Return two side-specific grades plus the decisive planes. Never collapse them silently, interpret them without age, or convert a posterior atrophy pattern into Alzheimer diagnosis, individual prognosis or treatment.
{
"categories": [
{
"outcome_code": "0",
"sulci": "closed_posterior_cingulate_and_parieto_occipital_sulci",
"parietal_and_precuneus": "closed_sulci_without_atrophy"
},
{
"outcome_code": "1",
"sulci": "mild_widening",
"parietal_and_precuneus": "mild_atrophy_without_evident_gyral_volume_loss"
},
{
"outcome_code": "2",
"sulci": "substantial_widening",
"parietal_and_precuneus": "substantial_atrophy_with_gyral_volume_loss"
},
{
"outcome_code": "3",
"sulci": "evident_or_severe_widening",
"parietal_and_precuneus": "end_stage_knife_blade_atrophy"
}
],
"applicability": {
"use_for": "Visual description of posterior cortical atrophy on structural brain MRI in a cognitive or dementia imaging assessment.",
"classification_unit": "each_cerebral_hemisphere_separately",
"required_inputs": [
"right_and_left_sides",
"sagittal_T1",
"axial_FLAIR_or_equivalent_structural_sequence",
"coronal_T1",
"posterior_cingulate_sulcus",
"parieto_occipital_sulcus",
"precuneus",
"parietal_lobe_sulci_and_gyri",
"patient_age"
],
"useful_companion_inputs": [
"cognitive_phenotype",
"medial_temporal_atrophy",
"global_cortical_atrophy",
"white_matter_hyperintensity_burden",
"focal_lesions",
"acquisition_quality",
"biomarker_results_if_clinically_obtained"
],
"outside_scope": [
"standalone_Alzheimer_diagnosis",
"standalone_treatment_selection",
"severity_staging_of_clinical_dementia",
"automatic_conversion_between_MRI_and_CT",
"screening_asymptomatic_people_for_Alzheimer_disease"
]
},
"original_rating_architecture": {
"scale": "four_point_ordinal_0_to_3",
"laterality": "rate_left_and_right_separately",
"per_side_rule": "For each hemisphere, inspect all three prescribed planes and use the highest or worst supported plane score as that side's final score.",
"original_research_aggregation": "The development study averaged left and right final scores for statistical analyses. That mean is a study variable, not a mandatory clinical output and must never replace the two side scores silently.",
"normative_research_aggregation": "The 2019 normative study used the higher hemisphere score. This is a different named research aggregation and is not interchangeable with the original bilateral mean.",
"aggregation_guard": "Default API output is right and left. If a caller requests mean, maximum or another aggregate, name the method, return both raw side scores and never label the aggregate as the original Koedam grade."
},
"plane_specific_landmarks": {
"sagittal_T1": [
"posterior_cingulate_sulcus_widening",
"parieto_occipital_sulcus_widening",
"precuneus_atrophy_on_left_and_right_paramedian_images"
],
"axial_FLAIR": [
"posterior_cingulate_sulcus_widening",
"parietal_sulcal_dilation_and_gyral_volume"
],
"coronal_T1": [
"posterior_cingulate_sulcus_widening",
"parietal_lobe_sulci_and_gyral_volume"
],
"whole_pattern_rule": "Use the combined sulcal and gyral pattern across contiguous images. A single enlarged sulcus, focal infarct, postoperative defect or asymmetric destructive lesion must not be converted automatically into a degenerative PA grade."
},
"classification_sequence": [
"confirm_MRI_protocol_quality_and_patient_age",
"exclude_or_separately_label_focal_structural_lesions_that_distort_the_landmarks",
"score_right_sagittal_axial_and_coronal_views",
"score_left_sagittal_axial_and_coronal_views",
"select_the_highest_supported_plane_score_for_each_side",
"retain_right_left_asymmetry_and_the_decisive_plane",
"apply_age_aware_normative_context_without_changing_the_raw_grade",
"integrate_with_other_atrophy_patterns_and_clinical_or_biomarker_context"
],
"decisive_boundaries": {
"grade_0_1": "Any reproducible mild opening of the target sulci with mild posterior volume loss supports grade 1; fully closed target sulci without posterior atrophy support grade 0.",
"grade_1_2": "Grade 1 has no evident gyral volume loss. Substantial sulcal widening plus visible loss of parietal or precuneus gyral volume is grade 2.",
"grade_2_3": "Grade 3 requires end-stage or knife-blade posterior gyral atrophy, not merely substantial grade-2 volume loss.",
"plane_disagreement": "A higher valid score in one prescribed plane determines that side's final grade; record the decisive plane instead of averaging planes.",
"asymmetry": "Do not average away a focal high score. Report both hemispheres and investigate marked asymmetry or a focal structural explanation."
},
"reliability_and_derivation": {
"development_sample": "The original study included 118 relatively young participants: 60 with Alzheimer disease, 20 with other dementias and 38 controls, without neuropathologic confirmation.",
"intra_rater": "Weighted kappa was 0.93 and 0.95 for the two readers.",
"inter_rater": "Pairwise weighted kappas were 0.84, 0.70 and 0.65; mean inter-rater kappa was 0.73.",
"group_findings": "Mean bilateral PA was 1.6 plus or minus 0.9 in the Alzheimer group, 0.6 plus or minus 0.7 in controls and 0.8 plus or minus 0.8 in other dementias; PA also related independently to lower MMSE in that selected sample.",
"threshold_guard": "A bilateral mean above 1 produced 58 percent sensitivity and 95 percent specificity in the derivation study. This was a study cutoff, not a universal diagnostic threshold or per-patient probability.",
"generalizability_guard": "Selected diagnostic groups, relatively young age and absent pathologic confirmation limit transportability. Do not convert the reported group means, kappas or threshold into certainty for a new patient."
