DETECTION AND CLASSIFICATION OF OTOSCOPIC IMAGES
Systems and methods for the automatic detection of the tympanic membrane and classification of the middle ear. Embodiments can comprise one or more computing devices and otoscope assembly. Computing devices can be configured to store and/or execute various engines of system including user interface, data store, detection model, segmentation model, and diagnosis model. Embodiments methods are configured for use by a non-healthcare professional, such as a parent or caregiver to automatically detect and classify images depicting the tympanic membrane and/or middle ear space.
1 . A method of identifying a location for tympanostomy tube placement at a tympanic membrane of a patient, the method comprising the steps of:
receiving, at one or more programmable computing devices, ear image data;
determining, by one or more programmable computing devices, whether the ear image data includes at least a portion of the tympanic membrane;
when the ear image data is determined to include at least a portion of the tympanic membrane, determining, by one or more programmable computing devices based on the ear image data, a region at the tympanic membrane for placing a tympanostomy tube; and
providing, by one or more programmable computing devices, an output indicative of the region at the tympanic membrane for placing the tympanostomy tube.
2 . The method of claim 1 , wherein the determined region at the tympanic membrane for placing the tympanostomy tube corresponds to a quadrant, within the ear image data, determined to include the at least the portion of the tympanic membrane.
3 . The method of claim 2 , wherein the quadrant determined to include the at least the portion of the tympanic membrane comprises an anterior or interior quadrant within the ear image data determined to include the at least the portion of the tympanic membrane.
4 . The method of claim 2 , wherein the provided output indicative of the region at the tympanic membrane for placing the tympanostomy tube comprises a visual annotation included at the quadrant determined to include the at least the portion of the tympanic membrane.
5 . The method of claim 4 , wherein the visual annotation is included in association with, and at a relative location within, the ear image data.
6 . The method of claim 4 , wherein the one or more programmable computing devices comprise an otoscope assembly and a remote programmable computing device, and wherein the visual annotation is included at the quadrant at the tympanic membrane for visual perception at the tympanic membrane.
7 . The method of claim 6 ,
wherein the otoscope assembly comprises: an otoscope, an otoscope speculum configured to couple to the otoscope, a camera configured to capture image data when the otoscope speculum is inserted into an ear canal, a light source configured to illuminate an ear canal when the otoscope speculum is inserted into the ear canal, an incision tool that is configured to create an incision at the tympanic membrane, and a tube placement tool that is configured to hold a tympanostomy tube and to place the tympanostomy tube at the incision at the tympanic membrane, and
wherein the visual annotation, included at the quadrant at the tympanic membrane for visual perception at the tympanic membrane, is provided by the light source of the otoscope assembly.
8 . The method of claim 7 , wherein the tube placement tool comprises a mechanical arm that is movable relative to otoscope.
9 . The method of claim 1 ,
wherein determining, by one or more programmable computing devices, whether the ear image data includes at least a portion of the tympanic membrane comprises determining, by the one or more programmable computing devices, a first predetermined probability range that the ear image data includes at least a portion of the tympanic membrane and a second, lower predetermined probability range that the ear image data includes at least a portion of the tympanic membrane,
when the one or more programmable computing devices determine that the ear image data includes at least a portion of the tympanic membrane within the first predetermined probability range, the one or more programmable computing devices determine that the ear image data includes at least a portion of the tympanic membrane, and
when the one or more programmable computing devices determine that the ear image data includes at least a portion of the tympanic membrane within the second, lower predetermined probability range, the one or more programmable computing devices determine that the ear image data does not include at least a portion of the tympanic membrane.
10 . The method of claim 9 , when the one or more programmable computing devices determine that the ear image data includes at least a portion of the tympanic membrane within the second, lower probability range, outputting, by the one or more programmable computing devices an insufficient ear image data capture notification indicative of the determination by the one or more programmable computing devices that the ear image data does not include at least a portion of the tympanic membrane.
11 . The method of claim 9 , wherein, when the one or more programmable computing devices determine that the ear image data includes at least a portion of the tympanic membrane by determining that the ear image data includes at least a portion of the tympanic membrane within the first predetermined probability range, the one or more programmable computing devices then determine the region at the tympanic membrane for placing a tympanostomy tube.
12 . The method of claim 11 ,
wherein determining, by one or more programmable computing devices, the region at the tympanic membrane for placing a tympanostomy tube comprises determining, by the one or more programmable computing device, a first predetermined probability range that the output indicative of the region at the tympanic membrane for placing the tympanostomy tube is accurate and a second, lower predetermined probability range that the output indicative of the region at the tympanic membrane for placing the tympanostomy tube is accurate,
when the one or more programmable computing devices determine that the output indicative of the region at the tympanic membrane for placing the tympanostomy tube is within the first predetermined probability range, the one or more programmable computing devices determine that the output indicative of the region at the tympanic membrane for placing the tympanostomy tube is accurate, and
when the one or more programmable computing devices determine that the output indicative of the region at the tympanic membrane for placing the tympanostomy tube is within the second, lower predetermined probability range, the one or more programmable computing devices determine that the output indicative of the region at the tympanic membrane for placing the tympanostomy tube is not accurate.
13 . The method of claim 1 , wherein determining, by one or more programmable computing devices, whether the ear image data includes at least a portion of the tympanic membrane comprises using a machine learning classifier trained to determine whether pixels of image data depict a portion of the tympanic membrane.
14 . The method of claim 13 , wherein the machine learning classifier has been trained with prior myringotomy-related surgical findings.
15 . A method for capturing and classifying otoscopic images, the method comprising the steps of:
capturing an image with an otoscope assembly, the otoscope assembly including
an otoscope,
an otoscope speculum,
a camera, and
a light; and
transmitting the image to a computing device, wherein the computing device
detects the presence or absence of the tympanic membrane,
classifies a condition of the middle ear space, tympanic membrane, and/or ear canal, and
presents the classification to a user.
16 . The method of claim 15 , wherein the computing device determines a probability that it has received an image of a tympanic membrane.
17 . The method of claim 15 , wherein a detected tympanic membrane results in the computing device drawing a box around the tympanic membrane in the image, and wherein the computing device determines a probability that it has placed the box around the tympanic membrane in the image.
18 . The method of claim 17 , wherein the computing device determines the presence of one or more conditions of the middle ear space, and wherein the one or more conditions of the middle ear space are selected from the group consisting of: normal/healthy, aerated middle ear, otitis media with effusion, and acute otitis media in the middle ear space.
19 . The method of claim 15 , wherein the computing device determines the presence or absence of the tympanic membrane at the ear canal, and wherein the computing device determines: (i) a presence of wax in the ear canal; (ii) an extent of an occlusion formed by the wax determined to be present in the ear canal; and (iii) a position, relative to the ear canal, of the wax determined to be present in the ear canal.
20 . A system comprising:
an otoscope assembly having otoscope, otoscope speculum, camera, and light;
a computing device in communication with a user interface, the computing device configured to provide
a data store comprising databases, file systems, memories, or other storage systems configured to store and provide data items,
a detection model comprising a ResNet classifier configured and trained to determine whether or not an input image data from a captured image depicts a tympanic membrane,
a segmentation model configured to receive each captured image and identify pixel regions within the captured image that include the tympanic membrane,
a diagnosis model configured to receive one or more of the captured and segmented images and produce an output that can be indicative of the presence of one or more conditions of the middle ear space that comprise: normal, having fluid, having infection, or the ear canal having blockage.