Patent Application
20220202353
The present invention is directed, in general, to the dermatology field. In particular, the invention relates to a dermatoscopy device, and to a method, for checking skin lesions.
International patent application WO-A1-2013169326 discloses an imaging station/booth for automated total body imaging having a small footprint and capable of quickly, efficiently, effectively, and consistently capturing multiple body images of a user or patient over time with minimal assistance from medical staff.
International patent application WO-A1-2018160720 relates to three-dimensional image capture, particularly with the use of polarized light. The capture of three-dimensional (3D) images of a subject, such as a human, typically entails the use of multiple cameras and light sources, arranged around the subject, that are activated simultaneously while the subject remains still. The number of cameras and light sources required depends in direct relation on the size of the area to be imaged.
International patent application WO-A1-2010107467 discloses an imaging station and method for repeatable alignment of images. In some embodiments, the imaging station includes a rotatable stage and a monostand. The rotatable stage may include a structure to support sets of handles for positioning of a subject's hands for at least some of the body poses. A camera is positioned on the monostand for capturing images of the subject and is configured to allow the camera to be repeatably positioned with respect to the rotatable stage for each body pose.
US patent US-B1-U.S. Pat. No. 9,980,649 discloses a scan head for scanning skin. The scan head includes a frame and a camera coupled to the frame. A controllable probe is coupled to the frame and is configured to change an orientation of hair on the skin to be examined and imaged with the camera.
International patent application WO-A1-2016188659 discloses a method and a device for producing a series of images of an object with respective predetermined imaging parameters by means of a partially automated imaging system, comprising a database, an imaging unit, a computing unit, a storage unit and an output unit, wherein it is ensured that a manually to be adjusted imaging parameter corresponds to an actual imaging parameter.
Scientific document “A new total body scanning system for automatic change detection in multiple pigmented skin lesion”, K. Korotkov et al. discloses a new photogrammetry-based total body scanning system that allows for skin surface image acquisition using cross-polarized light. The system is equipped with 21 high-resolution cameras and a turntable. This scanner automatically acquires a set of overlapping images, covering 85%-90% of the patient's skin surface. These images are used for the automated mapping of PSLs and their change estimation between explorations. The scanner can be applied for automated mapping and temporal monitoring of multiple lesions.
In the known dermatoscopy imaging devices/stations, a single dermatoscopic image is acquired invasively, i.e. not acquired from a distance at least greater than 1-3 mm. Moreover, the known imaging devices/stations do not allow operating in fully controlled lighting conditions using a single image acquisition system comprising different types of cameras.
New imaging devices and method for checking skin lesions in a location with controlled light and image acquisition are therefore needed.
The object of this invention is fulfilled by a dermatoscopy device with the characteristics of claim 1 and by a method with the characteristics of claim 10.
The present invention proposes, according to one aspect, a dermatoscopy device for checking skin lesions. Similar to the devices known in the field, the proposed dermatoscopy device comprises:
Unlike the known devices in the field, the image acquisition system of the proposed device comprises two different types of cameras. A first type of the cameras includes a stereoscopic camera or two cameras configured to work/operate as a stereoscopic camera. A second type of the cameras comprises a dermatoscopic camera (e.g. a long-range dermatoscopic camera). The camera(s) of the first type is/are configured to operate simultaneously (if there are more than one) and independently of the dermatoscopic camera.
The lighting elements comprise at least two lighting elements each one attached on an extreme of the support of the displacement element. Thus, in the proposed device, the images are always acquired under controlled lighting conditions.
Moreover, in the proposed device the cited processing comprises analyzing, for each body region, at least one image acquired by the camera(s) of the first type; executing a scanning algorithm that detects skin structures (e.g. freckles, skin marks, among others) in said image and that checks whether said skin structures fulfill a given criteria; upon a given skin structure of a given body region is determined as fulfilling the given criteria, executing a motion algorithm that automatically moves the displacement element to the given body region; and providing a triggering order to the image acquisition system. Therefore, as a result, the dermatoscopic camera can acquire, at or from a determined distance, at least one image of the given skin structure.
