Methods and systems for generating a diagnostic report

1 . A method for generating a diagnostic report, comprising:

obtaining raw scanning data of a target object, a reconstructed image of the target object, and clinical information of the target object, wherein the raw scanning data is obtained by scanning the target object using a medical imaging device, and the reconstructed image is generated based on the raw scanning data;

determining, based on the raw scanning data, first lesion information of the target object;

determining, based on the reconstructed image and the first lesion information, second lesion information of the target object;

determining whether the target object includes a lesion based on the first lesion information and the second lesion information;

in response to determining that the target object includes the lesion, generating, based on the raw scanning data and the reconstructed image, a target reconstructed image of the lesion using a simulation scanning model;

obtaining third lesion information by performing feature extraction on the target reconstructed image; and

generating a structured diagnostic report for the target object using a report generation model based on the first lesion information, the second lesion information, the third lesion information, and the clinical information.

2 . The method of claim 1 , wherein:

the first lesion information is determined by processing the raw scanning data using a raw data analysis model;

the second lesion information is determined by processing the reconstructed image and the first lesion information using an image analysis model; and

the third lesion information is determined by processing the target reconstructed image using a lesion analysis model, wherein the raw data analysis model, the image analysis model, and the lesion analysis model are trained machine learning models.

3 . The method of claim 2 , wherein the raw data analysis model and the image analysis model are obtained by:

obtaining sample raw data, a sample reconstructed image corresponding to the sample raw data, first label lesion information corresponding to the sample reconstructed image, and second label lesion information corresponding to the sample reconstructed image;

obtaining the raw data analysis model by training a first preliminary model using the sample raw data and the first label lesion information;

determining sample first lesion information corresponding to the sample raw data based on the sample raw data and the raw data analysis model; and

obtaining the image analysis model by training a second preliminary model using the sample reconstructed image, the sample first lesion information, and the second label lesion information.

4 . The method of claim 1 , wherein

in response to determining that historical diagnostic data relating to the lesion exists, the method further comprises:

determining lesion change information based on the historical diagnostic data, the first lesion information, the second lesion information, and the third lesion information; and

the generating a structured diagnostic report for the target object using a report generation model based on the first lesion information, the second lesion information, the third lesion information, and the clinical information includes:

generating the structured diagnostic report for the target object by processing the first lesion information, the second lesion information, the third lesion information, the clinical information, and the lesion change information using the report generation model.

5 . The method of claim 1 , wherein the simulation scanning model includes a simulation scanning module and an update module,

the simulation scanning module is configured to generate an initial target reconstructed image based on the raw scanning data, the reconstructed image, the first lesion information, the second lesion information, and an initial scanning protocol parameter; and

the update module is configured to generate the target reconstructed image based on the initial target reconstructed image.

6 . The method of claim 5 , wherein to generate the target reconstructed image based on the initial target reconstructed image, the update module is configured to:

determine whether the initial target reconstructed image satisfies a preset condition;

in response to determining that the initial target reconstructed image does not satisfy the preset condition, adjust the initial scanning protocol parameter to generate a target scanning protocol parameter; and

generate the target reconstructed image based on the target scanning protocol parameter.

7 . The method of claim 1 , wherein the simulation scanning model is obtained by:

obtaining a plurality of third training samples and a third preliminary model, each of the plurality of third training samples including sample raw data, a sample reconstructed image, sample first lesion information, sample second lesion information, a sample scanning protocol parameter, and a reference reconstructed image; and

generating the simulation scanning model by training the third preliminary model using the plurality of third training samples based on a first loss function and a second loss function, wherein

during training, for each of the plurality of third training samples:

a value of the first loss function is determined based on the reference reconstructed image and a predicted target reconstructed image generated by the third preliminary model; and

a value of the second loss function is determined based on predicted raw scanning data and the sample raw data, the predicted raw scanning data being determined based on the predicted target reconstructed image.