},
"age_aware_normative_context": {
"reference_sample": "A 2019 study rated 936 cognitively intact adults aged 20-84 years and used the higher hemisphere for its normative summary.",
"common_low_grade": "PA grade 1 occurred in 38 percent of cognitively intact participants; a low positive grade is therefore not automatically pathologic.",
"ninetieth_percentile": [
{
"age_range": "20_to_59_years",
"highest_hemisphere_grade_at_or_below": "1"
},
{
"age_range": "60_to_84_years",
"highest_hemisphere_grade_at_or_below": "2"
}
],
"interpretation_guard": "These values describe a cognitively intact reference sample and compatibility with normal cognition, not proof of healthy brain tissue. Age modifies interpretation but never rewrites the observed grade."
},
"pattern_and_differential_boundary": {
"Alzheimer_pattern": "Posterior-predominant atrophy can support an Alzheimer-type imaging pattern, including atypical or younger-onset presentations, but is neither specific nor required for Alzheimer disease.",
"focal_mimics": [
"prior_infarction",
"encephalomalacia",
"trauma_or_surgery",
"developmental_asymmetry",
"mass_effect",
"hydrocephalus_or_CSf_dynamics",
"other_neurodegenerative_patterns"
],
"complementary_scales": "Report medial temporal atrophy, global cortical atrophy and Fazekas white-matter burden as separate axes. Their numbers are not interchangeable with Koedam and must not be summed into an invented dementia score.",
"CT_boundary": "The original scale and plane rules were developed on MRI. If applied to CT, label it as an adaptation and do not claim original validation or equivalence."
},
"current_diagnosis_and_care_boundary": {
"biological_AD_rule": "Current Alzheimer criteria define and stage disease biologically through core biomarkers. A Koedam grade cannot establish amyloid or tau pathology, determine biological stage or substitute for a complete clinical assessment.",
"symptomatic_workup": "When cognitive symptoms and an age-discordant or characteristic posterior pattern are present, the result can support specialist cognitive evaluation and appropriately selected biomarker work-up; the grade alone does not specify which test.",
"asymptomatic_guard": "Do not use Koedam to screen or label an asymptomatic person with Alzheimer disease, and do not infer future dementia from a grade alone.",
"treatment_guard": "No grade independently starts, stops or selects symptomatic therapy, disease-modifying therapy, counseling, driving restrictions or surveillance. Eligibility depends on clinical syndrome, confirmed biology, contraindications and current specialist guidance."
},
"agent_output_contract": [
"Koedam_2011_scale_identity",
"MRI_sequences_and_technical_adequacy",
"patient_age",
"right_sagittal_axial_coronal_scores_and_decisive_plane",
"left_sagittal_axial_coronal_scores_and_decisive_plane",
"right_and_left_final_worst_plane_scores",
"named_aggregation_method_only_if_requested",
"posterior_cingulate_parieto_occipital_precuneus_and_parietal_findings",
"asymmetry_and_focal_confounders",
"age_aware_normative_context",
"complementary_MTA_GCA_and_Fazekas_context",
"non_diagnostic_non_treatment_warning",
"uncertainty_and_missing_inputs"
],
"missing_input_behavior": [
"If one prescribed plane is absent, report the available per-plane scores and a provisional per-side maximum; do not claim the complete original protocol.",
"If laterality is not preserved, do not invent a bilateral mean or maximum; return laterality unresolved.",
"If age is unknown, return the morphology score but withhold age-normative interpretation.",
"If a focal lesion distorts one landmark, mark that region unassessable and report the lesion separately rather than treating it as diffuse degenerative atrophy.",
"If the caller supplies one unlabeled 0-3 value, ask whether it is right, left, a side maximum, a bilateral mean or a bilateral maximum before grounding downstream reasoning."
],
"supporting_sources": [
{
"role": "original_scale_and_validation",
"citation": "Koedam et al. Eur Radiol. 2011;21:2618-2625",
"doi": "10.1007/s00330-011-2205-4",
"pmcid": "PMC3217148"
},
{
"role": "age_normative_context",
"citation": "Cotta Ramusino et al. NeuroImage Clinical. 2019;24:101936",
"doi": "10.1016/j.nicl.2019.101936",
"pmcid": "PMC6690662"
},
{
"role": "biological_Alzheimer_boundary",
"citation": "Jack et al. Alzheimer's & Dementia. 2024",
"doi": "10.1002/alz.13859",
"pmid": "38934362"
}
],
"source_locator": "Koedam et al. Eur Radiol 2011, DOI 10.1007/s00330-011-2205-4, PMC3217148, Methods for planes, landmarks, grades and worst-plane rule and Results for reliability and derivation performance; Cotta Ramusino et al. 2019, DOI 10.1016/j.nicl.2019.101936, PMC6690662, PA age distributions and 90th percentiles; Jack et al. 2024, DOI 10.1002/alz.13859, biological diagnosis boundary."
}