Radial Arm Maze

Overview

The radial arm maze (RAM) is an eight-arm apparatus used to assess both working memory and reference memory in a single testing session. In the standard protocol, a subset of arms is baited with food rewards, and the animal must visit all baited arms without revisiting any. Re-entries into already-visited baited arms constitute working memory errors, while entries into never-baited arms constitute reference memory errors — allowing simultaneous, independent measurement of two distinct memory systems.

This dual-error dissociation makes the RAM uniquely informative: working memory errors reflect trial-specific, rapidly decaying information held online (prefrontal cortex and hippocampus-dependent), while reference memory errors reflect failure to learn the fixed rule about which arms are baited (hippocampus and basal forebrain-dependent). Pharmacological and lesion studies have shown that cholinergic blockade impairs both error types, while hippocampal lesions preferentially impair working memory.

ConductMaze automates the radial arm maze through sensors at each arm entrance that detect entries and exits, motorized guillotine doors for controlled arm access, and automated reward dispensers at arm ends. The software scores working and reference memory errors in real time, manages arm-baiting patterns, and supports configurable delay intervals for delayed working memory variants.

Trial Flow

start

Bait Arms

Place rewards in designated subset of arms (e.g., 4 of 8)

process

Place Subject

Place animal on center platform, doors closed

process

Doors Open

All doors open simultaneously; free-choice phase

output

Arm Entry

Sensor logs arm chosen, timestamp, reward obtained

decision

Error Classification

Working error (revisit) or reference error (unbaited arm)

decision

All Baits Collected

Subject found all rewards, or max time reached

end

Session End

Record total errors, choices to criterion, time

Parameters

ParameterTypeDefaultDescription
Number of Armsinteger8Total arms in maze (typically 8)
Baited Armsinteger4Number of arms containing food rewards
Max Trial Durationseconds600Maximum session time
Arm Lengthcm50Length of each arm from center to reward location
Food Deprivationenum85% BWFood restriction level to motivate performance
Delay Intervalseconds0Delay between first 4 choices and remaining choices (0 = no delay)
Training Sessionsinteger15Number of sessions to criterion performance

Metrics

MetricUnitDescription
Working Memory ErrorscountRe-entries into already-visited baited arms — trial-specific memory
Reference Memory ErrorscountEntries into never-baited arms — rule memory
Choices to CriterioncountTotal arm entries to collect all baits — overall efficiency
Time to CompletionsecondsTotal time to collect all available rewards
First-Choice Accuracy%Percentage of first 4 choices that are correct
Perseverative ReturnscountConsecutive re-entries to same arm — compulsive behavior index
Choice Pattern EntropybitsRandomness of arm selection — spatial strategy measure

Sample Data

SubjectGroupSessionWM_ErrorsRM_ErrorsChoicesTime_s

Representative data for illustration purposes. Actual values will vary by species, strain, and experimental conditions.

Applications

  • 1
    Cholinergic pharmacologyscopolamine-induced memory impairment and reversal by cholinesterase inhibitors
  • 2
    Hippocampal lesion studiesdissociating working from reference memory circuits
  • 3
    Aging researchage-related working memory decline and compensatory strategies
  • 4
    Alzheimer modelsamyloid and tau effects on dual memory systems
  • 5
    Nutritional neurosciencediet and supplement effects on spatial working memory

Ready to Automate Your Behavioral Protocols?

Contact us for a demo and pricing information.