Therefore, present invention provides a medical device that provides automated and accurate clinical and dermatoscopic images of a patient's skin. Moreover, the dermatoscopic images of individual skin lesions are acquired in a standardized way (controlled light and acquisition distance). In addition, the present invention provides the physician/doctor with the necessary information such as the color, dimension, shape, asymmetry and evolution of each of the skin structures in order to make an accurate diagnosis.
The camera(s) of the first type can be a passive stereo camera(s), an active stereo camera(s) with a structured light source or a laser, a time-of-flight (TOF) camera(s) or the combination of a LIDAR and a single camera.
In an embodiment, the determined distance is comprised in a range between 100-1000 mm from the user, particularly between 300-600 mm. Therefore, the acquisition is performed non-invasively.
In an embodiment, the displacement element is arranged on a guiding railway of a guiding system. The guiding system is arranged on an interior wall of the enclosed chamber. Particularly, the guiding system comprises two vertical guide members.
In another embodiment, each of the lighting elements comprises a light emitting diode (LED). The lighting elements also include a light confinement chamber in order to only illuminate a given zone when the image/s is/are acquired. Furthermore, the lighting elements also include a polarization filter, which is 90° out of phase with respect to a polarization filter of the dermatoscopic camera in order to generate a cross polarization.
The lighting elements in an embodiment are separated a certain distance on said support in order to operate at a given angle (between 35 and 55°).
In yet another embodiment, the proposed device also has a positioning unit to indicate a position in which the user has to be placed inside the enclosed chamber. For example, the position unit can be a floor display or even a loudspeaker which guides the user to the different positions. Alternatively, the positioning unit can be configured to automatically move the user towards the interior wall of the enclosed chamber.
The displacement element in some embodiments can also include an articulable or pivotable element that allows the displacement element to be moved in multiple degrees of freedom.
Embodiments of the present invention also propose, according to another aspect, a method for checking skin lesions. The method comprises a) accommodating a user to be examined within an enclosed chamber of a dermatoscopy device; b) moving, via a displacement element of the dermatoscopy device, an image acquisition system of the dermatoscopy device towards different body parts of a given body side of the user; c) acquiring, by at least one camera of a first type of the image acquisition system, one or more images of the different body parts of the user, the at least one camera of the first type comprising a stereoscopic camera or two cameras operating as a stereoscopic camera, and the one or more images being stereoscopic images that are acquired under controlled lighting conditions; d) analyzing, by a computing unit, the acquired stereoscopic images by executing a scanning algorithm that checks whether skin structures included in the stereoscopic images fulfill a given criteria; e) upon a given skin structure of a given body region is determined as fulfilling the given criteria, executing, by the computing unit, a motion algorithm that automatically moves the displacement element towards the given body region; f) in response to a triggering order from the computing unit, acquiring, by a camera of a second type of the image acquisition system, which is a dermatoscopic camera, at/from a determined distance, at least one dermatoscopic image of the given skin structure; and g) providing the acquired at least one dermatoscopic image using the computing unit.
According to the proposed method, said step b) is performed for all the body sides of the user.
In an embodiment, the cited criteria is based on an ABCD rule of dermatoscopy based on the criteria asymmetry, A, border, B, color, C, and differential structure, D, of the skin structures.
In another embodiment, the scanning algorithm comprises a neural network. The neural network is particularly trained to automatically recognize if the skin structure looks like a benign or malign structure. In other embodiments, the scanning algorithm comprises a decision tree-based algorithm and/or a support vector machine.
The previous and other advantages and features will be more fully understood from the following detailed description of embodiments, with reference to the attached figures, which must be considered in an illustrative and non-limiting manner, in which:
The computing unit 30 is operatively connected to the displacement element 12 and to the image acquisition system 20 and is adapted and configured to implement/execute different algorithms in order to control operation thereof. It should be noted that in other embodiments, the computing unit 30 instead of being included in the dermatoscopy device 1 can be located remote therefrom. For example, in this latter case, the computing unit can be a PC, a cloud server, a Smartphone, a Tablet, etc. placed distant from the dermatoscopy device 1.
