマスク画像を生成し、
前記複数のマスク画像に基づき、前記注目箇所を特定する、
画像処理方法。 - 内視鏡に設けられた撮影部により検査対象を撮影した撮影画像を取得し、
前記撮影画像に対する注目箇所の候補領域を異なる粒度により表した複数のマスク画像を生成し、
前記複数のマスク画像に基づき、前記注目箇所を特定する処理をコンピュータに実行させるプログラムを格納した記憶媒体。
Priority Applications (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
EP21859349.9A EP4298986A4 (en) | 2021-02-26 | 2021-02-26 | IMAGE PROCESSING DEVICE AND METHOD, AND STORAGE MEDIUM |
US18/278,294 US20240127434A1 (en) | 2021-02-26 | 2021-02-26 | Image processing device, image processing method and storage medium |
JP2023501952A JP7485193B2 (ja) | 2021-02-26 | 2021-02-26 | 画像処理装置、画像処理方法及びプログラム |
PCT/JP2021/007346 WO2022180786A1 (ja) | 2021-02-26 | 2021-02-26 | 画像処理装置、画像処理方法及び記憶媒体 |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/JP2021/007346 WO2022180786A1 (ja) | 2021-02-26 | 2021-02-26 | 画像処理装置、画像処理方法及び記憶媒体 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2022180786A1 true WO2022180786A1 (ja) | 2022-09-01 |
Family
ID=83049000
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2021/007346 WO2022180786A1 (ja) | 2021-02-26 | 2021-02-26 | 画像処理装置、画像処理方法及び記憶媒体 |
Country Status (4)
Country | Link |
---|---|
US (1) | US20240127434A1 (ja) |
EP (1) | EP4298986A4 (ja) |
JP (1) | JP7485193B2 (ja) |
WO (1) | WO2022180786A1 (ja) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2024084838A1 (ja) * | 2022-10-18 | 2024-04-25 | 日本電気株式会社 | 画像処理装置、画像処理方法及び記憶媒体 |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2020003992A1 (ja) * | 2018-06-28 | 2020-01-02 | 富士フイルム株式会社 | 学習装置及び学習方法、並びに、医療画像処理装置 |
WO2020003607A1 (ja) | 2018-06-25 | 2020-01-02 | オリンパス株式会社 | 情報処理装置、モデル学習方法、データ認識方法および学習済みモデル |
WO2020012530A1 (ja) * | 2018-07-09 | 2020-01-16 | 日本電気株式会社 | 施術支援装置、施術支援方法、及びコンピュータ読み取り可能な記録媒体 |
US20200065961A1 (en) * | 2018-08-21 | 2020-02-27 | International Business Machines Corporation | Training classifier with data of different granularity |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP6584090B2 (ja) | 2015-02-23 | 2019-10-02 | Hoya株式会社 | 画像処理装置 |
-
2021
- 2021-02-26 JP JP2023501952A patent/JP7485193B2/ja active Active
- 2021-02-26 EP EP21859349.9A patent/EP4298986A4/en active Pending
- 2021-02-26 US US18/278,294 patent/US20240127434A1/en active Pending
- 2021-02-26 WO PCT/JP2021/007346 patent/WO2022180786A1/ja active Application Filing
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2020003607A1 (ja) | 2018-06-25 | 2020-01-02 | オリンパス株式会社 | 情報処理装置、モデル学習方法、データ認識方法および学習済みモデル |
WO2020003992A1 (ja) * | 2018-06-28 | 2020-01-02 | 富士フイルム株式会社 | 学習装置及び学習方法、並びに、医療画像処理装置 |
WO2020012530A1 (ja) * | 2018-07-09 | 2020-01-16 | 日本電気株式会社 | 施術支援装置、施術支援方法、及びコンピュータ読み取り可能な記録媒体 |
US20200065961A1 (en) * | 2018-08-21 | 2020-02-27 | International Business Machines Corporation | Training classifier with data of different granularity |
