Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
DETAILED ACTION
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
Claims 1-4, 6-10, 12-16 and 18 are rejected under 35 U.S.C. 103 as being unpatentable over Oishila Bandyopadhyay et all (Long-bone fracture detection in digital X-ray images based on digital-geometric techniques) in view of Aaron Pries, et all (DeepMorphing:Detectingbonestructuresinfluoroscopicx-rayimageswithpriorknowledge)).
Regarding claim 1, Oishila discloses a device for assisting with bone fracture detection, the device comprising a processing unit configured to (see abstract, automated fracture detection is an essential part of a computer-aided tele-medicine system. In this paper, we have proposed a unified technique for the detection and evaluation of orthopedic fractures in long-bone digital X-ray image. We have also developed a software tool that can be conveniently used by paramedics or specialist doctors. The proposed tool first segments the bone region of an input digital X-ray image from its surrounding flesh region and then generates the bone-contour using an adaptive thresholding approach. Next, it performs unsupervised correction of bone-contour discontinuities that might have been generated because of segmentation errors, and finally detects the presence of fracture in the bone.
Also page 9, column 2, Fig. 12 concavity curve (a) Input X-ray image of humerus, (b) bone-contour of (a), (c) concavity Index plot of (b), (d) input X-ray image of fractured humerus, (e) bone-contour of (d), and (f) concavity Index plot of (e));
receive image data of a medical image from a device selected from the group consisting of a C-arm based X-ray device, a diagnostic X-ray device, a computer tomography device, a magnet resonance device, an ultrasound device, a data detection device or an image storage device (see abstract, also page 9, column 2, Fig. 12, automated fracture detection is an essential part of a computer-aided tele-medicine system. In this paper, we have proposed a unified technique for the detection and evaluation of orthopedic fractures in long-bone digital X-ray image. We have also developed a software tool that can be conveniently used by paramedics or specialist doctors. The proposed tool first segments the bone region of an input digital X-ray image from its surrounding flesh region and then generates the bone-contour using an adaptive thresholding approach. Next, it performs unsupervised correction of bone-contour discontinuities that might have been generated because of segmentation errors, and finally detects the presence of fracture in the bone. Fig. 12 concavity curve (a) Input X-ray image of humerus, (b) bone-contour of (a), (c) concavity Index plot of (b), (d) input X-ray image of fractured humerus, (e) bone-contour of (d), and (f) concavity Index plot of (e));
identify a bone structure in the medical image; utilize a neural net for detecting a bone fracture based on a bone feature of the bone structure, wherein the neural net is trained on the basis of a multiplicity of images including at least one of medical images and simulated medical images of bones of interest (see above, also page 3, column 1, section 1. automated X-ray analysis is an integral part of telemedicine that is widely used in the rural areas where there is an acute shortage of medical practitioners. An analysis of an orthopedic X-ray image can help the paramedical staff in remote areas to identify the severity of the condition and take proper action in a timely fashion. X-ray analysis can be done in two different phases: (i) segmentation of bone region from the surrounding flesh component followed by the extraction of bone-contour, and (ii) identification of the presence of fracture in the bone X-ray, if any.
