Fix timeout issues when querying instances of a local study in Orthanc

pixl_imaging/src/pixl_imaging/_orthanc.py

@@ -64,9 +64,16 @@ async def get_local_study(self, study_id: str) -> Any:
         """Query local Orthanc instance for study."""
         return await self._get(f"/studies/{study_id}")
 
+    async def get_local_study_statistics(self, study_id: str) -> Any:
+        """Query local Orthanc instance for study statistics."""
+        return await self._get(f"/studies/{study_id}/statistics")
+
     async def get_local_study_instances(self, study_id: str) -> Any:
         """Get the instances of a study."""
-        return await self._get(f"/studies/{study_id}/instances")
+        return await self._get(
+            f"/studies/{study_id}/instances?short=true",
+            timeout=self.dicom_timeout,  # this API call can sometimes take several minutes
+        )
 
     async def query_remote(self, data: dict, modality: str) -> Optional[str]:
         """Query a particular modality, available from this node"""
@@ -154,13 +161,15 @@ async def job_state(self, job_id: str) -> Any:
         # See: https://book.orthanc-server.com/users/advanced-rest.html#jobs-monitoring
         return await self._get(f"/jobs/{job_id}")
 
-    async def _get(self, path: str) -> Any:
+    async def _get(self, path: str, timeout: int | None = None) -> Any:
+        # Optionally override default http timeout
+        http_timeout = timeout or self.http_timeout
         async with (
             aiohttp.ClientSession() as session,
             session.get(
                 f"{self._url}{path}",
                 auth=self._auth,
-                timeout=self.http_timeout,
+                timeout=http_timeout,
             ) as response,
         ):
             return await _deserialise(response)

pixl_imaging/src/pixl_imaging/_processing.py

@@ -283,15 +283,9 @@ async def _get_missing_instances(
 
     Return a list of missing instance UIDs (empty if none missing)
     """
-    # First get all SOPInstanceUIDs for the study that are in Orthanc Raw
-    orthanc_raw_sop_instance_uids = []
-    for resource in resources:
-        study_instances = await orthanc_raw.get_local_study_instances(study_id=resource)
-        orthanc_raw_sop_instance_uids.extend(
-            [instance["MainDicomTags"]["SOPInstanceUID"] for instance in study_instances]
-        )
+    missing_instances: list[dict[str, str]] = []
 
-    # Now query the VNA / PACS for the study instances
+    # First query the VNA / PACS for the study instances
     study_query_answers = await orthanc_raw.get_remote_query_answers(study_query_id)
     instances_queries_and_answers = []
     for answer_id in study_query_answers:
@@ -302,13 +296,25 @@ async def _get_missing_instances(
         instances_queries_and_answers.extend(
             [(instances_query_id, answer) for answer in instances_query_answers]
         )
+    num_remote_instances = len(instances_queries_and_answers)
 
-    missing_instances: list[dict[str, str]] = []
+    num_local_instances = 0
+    for resource in resources:
+        study_statistics = await orthanc_raw.get_local_study_statistics(study_id=resource)
+        num_local_instances += int(study_statistics["CountInstances"])
 
-    if len(instances_queries_and_answers) == len(orthanc_raw_sop_instance_uids):
+    if num_remote_instances == num_local_instances:
         logger.debug("No missing instances for study {}", study.message.study_uid)
         return missing_instances
 
+    # Get all SOPInstanceUIDs for the study that are in Orthanc Raw
+    orthanc_raw_sop_instance_uids = []
+    for resource in resources:
+        study_instances = await orthanc_raw.get_local_study_instances(study_id=resource)
+        orthanc_raw_sop_instance_uids.extend(
+            [instance["MainDicomTags"]["0008,0018"] for instance in study_instances]
+        )
+
     # If the SOPInstanceUID is not in the list of instances in Orthanc Raw
     # retrieve the instance from the VNA / PACS
     query_tags = ["0020,000d", "0020,000e", "0008,0018"]