Non-Patent Citations (3)
Title |
---|
REBECCA C FITZGERALDMASSIMILIANO DI PIETROKRISH RAGUNATH ET AL., BRITISH SOCIETY OF GASTROENTEROLOGY GUIDELINES ON THE DIAGNOSIS AND MANAGEMENT OF BARRETT'S OESOPHAGUS, 28 October 2013 (2013-10-28), pages 17, Retrieved from the Internet <URL:https://www.bsg.org.uk/wp-content/uploads/2019/12/BSG-guidelines-on-the-diagnosis-and-management-of-Barretts-oesophagus.pdf> |
See also references of EP4298986A4 |
T.-Y LINP. DOLLARR. GIRSHICKK. HEB. HARIHARANS. BELONGIE: "Feature pyramid networks for object detection", CVPR, 2017 |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2024084838A1 (ja) * | 2022-10-18 | 2024-04-25 | 日本電気株式会社 | 画像処理装置、画像処理方法及び記憶媒体 |
Also Published As
Publication number | Publication date |
---|---|
EP4298986A4 (en) | 2024-03-13 |
EP4298986A1 (en) | 2024-01-03 |
JPWO2022180786A1 (ja) | 2022-09-01 |
US20240127434A1 (en) | 2024-04-18 |
JP7485193B2 (ja) | 2024-05-16 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP7346285B2 (ja) | 医療画像処理装置、内視鏡システム、医療画像処理装置の作動方法及びプログラム | |
CN110049709B (zh) | 图像处理装置 | |
US8502861B2 (en) | Image display apparatus | |
JP5374078B2 (ja) | 画像処理装置、画像処理方法および画像処理プログラム | |
JP5800468B2 (ja) | 画像処理装置、画像処理方法、および画像処理プログラム | |
JP5147308B2 (ja) | 画像抽出装置および画像抽出プログラム | |
CN111275041B (zh) | 内窥镜图像展示方法、装置、计算机设备及存储介质 | |
WO2022180786A1 (ja) | 画像処理装置、画像処理方法及び記憶媒体 | |
JP5004736B2 (ja) | 画像処理装置および画像処理プログラム | |
WO2023042273A1 (ja) | 画像処理装置、画像処理方法及び記憶媒体 | |
WO2016208016A1 (ja) | 画像処理装置、画像処理方法、及び画像処理プログラム | |
JP2006141734A (ja) | 画像処理装置、画像処理方法、及び画像処理プログラム | |
JP2007075157A (ja) | 画像表示装置 | |
CN114049934B (zh) | 辅助诊断方法、装置、***、设备和介质 | |
WO2022224446A1 (ja) | 画像処理装置、画像処理方法及び記憶媒体 | |
WO2023126999A1 (ja) | 画像処理装置、画像処理方法、及び、記憶媒体 | |
WO2023187886A1 (ja) | 画像処理装置、画像処理方法及び記憶媒体 | |
WO2024084838A1 (ja) | 画像処理装置、画像処理方法及び記憶媒体 | |
WO2023233453A1 (ja) | 画像処理装置、画像処理方法及び記憶媒体 | |
WO2024018581A1 (ja) | 画像処理装置、画像処理方法及び記憶媒体 | |
WO2023181353A1 (ja) | 画像処理装置、画像処理方法及び記憶媒体 | |
WO2022185369A1 (ja) | 画像処理装置、画像処理方法及び記憶媒体 | |
WO2023162216A1 (ja) | 画像処理装置、画像処理方法及び記憶媒体 | |
WO2022208615A1 (ja) | 画像処理装置、画像処理方法及び記憶媒体 | |
WO2024142490A1 (ja) | 画像処理装置、画像処理方法及び記憶媒体 |
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: 21859349 Country of ref document: EP Kind code of ref document: A1 |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2023501952 Country of ref document: JP |
|
WWE | Wipo information: entry into national phase |
Ref document number: 18278294 Country of ref document: US |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2021859349 Country of ref document: EP |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
ENP | Entry into the national phase |
Ref document number: 2021859349 Country of ref document: EP Effective date: 20230926 |