Also, page 11, section 4.3.3. detection of fracture types the proposed method categorizes a fracture on the basis of the severity of injury. Evidences from the medical literature indicate that a simple fracture results in a single break with loosely-connected fragments, whereas, in the case of complex or compound fracture, multiple bone splinters appear with external skin injury. Analysis on the segmented bone-contour to count the number of disconnected components in the fractured bone. A fracture that shows up a single connected component is “classified” as being simple and that with multiple or disjoint components is classified as being complex. Further, a fracture may appear in the top (proximal), middle (diaphyseal), or bottom (distal) region of the bone. Also, page 13, column 2, section 7. We have proposed an integrated tool for contour correction, fracture detection, line-of-break identification and for the classification of long-bone fractures based on digital geometric analysis of an input X-ray image. We have shown that certain geometric features can be successfully utilized for providing fast and accurate solutions to the automation of medical image analysis. Our experiments on several X-ray image databases demonstrate the suitability and effectiveness of the proposed approach for the purpose of fracture detection in human long-bone structures. This work can further be extended in the future for automated fracture classification and for virtual simulation of orthopedic surgery);
wherein the bone fracture detection is selected from the group consisting of detection of a point of a fracture line on an outline of the identified bone structure, detection of the fracture line, detection of a displacement of bone parts relative to each other, or a combination thereof (see above, also page 9, section 4.3., the proposed technique identifies the presence of fracture by analyzing the changes in concavity-indices of the pixels on the bone-contour. Once the presence of fracture is confirmed, the break-points and the region of break are identified. Finally, the line-of-break is reconstructed by joining the break-points to mark the region of fracture. Also, page 6, section 4.1.1 bone region segmentation from an X-ray image Extraction of bone region from overlapping muscles and tissue shadow is a challenging task in digital X-ray image segmentation. Each X-ray image has three distinct regions - a black background region, a low-intensity flesh region, and a high-intensity bone region. We have observed that traditional thresholding-based approaches such as ‘OTSU’ [25] often result in over-segmentation or under-segmentation for most of the X-ray images. Also, an “edge-detection” approach such as ‘Canny’ method is observed to generate multiple false edges because of the presence of flesh-shadow surrounding the bone components. Another widely used segmentation method is k-means clustering [15,16] that segments the region of interest (ROI) more efficiently than classical segmentation approaches such as ‘Canny’ or ‘OTSU’. We have evaluated k-means clustering and an entropy-based method for bone region segmentation.
Finally, page 3, column 2 paragraph, 3, among others, mathematical morphology has been widely used in bone-fracture detection. A computer-aided system for long-bone fracture detection had been proposed by Donnelley et al, where a scale-space approach is used for edge detection, Hough transform for parameter approximation, and gradient analysis for diathesis segmentation and fracture detection. A frequently used data mining technique called classification, had also been widely used to detect the presence of fracture. These systems combine various image features such shape, texture, and color, and deploy machine learning algorithms to identify fractures. Also, page 7, column 1, paragraph, first Fig. 6 – Segmentation and contour generation (a) Input digital X-ray image, (b) bone-contour using k-means segmentation, (c) counter using entropy-standard deviation-based segmentation. segmentation of input X-ray image obtained by the k-means segmentation, and Fig. 6(c) shows the contour of the segmented image obtained by the entropy-based segmentation and adaptive thresholding approach);
and generate a corresponding output based on the bone fracture detection of the neural net (see page 11, section 4.3.3. detection of fracture types the proposed method categorizes a fracture on the basis of the severity of injury. Evidences from the medical literature indicate that a simple fracture results in a single break with loosely-connected fragments, whereas, in the case of complex or compound fracture, multiple bone splinters appear with external skin injury. We perform connected component analysis on the segmented bone-contour to count the number of disconnected components in the fractured bone. A fracture that shows up a single connected component is “classified” as being simple and that with multiple or disjoint components is classified as being complex. Further, a fracture may appear in the top (proximal), middle (diaphyseal), or bottom (distal) region of the bone. The proposed approach identifies the location of fracture-points and categorizes the fracture accordingly. If the points appear within the upper 30% of a bone image, then itis termed as proximal region fracture. Fracture points within the lower 30% of the image indicate distal fracture, and those in the middle part of the bone (middle 40%) represent diaphyseal region fracture. Table 3 reports the type of fractures for several test images).
However, regarding claim 1, Oishila discloses page 11, section 4.3.3. detection of fracture types the proposed method categorizes a fracture on the basis of the severity of injury. Evidences from the medical literature indicate that a simple fracture results in a single break with loosely-connected fragments, whereas, in the case of complex or compound fracture, multiple bone splinters appear with external skin injury. Analysis on the segmented bone-contour to count the number of disconnected components in the fractured bone. A fracture that shows up a single connected component is “classified” as being simple and that with multiple or disjoint components is classified as being complex. Further, a fracture may appear in the top (proximal), middle (diaphyseal), or bottom (distal) region of the bone. Also, page 13, column 2, section 7. We have proposed an integrated tool for contour correction, fracture detection, line-of-break identification and for the classification of long-bone fractures based on digital geometric analysis of an input X-ray image. We have shown that certain geometric features can be successfully utilized for providing fast and accurate solutions to the automation of medical image analysis. Our experiments on several X-ray image databases demonstrate the suitability and effectiveness of the proposed approach for the purpose of fracture detection in human long-bone structures. This work can further be extended in the future for automated fracture classification and for virtual simulation of orthopedic surgery).
