Method and system for managing feature reading and scoring in ultrasound and/or optoacoustic images

A system for analyzing at least one of ultrasound (US) images or optoacoustic (OA) images (US/OA images), comprising:

• a display configured to display at least a first image from at least one of OA images or US images acquired in connection with an examination for a region of interest (ROI), the first image including a lesion, the first image overlaid with an interior ROI outline separating an internal zone from a boundary zone of the ROI, the first image overlaid with an exterior ROI outline separating the boundary zone from a peripheral zone;
• a graphical user interface (GUI);
• memory configured to store program instructions;
• and
• one or more processors configured to execute the programmable instructions to:
• obtain feature scores in connection with at least the boundary zone and peripheral zone of the first image;
• obtain a depth to a posterior margin of the mass;
• apply the feature scores and the depth to a posterior margin of the mass, to a classification model, to obtain at least one of a prognostic result or predictive result indicative of a trait of the lesion, wherein the depth to the posterior margin of the mass to the classification model accounts for reduced energy reaching the posterior margin of the mass;
• and
• output the at least one of the prognostic result or predictive result.

A system for analyzing at least one of ultrasound (US) images or optoacoustic (OA) images (US/OA images), comprising:

• a display configured to display at least a first image from at least one of OA images or US images acquired in connection with an examination for a region of interest (ROI), the first image including a lesion, the first image overlaid with an interior ROI outline separating an internal zone from a boundary zone of the ROI, the first image overlaid with an exterior ROI outline separating the boundary zone from a peripheral zone;
• a graphical user interface (GUI);
• memory configured to store program instructions;
• and
• one or more processors configured to execute the programmable instructions to:
• obtain feature scores in connection with at least the boundary zone and peripheral zone of the first image;
• apply the feature scores to a classification model to obtain at least one of a prognostic result or predictive result indicative of a trait of the lesion;
• and
• output the at least one of the prognostic result or predictive result,
• wherein the feature scores include a US internal zone sound transmission feature score, the memory configured to store the classification model, the classification model configured to overweight the US internal zone sound transmission feature score with respect to at least one other US feature score to obtain, as a prognostic result, an indication of lymph node metastasis, the classification model configured to recognize that shadowing in the US internal zone is indicative of an increased risk of lymph node metastasis, while enhanced sound transmission is indicative of a lowered risk of lymph node metastasis.

A computer implemented method for analyzing at least one of ultrasound (US) images or optoacoustic (OA) images (US/OA images), comprising:

• utilizing one or more processors configured to execute programmable instructions for,
• displaying at least a first image from at least one of OA images or US images acquired in connection with an examination for a region of interest (ROI), the first image overlaid with an interior outline separating an internal zone from a boundary zone of the ROI, the first image overlaid with an exterior outline separating the boundary zone from a peripheral zone of the ROI;
• receiving feature scores in connection with the internal zone, boundary zone and peripheral zone of the first image;
• applying the feature scores to a classification model to obtain at least one of a prognostic result or predictive result indicative of a trait of the lesion;
• and
• outputting the at least one of the prognostic result or predictive result,
• wherein the classification model defines a relation between a combination of the feature scores and a positive predictive value of the at least one of the prognostic result or predictive result, the relation representing a sigmoidal curve having initial, intermediate and final portions, the initial portion corresponding to low positive predictive values (PPVs) for a select combination of the feature scores and exhibiting a first slope increasing at a first rate over the initial portion, the intermediate portion corresponding to intermediate PPVs for the select combination of the feature scores and exhibiting a second slope increasing at a second rate over the intermediate portion, the final portion corresponding to high PPVs for the select combination of the feature scores and exhibiting a third slope decreasing at a third rate over the final portion, the second slope and second rate being greater than the first slope and first rate, respectively.

A computer implemented method for analyzing at least one of ultrasound (US) images or optoacoustic (OA) images (US/OA images), comprising:

• utilizing one or more processors configured to execute programmable instructions for,
• displaying at least a first image from at least one of OA images or US images acquired in connection with an examination for a region of interest (ROI), the first image overlaid with an interior outline separating an internal zone from a boundary zone of the ROI, the first image overlaid with an exterior outline separating the boundary zone from a peripheral zone of the ROI;
• receiving feature scores in connection with the internal zone, boundary zone and peripheral zone of the first image;
• obtaining a depth to a posterior margin of the mass;
• applying the feature scores and the depth to the posterior margin of the mass, to a classification model, to obtain at least one of a prognostic result or predictive result indicative of a trait of the lesion, wherein the depth to the posterior margin of the mass accounts for reduced energy reaching the posterior margin of the mass;
• and
• outputting the at least one of the prognostic result or predictive result.

A computer implemented method for analyzing at least one of ultrasound (US) images or optoacoustic (OA) images (US/OA images), comprising:

• utilizing one or more processors configured to execute programmable instructions for,
• displaying at least a first image from at least one of OA images or US images acquired in connection with an examination for a region of interest (ROI), the first image overlaid with an interior outline separating an internal zone from a boundary zone of the ROI, the first image overlaid with an exterior outline separating the boundary zone from a peripheral zone of the ROI;
• receiving feature scores in connection with the internal zone, boundary zone and peripheral zone of the first image;
• applying the feature scores to a classification model to obtain at least one of a prognostic result or predictive result indicative of a trait of the lesion;
• and
• outputting the at least one of the prognostic result or predictive result,
• wherein the feature scores include a US internal zone sound transmission feature score, the memory configured to store the classification model, the classification model configured to overweight the US internal zone sound transmission feature score with respect to at least one other US feature score to obtain, as a prognostic result, an indication of lymph node metastasis, the classification model configured to recognize that shadowing in the US internal zone is indicative of an increased risk of lymph node metastasis, while enhanced sound transmission is indicative of a lowered risk of lymph node metastasis.

A system, comprising:

• a graphical user interface (GUI) configured to receive user inputs;
• memory configured to store program instructions and to store at least a first image from at least one of OA images or US images acquired in connection with an examination for a region of interest (ROI);
• one or more processors configured to execute the programmable instructions to obtain at least one of a prognostic result or predictive result indicative of a trait of the lesion based on feature scores related to at least one of an internal zone, a boundary zone or a peripheral zone of the ROI in the first image;
• and
• a display configured to display indicia indicative of the at least one of the prognostic result or predictive result, wherein the indicia displayed includes a graphical indicia than extends from 0% false negative ratio (FNR) to 100% likelihood of malignancy (LOM), the graphical indicia separated at a select FNR level into a first segment corresponding to predicted FNR and a second segment corresponding to predicted LOM.

A method, comprising:

• under control of one or more processors configured with program instructions for,
• providing a graphical user interface (GUI) configured to receive user inputs related to at least a first image from at least one of OA images or US images acquired in connection with an examination for a region of interest (ROI);
• obtaining at least one of a prognostic result or predictive result indicative of a trait of the lesion based on feature scores related to at least one of an internal zone, a boundary zone or a peripheral zone of the ROI in the first image;
• and
• displaying indicia indicative of the at least one of the prognostic result or predictive result, wherein the indicia displayed includes a graphical indicia than extends from 0% false negative ratio (FNR) to 100% likelihood of malignancy (LOM), the graphical indicia separated at a select FNR level into a first segment corresponding to predicted FNR and a second segment corresponding to predicted LOM.