Determination of a Probability Indicator Value

Simple SummaryContent extracted from patent full text and abstract with AI.

This invention introduces a method and system that uses image data from tomographic imaging devices (such as MRI, CT, or ultrasound) to automatically analyze an organ or tissue structure, extract detailed imaging features (such as texture, shape, and intensity), and compare these features against reference data from a database. A machine learning algorithm, such as a Random Forest classifier, is used to determine the probability that a person has a genetic predisposition (for example, BRCA1/BRCA2 mutation) related to a specific illness like cancer. The output is a probability indicator value that helps identify those likely to have a hereditary risk, supporting early and targeted interventions.

Use CasesContent extracted from patent full text and abstract with AI.

• Pre-screening of patients for genetic counseling and testing for hereditary cancers such as breast, ovarian, prostate, or colon cancers.
• Decision support for clinicians in personalizing cancer surveillance and treatment, especially in cases with inconclusive family history or clinical data.
• Reducing unnecessary genetic testing by prioritizing patients with a higher image-based likelihood of genetic predisposition.
• Enhanced patient management in oncology units—identifying patients requiring intensive monitoring or preventive interventions based on predicted hereditary risk.
• Integration into tomographic imaging devices and hospital IT systems to provide real-time risk assessments during routine scans.

BenefitsContent extracted from patent full text and abstract with AI.

• Reduces the need for expensive and wide-scale genetic testing by providing a more focused, image-based pre-selection tool.
• Enables earlier identification of genetically predisposed cases, allowing timely and tailored medical care, lead to better treatment outcomes.
• Improves accuracy and reliability compared to approaches based only on family history or clinical background, especially when such data are lacking or incomplete.
• Automates the analysis, reducing workload and subjectivity for clinicians and increasing reproducibility.
• Can be implemented with existing imaging data, making adoption cost-effective and minimally disruptive to current clinical workflows.
• Potential for continual improvement as algorithms are self-learning and can integrate new data over time.

Technical Classifications (CPCs)

Inventors & Applicants

Patent Abstract

The present invention concerns a method (Z) of determining a probability indicator value (PI) representing a probability of a presence of a genetic predisposition of a being for a specific illness. The method (Z) comprises the steps of providing (Y) a number of input image data (ID) of an organ or structure which potentially shows phenomena of the specific illness from a tomographic imaging device (5), analyzing (X) the input image data (ID) by extracting therefrom a number of low-level phenotypical features (PF, PF1, PF2, PF3), providing (W) a plurality of phenotypical reference features (PRF) derived from reference image data of the same organ or structure of a number of reference beings and/or a database (DB) with a number of datasets (DS) representing the phenotypical reference features (PRF), the plurality of phenotypical reference features (PRF) comprising such phenotypical reference features (PRF) which correspond with the phenotypical features (PF, PF1, PF2, PF3) of the input image data (ID), processing (U) the phenotypical features (PF, PF1, PF2, PF3) of the input image data (ID) through a classification and/or comparison algorithm based on the number of phenotypical reference features (PRF) and/or on the datasets (DS), deriving (T) the probability indicator value (PI) from the processing (U). The invention also concerns a determination assembly (9) for that purpose.