However, Oishila does not explicitly state “neural net”.
On the other hand, Aaron in the same field of “detecting bone structures in X-ray images”, teaches (see page 2, first paragraphs, that uses “deep learning” combined with prior information about the shape of the bone. We will use fluoroscopic X-ray images for our evaluation, but our approach can be applied to other 2D and 3D modalities as well. A classical solution to localize anatomical points on a contour is the active shape model (ASM) algorithm. for the ASM algorithm, the variation of the outline and the gray-level appearances near the outline are learned from training data. Also Fig. 3, page 5, initialization of the model: Prediction of a DNN (Deep Neural Network), with the mean shape fitted to the foreground pixels. (b) Update along lines orthogonal to the current shape. With the mean shape fitted to the foreground pixels. Figure 3: Steps during the shape model fit, that we reach a good tradeoff between fitting the observed data and staying close to shape variations observed in the training data. This will be described in Section 3.2.
It would have been obvious to one of ordinary skill in the art before the effective filing date of the claimed invention, that to combine the detecting and classifying bone fractures in X-ray images by detecting the bone contour and outline, and classifying the bone fracture as taught by the Oishila with the classification using deep learning neural network to detect and classify fractures as taught by Aaron for a faster, more efficient system and method for detecting, identifying and classifying bone fractures.
Regarding claim 2, Oishila discloses the device of claim 1, wherein the neural net is further trained with information from a data base that is selected from a group consisting of a bone model, possible fracture lines and their likelihood, possible bone fragmentations and their likelihood, a classification of bone fractures or a combination thereof (see claim 1, also page 11, section 4.3.3. detection of fracture types the proposed method categorizes a fracture on the basis of the severity of injury. Evidences from the medical literature indicate that a simple fracture results in a single break with loosely-connected fragments, whereas, in the case of complex or compound fracture, multiple bone splinters appear with external skin injury. Analysis on the segmented bone-contour to count the number of disconnected components in the fractured bone. A fracture that shows up a single connected component is “classified” as being simple and that with multiple or disjoint components is classified as being complex. Further, a fracture may appear in the top (proximal), middle (diaphyseal), or bottom (distal) region of the bone. Also, page 13, column 2, section 7. We have proposed an integrated tool for contour correction, fracture detection, line-of-break identification and for the classification of long-bone fractures based on digital geometric analysis of an input X-ray image. We have shown that certain geometric features can be successfully utilized for providing fast and accurate solutions to the automation of medical image analysis. Our experiments on several X-ray image databases demonstrate the suitability and effectiveness of the proposed approach for the purpose of fracture detection in human long-bone structures. This work can further be extended in the future for automated fracture classification and for virtual simulation of orthopedic surgery).
Regarding claim 3, Oishila discloses the device of claim 1, wherein the output includes information of the likely location of the detected bone fracture (see claim 1, also abstract, automated fracture detection is an essential part of a computer-aided tele-medicine system. In this paper, we have proposed a unified technique for the detection and evaluation of orthopedic fractures in long-bone digital X-ray image. We have also developed a software tool that can be conveniently used by paramedics or specialist doctors. The proposed tool first segments the bone region of an input digital X-ray image from its surrounding flesh region and thengeneratesthebone-ontourusinganadaptivethresholdingapproach.Next, it performs unsupervised correction of bone-contour discontinuities that might have been generated because of segmentation errors, and finally detects the presence of fracture in the bone. Moreover, the method can also localize the line-of-break for easy visualization of the fracture, identify its orientation, and assess the extent of damage in the bone .Several concepts from digital geometry such as relaxed straightness and concavity index are utilized to correct contour imperfections, and to “detect fracture locations” and type).
