算是語意辨認的一種 benchmark，用來評估模型是否能理解精確的放射科報告內容上下文。
Prompt:
Your task is to extract medical entities and relations from a given radiology report. I’ll provide you with 1) the problem setup, 2) the radiology report, and 3) the output format.
1) Problem setup: For each report, you will be asked to identify 7 types of medical entities:
(1) observation::present, which is used for visual features, pathophysiologic processes, or diagnosable diseases that are present;
(2) observation::absent, which is used for visual features, pathophysiologic processes, or diagnosable diseases that are absent;
(3) observation::uncertain, which is used for visual features, pathophysiologic processes, or diagnosable diseases where you are uncertain about presence or absence;
(4) observation::measurement::present, which refers to a measurement associated with visual features, pathophysiologic processes, or diseases;
(5) anatomy::present, which refers to an anatomical body part that is present;
(6) anatomy::absent, which refers to an anatomical body part that is absent;
(7) anatomy::measurement::present, which refers to a measurement associated with an anatomical body part;
For each report, you will also be asked to identify 3 types of relations between entities:
(1) suggestive_of, which is a relation between two Observation entities indicating that the presence of the second Observation is inferred from the first Observation.
(2) located_at, which is a relation between an Observation entity and an Anatomy entity indicating that the Observation is related to the Anatomy
(3) modify, which is a relation between two Observation entities or two Anatomy entities indicating that the first entity modifies the scope of or quantifies the degree of the second entity.
2) Radiology report:
<Report>
3) Output format:
Please strictly follow this output format. Entities must be short substrings (often just 1 word) from the radiology report with no changes to formatting. Each relation exists between a pair of identified entities. Please list entities and relations in the order they appear in the radiology report.
[Entities]:
[[<entity>, <entity type>], [<entity>, <entity type>], ..., [<entity>, <entity type>]]
[Relations]:
[[<entity 1>, <entity 2>, <relation type>], [<entity 1>, <entity 2>, <relation type>], ..., [<entity 1>, <entity 2>, <relation type>]]