8 . The method of claim 1 , wherein the first lesion information includes a first lesion detection result and a first confidence map, the second lesion information includes a second lesion detection result and a second confidence map, the third lesion information includes a third lesion detection result and a third confidence map, and

the generating a structured diagnostic report for the target object using a report generation model based on the first lesion information, the second lesion information, the third lesion information, and the clinical information includes:

determining a first weight corresponding to the first lesion detection result, a second weight corresponding to the second lesion detection result, and a third weight corresponding to the third lesion detection result based on the clinical information, the first confidence map, the second confidence map, and the third confidence map; and

generating the structured diagnostic report for the target object by processing the first lesion detection result, the second lesion detection result, the third lesion detection result, the first weight, the second weight, and the third weight using the report generation model.

9 . The method of claim 8 , wherein the determining a first weight corresponding to the first lesion detection result, a second weight corresponding to the second lesion detection result, and a third weight corresponding to the third lesion detection result based on the clinical information, the first confidence map, the second confidence map, and the third confidence map includes:

determining a case type of the target object based on the clinical information;

determining a preset first weight corresponding to the first lesion detection result, a preset second weight corresponding to the second lesion detection result, and a preset third weight corresponding to the third lesion detection result based on the case type; and

determining the first weight, the second weight, and the third weight by adjusting the preset first weight, the preset second weight, and the preset third weight based on the first confidence map, the second confidence map, and the third confidence map.

10 . The method of claim 8 , wherein the structured diagnostic report for the target object includes a diagnostic result and a traceability basis, and the traceability basis indicates information on which the diagnostic result is based.

11 . The method of claim 10 , wherein the report generation model includes:

a feature fusion module, configured to determine a consistency detection result, fused feature information, and an initial traceability basis corresponding to the fused feature information by performing consistency detection and feature fusion based on the first lesion detection result, the second lesion detection result, the third lesion detection result, the first weight, the second weight, the third weight, and the clinical information; and

a structured report generation module, configured to:

determine the diagnostic result based on the fused feature information;

determine the traceability basis based on the initial traceability basis; and

generate the structured diagnostic report based on the consistency detection result, the diagnostic result, and the traceability basis.

12 . The method of claim 11 , wherein the structured diagnostic report for the target object further includes a treatment evaluation result, and the report generation model further includes a treatment prediction module,

the treatment prediction module is configured to:

determine one or more reference treatment plans for the lesion based on the first lesion detection result, the second lesion detection result, the third lesion detection result, the clinical information, and the diagnostic result; and

generate the treatment evaluation result based on the reconstructed image, the first lesion detection result, the second lesion detection result, the third lesion detection result, the clinical information, and the one or more reference treatment plans;

the structured report generation module is further configured to generate the structured diagnostic report based on the treatment evaluation result, the consistency detection result, the diagnostic result, and the traceability basis.

13 . The method of claim 12 , wherein the treatment evaluation result includes, for each reference treatment plan of the one or more reference treatment plans, at least one of an efficacy prediction indicator, a predicted medical image of the lesion, a side effect assessment, and a recommendation index.

14 . The method of claim 12 , wherein determining one or more reference treatment plans for the lesion based on the first lesion detection result, the second lesion detection result, the third lesion detection result, the clinical information, and the diagnostic result includes:

constructing a target optimization problem based on the first lesion detection result, the second lesion detection result, the third lesion detection result, the clinical information, and the diagnostic result, the target optimization problem including a first optimization objective, a second optimization objective, a third optimization objective, and a fourth optimization objective, wherein the first optimization objective is quantified by a first objective function for characterizing efficacy, the second optimization objective is quantified by a second objective function for characterizing safety, the third optimization objective is quantified by a third objective function for characterizing treatment cost, and the fourth optimization objective is quantified by a fourth objective function for characterizing compliance with treatment guidelines; and

generating the one or more reference treatment plans by solving, under a medical constraint, the target optimization problem using an optimization algorithm based on the first optimization objective, the second optimization objective, the third optimization objective, and the fourth optimization objective.