Regarding claim 4, Oishila discloses the device of claim 1, wherein the multiplicity of images shows bones with the same type of fracture (see claim 1, also page 3, column 2paragraph, 3, among others, mathematical morphology has been widely used in bone-fracture detection. A computer-aided system for long-bone fracture detection had been proposed by Donnelley et al, where a scale-space approach is used for edge detection, Hough transform for parameter approximation, and gradient analysis for diathesis segmentation and fracture detection. A frequently used data mining technique called classification, had also been widely used to detect the presence of fracture. These systems combine various image features such shape, texture, and color, and deploy machine learning algorithms to identify fractures.
Also, page 7, column 1, paragraph, first Fig. 6 – Segmentation and contour generation (a) Input digital X-ray image, (b) bone-contour using k-means segmentation, (c) counter using entropy-standard deviation-based segmentation. segmentation of input X-ray image obtained by the k-means segmentation, and Fig. 6(c) shows the contour of the segmented image obtained by the entropy-based segmentation and adaptive thresholding approach).
Regarding claim 6, Oishila discloses the device of claim 1, further comprising an input device, wherein the processing unit is further configured to receive a user-prompted input from the input device, wherein the user-prompted input is selected from a group consisting of bone structure identification, bone outline detection, fracture line detection, fracture line starting point detection, bone part displacement detection, bone fracture classification, or a combination thereof (see claim 1, also page 9, section 4.3., the proposed technique identifies the presence of fracture by analyzing the changes in concavity-indices of the pixels on the bone-contour. Once the presence of fracture is confirmed, the break-points and the region of break are identified. Finally, the line-of-break is reconstructed by joining the break-points to mark the region of fracture. Also, page 6, section 4.1.1 bone region segmentation from an X-ray image Extraction of bone region from overlapping muscles and tissue shadow is a challenging task in digital X-ray image segmentation. Each X-ray image has three distinct regions - a black background region, a low-intensity flesh region, and a high-intensity bone region. We have observed that traditional thresholding-based approaches such as ‘OTSU’ [25] often result in over-segmentation or under-segmentation for most of the X-ray images. Also, an “edge-detection” approach such as ‘Canny’ method is observed to generate multiple false edges because of the presence of flesh-shadow surrounding the bone components. Another widely used segmentation method is k-means clustering [15,16] that segments the region of interest (ROI) more efficiently than classical segmentation approaches such as ‘Canny’ or ‘OTSU’. We have evaluated k-means clustering and an entropy-based method for bone region segmentation.
Finally, page 3, column 2 paragraph, 3, among others, mathematical morphology has been widely used in bone-fracture detection. A computer-aided system for long-bone fracture detection had been proposed by Donnelley et al, where a scale-space approach is used for edge detection, Hough transform for parameter approximation, and gradient analysis for diathesis segmentation and fracture detection. A frequently used data mining technique called classification, had also been widely used to detect the presence of fracture. These systems combine various image features such shape, texture, and color, and deploy machine learning algorithms to identify fractures. Also, page 7, column 1, paragraph, first Fig. 6 – Segmentation and contour generation (a) Input digital X-ray image, (b) bone-contour using k-means segmentation, (c) counter using entropy-standard deviation-based segmentation. segmentation of input X-ray image obtained by the k-means segmentation, and Fig. 6(c) shows the contour of the segmented image obtained by the entropy-based segmentation and adaptive thresholding approach).
