公开(公告)号：US20250148773A1

公开(公告)日：2025-05-08

申请号：US18935742

申请日：2024-11-04

Applicant: Qure.ai Technologies Private Limited

Inventor： Preetham Putha ,  Swetha Tanamala ,  Sahil Sahil ,  Ravi Kumar Kushawaha

IPC: G06V10/82 ,  G06V10/26 ,  G06V10/764 ,  G06V40/10 ,  G16H15/00

Abstract: The present disclosure relates to a system (100) and method for providing bounding box to fractured body parts in musculoskeletal X-ray images. The system (100) comprises a memory unit (203) and processor (201) with a data collection module (205) gathering input X-ray data (101). Additionally, it includes a fracture detection module (206) consisting of a segmentation model (207) and an object detection model (208). The segmentation model (207) determines segmentation scores to each pixel in the X-ray image and generates a segmentation mask (404) by comparing the segmentation score of each pixel (403) with a segmentation threshold of a fracture. Simultaneously, the object detection model (208) determines the rectangular coordinates of one or more body parts within the X-ray image (101). The fracture detection module (206) then overlaps the segmentation mask (404) with the body part coordinates, resulting in bounding boxes for each fractured body part.

公开(公告)号：US12299887B1

公开(公告)日：2025-05-13

申请号：US18935765

申请日：2024-11-04

Applicant: Qure.ai Technologies Private Limited

Inventor： Preetham Putha ,  Swetha Tanamala ,  Sahil Sahil ,  Ravi Kumar Kushawaha

IPC: G06T7/00 ,  G06V10/764 ,  G06V10/82 ,  G06V40/10

Abstract: An invention relates to a system (100) for classifying presence of fracture and body parts on musculoskeletal X-ray. The system (100) is based on an artificial intelligence for targeting fracture classification and body part. Further, the system comprises a data collection module (205) and a fracture classification module (206). Furthermore, the fracture classification module (206) may comprise a classification model (207) and a segmentation model (208). The classification model (207) may be configured to generate a classification score and the segmentation model (208) may be configured to generate segmentation score. The fracture classification module (206) may be configured to generate a fracture score (403) and a body part score (404). The fracture classification module (206) may be configured to compare said scores (403, 404) against threshold values to classify fracture presence and classify body parts. The system (100) may enhance the efficiency and accuracy of fracture diagnosis in X-ray images.

公开(公告)号：US20250148596A1

公开(公告)日：2025-05-08

申请号：US18935726

申请日：2024-11-04

Applicant: Qure.ai Technologies Private Limited

Inventor： Preetham Putha ,  Swetha Tanamala ,  Sahil Sahil ,  Ravi Kumar Kushawaha

IPC: G06T7/00 ,  G06V10/764 ,  G06V10/82

Abstract: The present subject matter relates to a system (100) and a method (300) for classifying a type of fracture on musculoskeletal X-ray image (101). The system (100) employs a data collection module, a fracture-type classification module including a classification model (207) and a segmentation model (208) based on an artificial intelligence. The data collection module (205) may collect data corresponding to input X-ray images (101), which are then analyzed by a fracture-type classification module (206) using the classification model (207) and the segmentation model (208). Further, the classification model (207) may calculate the classification score and the segmentation module (208) may calculate the segmentation score. The fracture-type classification module may generate the fracture-type score based on the classification score and segmentation score. The fracture-type classification module may be configured for comparing the fracture-type score with a threshold value for each fracture part separately to classify the type of fracture.

公开(公告)号：US20250148597A1

公开(公告)日：2025-05-08

申请号：US18935765

申请日：2024-11-04

Applicant: Qure.ai Technologies Private Limited

Inventor： Preetham Putha ,  Swetha Tanamala ,  Sahil Sahil ,  Ravi Kumar Kushawaha

IPC: G06T7/00 ,  G06V10/764 ,  G06V10/82 ,  G06V40/10

Abstract: An invention relates to a system (100) for classifying presence of fracture and body parts on musculoskeletal X-ray. The system (100) is based on an artificial intelligence for targeting fracture classification and body part. Further, the system comprises a data collection module (205) and a fracture classification module (206). Furthermore, the fracture classification module (206) may comprise a classification model (207) and a segmentation model (208). The classification model (207) may be configured to generate a classification score and the segmentation model (208) may be configured to generate segmentation score. The fracture classification module (206) may be configured to generate 10 a fracture score (403) and a body part score (404). The fracture classification module (206) may be configured to compare said scores (403, 404) against threshold values to classify fracture presence and classify body parts. The system (100) may enhance the efficiency and accuracy of fracture diagnosis in X-ray images.