15 . The method of claim 14 , wherein the generating the one or more reference treatment plans by solving, under a medical constraint, the target optimization problem using an optimization algorithm based on the first optimization objective, the second optimization objective, the third optimization objective, and the fourth optimization objective includes:

obtaining a value preference of the target object;

for each optimization objective of the first optimization objective, the second optimization objective, the third optimization objective, and the fourth optimization objective, determining an optimization weight corresponding to the optimization objective based on the value preference;

generating the one or more reference treatment plans by solving, under the medical constraint, the target optimization problem using the optimization algorithm based on the first optimization objective, the second optimization objective, the third optimization objective, the fourth optimization objective, and the optimization weight corresponding to each optimization objective.

16 . A system for generating a diagnostic report, comprising:

at least one storage device configured to store computer instructions; and

at least one processor in communication with the at least one storage device, wherein when executing the computer instructions, the at least one processor is configured to cause the system to perform operations including:

obtain raw scanning data of a target object, a reconstructed image of the target object, and clinical information of the target object, wherein the raw scanning data is obtained by scanning the target object using a medical imaging device, and the reconstructed image is generated based on the raw scanning data;

determine, based on the raw scanning data, first lesion information of the target object;

determine, based on the reconstructed image and the first lesion information, second lesion information of the target object;

determine whether the target object includes a lesion based on the first lesion information and the second lesion information;

in response to determining that the target object includes the lesion, generate, based on the raw scanning data and the reconstructed image, a target reconstructed image of the lesion using a simulation scanning model;

obtain third lesion information by performing feature extraction on the target reconstructed image; and

generate a structured diagnostic report for the target object using a report generation model based on the first lesion information, the second lesion information, the third lesion information, and the clinical information.

17 . The system of claim 16 , wherein

the first lesion information is determined by processing the raw scanning data using a raw data analysis model;

the second lesion information is determined by processing the reconstructed image and the first lesion information using an image analysis model; and

the third lesion information is determined by processing the target reconstructed image using a lesion analysis model, wherein the raw data analysis model, the image analysis model, and the lesion analysis model are trained machine learning models.

18 . The system of claim 17 , wherein the raw data analysis model and the image analysis model are obtained by:

obtaining sample raw data, a sample reconstructed image corresponding to the sample raw data, first label lesion information corresponding to the sample reconstructed image, and second label lesion information corresponding to the sample reconstructed image;

obtaining the raw data analysis model by training a first preliminary model using the sample raw data and the first label lesion information;

determining sample first lesion information corresponding to the sample raw data based on the sample raw data and the raw data analysis model; and

obtaining the image analysis model by training a second preliminary model using the sample reconstructed image, the sample first lesion information, and the second label lesion information.

19 . The system of claim 16 , wherein,

in response to determining that historical diagnostic data related to the lesion exists, the at least one processor is further configured to:

determine lesion change information based on the historical diagnostic data, the first lesion information, the second lesion information, and the third lesion information; and

the generating a structured diagnostic report for the target object using a report generation model based on the first lesion information, the second lesion information, the third lesion information, and the clinical information includes:

generating the structured diagnostic report for the target object by processing the first lesion information, the second lesion information, the third lesion information, the clinical information, and the lesion change information using the report generation model.

20 . A non-transitory computer-readable storage medium, wherein the storage medium stores at least one set of instructions, and when one or more processors of a computing device execute the at least one set of instructions, the at least one set of instructions causes the computing device to perform a method, the method comprising:

obtaining raw scanning data of a target object, a reconstructed image of the target object, and clinical information of the target object, wherein the raw scanning data is obtained by scanning the target object using a medical imaging device, and the reconstructed image is generated based on the raw scanning data;

determining, based on the raw scanning data, first lesion information of the target object;

determining, based on the reconstructed image and the first lesion information, second lesion information of the target object;

determining whether the target object includes a lesion based on the first lesion information and the second lesion information;

in response to determining that the target object includes the lesion, generating, based on the raw scanning data and the reconstructed image, a target reconstructed image of the lesion using a simulation scanning model;

obtaining third lesion information by performing feature extraction on the target reconstructed image; and

generating a structured diagnostic report for the target object using a report generation model based on the first lesion information, the second lesion information, the third lesion information, and the clinical information.