Regarding claim 8, Oishila discloses the method of claim 7, wherein the neural net is trained with information from a data base that is selected from a group consisting of a bone model, possible fracture lines and their likelihood, possible bone fragmentations and their likelihood, a classification of bone fractures or a combination thereof (see claim 1, also page 11, section 4.3.3. detection of fracture type the proposed method categorizes a fracture on the basis of the severity of injury. Evidences from the medical literature indicate that a simple fracture results in a single break with loosely-connected fragments, whereas, in the case of complex or compound fracture, multiple bone splinters appear with external skin injury. Analysis on the segmented bone-contour to count the number of disconnected components in the fractured bone. A fracture that shows up a single connected component is “classified” as being simple and that with multiple or disjoint components is classified as being complex. Further, a fracture may appear in the top (proximal), middle (diaphyseal), or bottom (distal) region of the bone. Also, page 13, column 2, section 7. We have proposed an integrated tool for contour correction, fracture detection, line-of-break identification and for the classification of long-bone fractures based on digital geometric analysis of an input X-ray image. We have shown that certain geometric features can be successfully utilized for providing fast and accurate solutions to the automation of medical image analysis. Our experiments on several X-ray image databases demonstrate the suitability and effectiveness of the proposed approach for the purpose of fracture detection in human long-bone structures. This work can further be extended in the future for automated fracture classification and for virtual simulation of orthopedic surgery).
Regarding claim 9, Oishila discloses the method of claim 7, wherein the output includes information of the likely location of the detected bone fracture (see claim 1, also abstract, automated fracture detection is an essential part of a computer-aided tele-medicine system. In this paper, we have proposed a unified technique for the detection and evaluation of orthopedic fractures in long-bone digital X-ray image. We have also developed a software tool that can be conveniently used by paramedics or specialist doctors. The proposed tool first segments the bone region of an input digital X-ray image from its surrounding flesh region and thengeneratesthebone-ontourusinganadaptivethresholdingapproach.Next, it performs unsupervised correction of bone-contour discontinuities that might have been generated because of segmentation errors, and finally detects the presence of fracture in the bone. Moreover, the method can also localize the line-of-break for easy visualization of the fracture, identify its orientation, and assess the extent of damage in the bone .Several concepts from digital geometry such as relaxed straightness and concavity index are utilized to correct contour imperfections, and to “detect fracture locations” and type).
Regarding claim 10, Oishila discloses the method of claim 7, wherein the multiplicity of images shows bones with the same type of fracture (see claim 1, also page 3, column 2paragraph, 3, among others, mathematical morphology has been widely used in bone-fracture detection. A computer-aided system for long-bone fracture detection had been proposed by Donnelley et al, where a scale-space approach is used for edge detection, Hough transform for parameter approximation, and gradient analysis for diathesis segmentation and fracture detection. A frequently used data mining technique called classification, had also been widely used to detect the presence of fracture. These systems combine various image features such shape, texture, and color, and deploy machine learning algorithms to identify fractures.
Also, page 7, column 1, paragraph, first Fig. 6 segmentation and contour generation (a) Input digital X-ray image, (b) bone-contour using k-means segmentation, (c) counter using entropy-standard deviation based segmentation. segmentation of input X-ray image obtained by the k-means segmentation, and Fig. 6(c) shows the contour of the segmented image obtained by the entropy-based segmentation and adaptive thresholding approach).
Regarding claim 12, Oishila discloses the method of claim 7, further comprising receiving a user-prompted input by means of an input device, wherein the user-prompted input is selected from the group consisting of bone structure identification, bone outline detection, fracture line detection, fracture line starting point detection, bone part displacement detection, and bone fracture classification (see claim 1, also page 9, section 4.3., the proposed technique identifies the presence of fracture by analyzing the changes in concavity-indices of the pixels on the bone-contour. Once the presence of fracture is confirmed, the break-points and the region of break are identified. Finally, the line-of-break is reconstructed by joining the break-points to mark the region of fracture. Also, page 6, section 4.1.1 bone region segmentation from an X-ray image Extraction of bone region from overlapping muscles and tissue shadow is a challenging task in digital X-ray image segmentation. Each X-ray image has three distinct regions - a black background region, a low-intensity flesh region, and a high-intensity bone region. We have observed that traditional thresholding-based approaches such as ‘OTSU’ [25] often result in over-segmentation or under-segmentation for most of the X-ray images. Also, an “edge-detection” approach such as ‘Canny’ method is observed to generate multiple false edges because of the presence of flesh-shadow surrounding the bone components. Another widely used segmentation method is k-means clustering [15,16] that segments the region of interest (ROI) more efficiently than classical segmentation approaches such as ‘Canny’ or ‘OTSU’. We have evaluated k-means clustering and an entropy-based method for bone region segmentation.
