diff --git a/src/pyrecest/filters/abstract_filter.py b/src/pyrecest/filters/abstract_filter.py
index 0a049bd20..96eb05989 100644
--- a/src/pyrecest/filters/abstract_filter.py
+++ b/src/pyrecest/filters/abstract_filter.py
@@ -3,6 +3,7 @@
 import copy
 from abc import ABC, abstractmethod
 
+from pyrecest.backend import atleast_1d
 from pyrecest.utils.history_recorder import HistoryRecorder
 
 
@@ -33,8 +34,8 @@ def filter_state(self, new_state):
         self._filter_state = copy.deepcopy(new_state)
 
     def get_point_estimate(self):
-        """Get a point estimate"""
-        return self.filter_state.mean()
+        """Get a point estimate as a one-dimensional state vector."""
+        return atleast_1d(self.filter_state.mean())
 
     @property
     def dim(self) -> int:
diff --git a/tests/filters/test_track_manager_atomic_filter_state.py b/tests/filters/test_track_manager_atomic_filter_state.py
index dc4efecfc..235459975 100644
--- a/tests/filters/test_track_manager_atomic_filter_state.py
+++ b/tests/filters/test_track_manager_atomic_filter_state.py
@@ -3,7 +3,9 @@
 import unittest
 
 import numpy as np
+from pyrecest.distributions import VonMisesDistribution
 from pyrecest.filters.track_manager import TrackManager
+from pyrecest.filters.von_mises_filter import VonMisesFilter
 
 
 class TrackManagerAtomicFilterStateTest(unittest.TestCase):
@@ -32,3 +34,19 @@ def test_invalid_replacement_preserves_existing_track_bank(self):
             manager.tracks[0].get_point_estimate(),
             np.array([1.0]),
         )
+
+    def test_scalar_filter_point_estimate_stacks_as_one_dimensional_state(self):
+        circular_filter = VonMisesFilter()
+        circular_filter.filter_state = VonMisesDistribution(0.25, 2.0)
+        manager = TrackManager(
+            extract_confirmed_only=False,
+            keep_history=False,
+            log_prior_estimates=False,
+            log_posterior_estimates=False,
+        )
+        manager.initialize_from_states([circular_filter], confirmed=True)
+
+        point_estimate = manager.get_point_estimate()
+
+        self.assertEqual(point_estimate.shape, (1, 1))
+        np.testing.assert_allclose(point_estimate, np.array([[0.25]]))