Medical image information processing method and device, electronic equipment and storage medium

Abstract

The invention provides a medical image information processing method and device, electronic equipment and a storage medium, and relates to the technical field of computers. The processing method comprises the steps: performing field division on first medical image information used for training, and conducting multi-granularity labeling on fields according to a preset medical label, the medical label comprising a first label formed by combining a disease site and a case type; shielding the marked fields, and training an identification model according to a training set formed by the marked fields and the derivative fields; acquiring to-be-identified second medical image information, and performing classification identification on fields in the second medical image information through the trained identification model; and combining the labels belonging to the same key field after classification and identification, and outputting the key field corresponding to the combined label in a key value form. According to the technical scheme, the efficiency and reliability of extracting the key information of the medical image information are improved.

Description

technical field [0001] The present disclosure relates to the technical field of information processing, and in particular to a processing method, device, electronic equipment, and storage medium for medical image information. Background technique [0002] With the upgrading of Internet+, information technology and supporting hardware measures, the data in the fields of medical health and insurance have also shown a blowout and explosive growth. Features: large data volume, complex data types, numerous data forms, low data value density, etc. However, as the basic people's livelihood industry for the development of the country and society, medical health contains huge value in medical data. On the one hand, quickly extracting key information from medical and health data is conducive to rapid diagnosis for doctors or medical institutions. On the other hand, It is conducive to statistical analysis of cases to determine the cause of disease or transmission path. On the other ha...

AI Technical Summary

Problems solved by technology

[0004] 1. All text descriptions in the tumor imaging report revolve around the tumor, and there is a lot of redundant information. For example, the primary site and pathological type of the tumor may be described at the beginning, middle, and end of the tumor imaging report. Some descriptions The scope of the content overlaps, and the dimensions of some description content are inconsistent

[0005] 2. The tumor imaging report includes the primary site of the tumor, the size of the tumor, the site of tumor metastasis, the size of the site of tumor metastasis, the site of lymph node metastasis, the size, clinical stage, pathological stage, condition changes, edge / adjacent relationship, etc. Information that is closely related to tumor diagnosis, follow-up treatment, and prognosis. In addition, these information are long text information described subjectively by doctors. The position and length of each module information have no fixed format, so they cannot be directly applied. Extracted from tumor imaging reports

[0007] 4. The data structure of the tumor imaging report is complex, such as the text type "lung cancer", "mediastinum", the data value type "4cm", "2 pieces", the orientation word "7 o'clock", the sequence word "7th rib", etc. Wait

[0010] It can be seen that the tumor imaging report contains a lot of valuable medical information, and it is of great significance to extract and analyze this valuable information. Due to the high complexity of its data format, it is necessary to manually extract the key information in the tumor imaging report. Not only waste huge manpower cost and time cost, but also may have serious reliability problems due to the omission of manual operation

Images