Finally, page 3, column 2 paragraph, 3, among others, mathematical morphology has been widely used in bone-fracture detection. A computer-aided system for long-bone fracture detection had been proposed by Donnelley et al, where a scale-space approach is used for edge detection, Hough transform for parameter approximation, and gradient analysis for diathesis segmentation and fracture detection. A frequently used data mining technique called classification, had also been widely used to detect the presence of fracture. These systems combine various image features such shape, texture, and color, and deploy machine learning algorithms to identify fractures. Also, page 7, column 1, paragraph, first Fig. 6 – Segmentation and contour generation (a) Input digital X-ray image, (b) bone-contour using k-means segmentation, (c) counter using entropy-standard deviation-based segmentation. segmentation of input X-ray image obtained by the k-means segmentation, and Fig. 6(c) shows the contour of the segmented image obtained by the entropy-based segmentation and adaptive thresholding approach).
Regarding claim 14, Oishila discloses the non-transitory, computer-readable memory of claim 13, wherein the neural net is trained with information from a data base that is selected from a group consisting of a bone model, possible fracture lines and their likelihood, possible bone fragmentations and their likelihood, a classification of bone fractures or a combination thereof (see claim 1, also page 11, section 4.3.3. detection of fracture types the proposed method categorizes a fracture on the basis of the severity of injury. Evidences from the medical literature indicate that a simple fracture results in a single break with loosely-connected fragments, whereas, in the case of complex or compound fracture, multiple bone splinters appear with external skin injury. Analysis on the segmented bone-contour to count the number of disconnected components in the fractured bone. A fracture that shows up a single connected component is “classified” as being simple and that with multiple or disjoint components is classified as being complex. Further, a fracture may appear in the top (proximal), middle (diaphyseal), or bottom (distal) region of the bone. Also, page 13, column 2, section 7. We have proposed an integrated tool for contour correction, fracture detection, line-of-break identification and for the classification of long-bone fractures based on digital geometric analysis of an input X-ray image. We have shown that certain geometric features can be successfully utilized for providing fast and accurate solutions to the automation of medical image analysis. Our experiments on several X-ray image databases demonstrate the suitability and effectiveness of the proposed approach for the purpose of fracture detection in human long-bone structures. This work can further be extended in the future for automated fracture classification and for virtual simulation of orthopedic surgery).
Regarding claim 15, Oishila discloses the non-transitory, computer-readable memory of claim 13, wherein the output includes information of the likely location of the detected bone fracture (see claim 1, also abstract, automated fracture detection is an essential part of a computer-aided tele-medicine system. In this paper, we have proposed a unified technique for the detection and evaluation of orthopedic fractures in long-bone digital X-ray image. We have also developed a software tool that can be conveniently used by paramedics or specialist doctors. The proposed tool first segments the bone region of an input digital X-ray image from its surrounding flesh region and thengeneratesthebone-ontourusinganadaptivethresholdingapproach.Next, it performs unsupervised correction of bone-contour discontinuities that might have been generated because of segmentation errors, and finally detects the presence of fracture in the bone. Moreover, the method can also localize the line-of-break for easy visualization of the fracture, identify its orientation, and assess the extent of damage in the bone .Several concepts from digital geometry such as relaxed straightness and concavity index are utilized to correct contour imperfections, and to “detect fracture locations” and type).
Regarding claim 16, Oishila discloses the non-transitory, computer-readable memory of claim 13, wherein the multiplicity of images shows bones with the same type of fracture (see claim 1, also page 3, column 2paragraph, 3, among others, mathematical morphology has been widely used in bone-fracture detection. A computer-aided system for long-bone fracture detection had been proposed by Donnelley et al, where a scale-space approach is used for edge detection, Hough transform for parameter approximation, and gradient analysis for diathesis segmentation and fracture detection. A frequently used data mining technique called classification, had also been widely used to detect the presence of fracture. These systems combine various image features such shape, texture, and color, and deploy machine learning algorithms to identify fractures.
