Combining domain knowledge and base model for single medical image feature localization

• Main Authors: SHANBHAG, DHAYANAN, SAXENA ANURAG, ANAND DEEPAK
• Format: Patent
• Language: Chinese; English
• Published: 22.08.2025
• Subjects: CALCULATING; COMPUTING; COUNTING; HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATIONTECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING ORPROCESSING OF MEDICAL OR HEALTHCARE DATA; IMAGE DATA PROCESSING OR GENERATION, IN GENERAL; INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTEDFOR SPECIFIC APPLICATION FIELDS; PHYSICS

The invention relates to combining domain knowledge and base models for single medical image feature localization. Single medical image feature localization techniques employing a pre-trained base model and domain knowledge are provided. In an example, a computer-implemented method may include determining an anatomical region of a subject for target features based on a reference spatial relationship between the target features as defined in reference spatial information and based on matching reference pixel features respectively associated with the target features with a corresponding subset of pixel features of a target medical image a position within the target medical image, where the reference pixel feature includes a template image pixel feature extracted from a tagged version of the target feature as included in a template medical image depicting an anatomical region of the reference object. The method further includes generating marker information for the target feature identifying the target feature a