In the particular embodiment of
In other embodiments (not showed), the image acquisition system 20 includes a single stereoscopic camera, for example consisting of a structured-light 3D sensor, or active stereo camera with a structured light source or a laser, a passive stereo camera, a TOF camera, etc.
In other embodiments (not showed either) the dermatoscopy device 1 includes a higher number of lighting elements, for example four or more lighting elements.
With regard to the lighting elements 13, 14, each one comprises a high-power LED, with a light confinement chamber and a refrigerating system (i.e. each LED element has a radiator and a fan attached). Moreover, each lighting element 13, 14 has a polarization filter having a 90° offset with respect to a polarization filter of the dermatoscopic camera 23. Thus, a cross polarization is generated.
The lighting elements 13, 14 are arranged on the cited support 11 separated from the image acquisition system 20 in order to operate at a given angle (between 35° and 55°) considering the operating distance of the dermatoscopic camera 23.
In the embodiment of
With reference now to
At step 501, the computing unit analyzes the acquired stereoscopic image(s) 31 by executing a scanning algorithm that checks whether skin structures (e.g. actinic keratosis, moles, skin marks, etc.) included in said stereoscopic image(s) 31 fulfill a given criteria (step 502). If no skin structures fulfilling the criteria are found, the computing unit simply provides (step 503) the results of the previous analysis. Otherwise, i.e. if a given skin structure has been determined as fulfilling the given criteria, the computing unit, at step 504, executes a motion algorithm that automatically moves (and optionally rotates) the displacement element 12, and so the image acquisition system 20 and the lighting elements 13, 14, towards the body part where the skin structure is located. Then, at step 505, the computing unit triggers a command/order to the image acquisition unit 20 in order the dermatoscopic camera 23 acquiring, at a determined distance from the user, particularly at approx. 300-600 mm from the user, at least one dermatoscopic image 32 of said skin structure. Once the dermatoscopic image 32 is acquired, the computing unit, at step 506, provides the dermatoscopic image, for example via a display thereof, via a specific user interface, by printing it, etc. In this sense, a physician/doctor can latter use the dermatoscopic image to make an accurate diagnosis of the skin lesion.
It should be noted that, previous steps are performed for all the body sides of the user, if necessary. For example, for the lateral and dorsal sides of the user, and for different body parts, such as the legs, arms, head, back, etc., of each body side. If the user has to be moved to a given position in order the image/images being acquired from a specific body part or body side, a further step of the method could be an indication of such a position.
In certain embodiments, said criteria is based on an ABCD rule of dermatoscopy based on the criteria asymmetry, A, border, B, color, C, and differential structure, D, of the skin structures. Complementarily or alternatively, in certain embodiments, the scanning algorithm can comprise a neural network, a decision tree-based algorithm and/or a support vector machine.
In an embodiment, the scanning algorithm is executed/implemented on at least one clinical image and particularly also uses additional features of the skin lesion (for example, size and/or volume of the skin lesion). Then, the algorithm performs, at least, a classification task, and as a result, provides a value of “potential malignity” of a given lesion. The criteria can be then based on that value and a given threshold.
In other embodiments, in the particular case that two cameras 21, 22 configured to operate as a stereoscopic camera are used, the image(s) 31 acquired by each camera are analyzed in parallel. Once the scanning algorithm has detected a skin structure in one of the stereoscopic images 31, it matches the position thereof in the other stereoscopic image 31 in order to identify the exact position of the skin structure.
Those skilled in the art will recognize that the present teachings are amenable to a variety of modifications and/or enhancements. All applications, modifications and alterations required to be protected in the claims may be within the protection scope of the present disclosure.
The scope of the present invention is defined in the following set of claims.
| Number | Date | Country | Kind |
|---|---|---|---|
| 19382687.2 | Aug 2019 | EP | regional |
| Filing Document | Filing Date | Country | Kind |
|---|---|---|---|
| PCT/EP2020/069850 | 7/14/2020 | WO | 00 |