Also, page 7, column 1, paragraph, first Fig. 6 – Segmentation and contour generation (a) Input digital X-ray image, (b) bone-contour using k-means segmentation, (c) counter using entropy-standard deviation based segmentation. segmentation of input X-ray image obtained by the k-means segmentation, and Fig. 6(c) shows the contour of the segmented image obtained by the entropy-based segmentation and adaptive thresholding approach).
Regarding claim 18, Oishila discloses the non-transitory, computer-readable memory of claim 13, wherein the set of instruction that, when executed on a processing unit, performs a method further comprising receiving a user-prompted input by means of an input device, wherein the user-prompted input is selected from the group consisting of bone structure identification, bone outline detection, fracture line detection, fracture line starting point detection, bone part displacement detection, and bone fracture classification (see claim 1, also page 9, section 4.3., the proposed technique identifies the presence of fracture by analyzing the changes in concavity-indices of the pixels on the bone-contour. Once the presence of fracture is confirmed, the break-points and the region of break are identified. Finally, the line-of-break is reconstructed by joining the break-points to mark the region of fracture. Also, page 6, section 4.1.1 bone region segmentation from an X-ray image Extraction of bone region from overlapping muscles and tissue shadow is a challenging task in digital X-ray image segmentation. Each X-ray image has three distinct regions - a black background region, a low-intensity flesh region, and a high-intensity bone region. We have observed that traditional thresholding-based approaches such as ‘OTSU’ [25] often result in over-segmentation or under-segmentation for most of the X-ray images. Also, an “edge-detection” approach such as ‘Canny’ method is observed to generate multiple false edges because of the presence of flesh-shadow surrounding the bone components. Another widely used segmentation method is k-means clustering [15,16] that segments the region of interest (ROI) more efficiently than classical segmentation approaches such as ‘Canny’ or ‘OTSU’. We have evaluated k-means clustering and an entropy-based method for bone region segmentation.
Finally, page 3, column 2 paragraph, 3, among others, mathematical morphology has been widely used in bone-fracture detection. A computer-aided system for long-bone fracture detection had been proposed by Donnelley et al, where a scale-space approach is used for edge detection, Hough transform for parameter approximation, and gradient analysis for diathesis segmentation and fracture detection. A frequently used data mining technique called classification, had also been widely used to detect the presence of fracture. These systems combine various image features such shape, texture, and color, and deploy machine learning algorithms to identify fractures. Also, page 7, column 1, paragraph, first Fig. 6 – Segmentation and contour generation (a) Input digital X-ray image, (b) bone-contour using k-means segmentation, (c) counter using entropy-standard deviation-based segmentation. segmentation of input X-ray image obtained by the k-means segmentation, and Fig. 6(c) shows the contour of the segmented image obtained by the entropy-based segmentation and adaptive thresholding approach).
With regard to claims 7 and 13, the arguments analogous to those presented above for claims 1, 2, 3, 4, 6, 8, 9, 10, 12, 14, 15, 16 and 18 are respectively applicable to claims 7 and 13.
Allowable Subject Matter
Claims 5, 11 and 17 are objected to as being dependent upon a rejected base claim, but would be allowable if rewritten in independent form including all of the limitations of the base claim and any intervening claims.
Contact Information
Any inquiry concerning this communication or earlier communications from the examiner should be directed to Seyed Azarian whose telephone number is (571) 272-7443. The examiner can normally be reached on Monday through Thursday from 6:00 a.m. to 7:30 p.m.
If attempts to reach the examiner by telephone are unsuccessful, the examiner's supervisor, Matthew Bella, can be reached at (571) 272-7778. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of an application may be obtained from the Patent Application information Retrieval (PAIR) system. Status information for published application may be obtained from either Private PAIR or Public PAIR.
Status information about the PAIR system, see http:// pair-direct.uspto.gov. Should you have questions on access to the Private PAIR system, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free).
/SEYED H AZARIAN/Primary Examiner, Art Unit 2667
July 21, 2026
Prosecution Insights