Scheltens MTA Scheltens medial temporal atrophy visual rating scale

{
  "categories": [
    {
      "outcome_code": "0",
      "choroid_fissure": "normal_width",
      "temporal_horn": "normal_width",
      "hippocampal_height": "normal"
    },
    {
      "outcome_code": "1",
      "choroid_fissure": "mildly_widened",
      "temporal_horn": "normal_width",
      "hippocampal_height": "normal"
    },
    {
      "outcome_code": "2",
      "choroid_fissure": "moderately_widened",
      "temporal_horn": "mildly_widened",
      "hippocampal_height": "mildly_reduced"
    },
    {
      "outcome_code": "3",
      "choroid_fissure": "markedly_widened",
      "temporal_horn": "moderately_widened",
      "hippocampal_height": "moderately_reduced"
    },
    {
      "outcome_code": "4",
      "choroid_fissure": "markedly_widened",
      "temporal_horn": "markedly_widened",
      "hippocampal_height": "severely_reduced_end_stage_pattern"
    }
  ],
  "applicability": {
    "use_for": "Visual description of medial temporal lobe atrophy during a cognitive-impairment or dementia imaging assessment.",
    "classification_unit": "one_medial_temporal_lobe_per_side",
    "required_inputs": [
      "patient_age",
      "right_coronal_MTA_score",
      "left_coronal_MTA_score",
      "sequence_and_plane",
      "image_adequacy",
      "named_interpretation_protocol_if_a_cutoff_is_used"
    ],
    "outside_scope": [
      "standalone_Alzheimer_disease_diagnosis",
      "biological_AD_staging",
      "treatment_eligibility",
      "cognitive_or_functional_severity_staging",
      "etiology_from_atrophy_alone"
    ]
  },
  "technique_and_rating_gate": {
    "original_basis": "Rate each side on a coronal T1-weighted image oriented parallel to the brainstem axis at the level through the cerebral aqueduct, assessing all three anatomic axes together.",
    "sequence_rule": "Record the sequence, reconstruction and plane. Axial ratings, CT ratings and automated volumetric outputs are adaptations or separate methods and must not be presented as the original coronal MRI score without naming the protocol.",
    "side_rule": "Assign right and left scores independently; asymmetry is clinically meaningful and must not be discarded.",
    "inadequacy_rule": "If head rotation, obliquity, motion, incomplete hippocampal coverage or a noncomparable plane prevents confident scoring, return side-specific ungradable or a bounded range rather than inventing a number."
  },
  "bilateral_output_contract": {
    "required_report": "Return MTA right and MTA left separately, plus technique and any relevant asymmetry.",
    "aggregation_guard": "Do not silently average, sum or select the worst side. If a study or local protocol uses a bilateral mean, maximum or another aggregation, name that operation before applying its cutoff.",
    "longitudinal_guard": "Compare serial scores only when acquisition, plane and rating protocol are sufficiently comparable; a one-point change can reflect technique or reader variation."
  },
  "age_and_cutoff_protocols": {
    "no_universal_threshold": "There is no universal age-independent abnormal cutoff. Thresholds differ by age band, modality, cohort and side-combination protocol, including whether right and left scores are averaged or otherwise combined.",
    "Claus_2017_CT_bilateral_mean": "In one large memory-clinic CT validation using the mean of right and left ratings, proposed abnormal thresholds were at least 1.0 below age 65, at least 1.5 at 65-74, and at least 2.0 at 75-84 and 85 or older; usefulness was limited in the oldest group.",
    "Molinder_2021_MRI_cohort": "A heterogeneous MRI memory-clinic cohort proposed thresholds of 1 below age 75 and 1.5 at age 75 or older within its own methods; these must not be merged with the Claus CT protocol.",
    "normative_warning": "Normative data show increasing scores with age, including high scores in some cognitively healthy adults over 80, so even grade 3 or 4 is not automatically Alzheimer disease."
  },
  "diagnostic_and_etiologic_boundary": {
    "structural_marker": "MTA is a structural neurodegeneration marker and can support a pattern-based differential diagnosis when integrated with symptoms, cognition, other regional atrophy and vascular or structural findings.",
    "not_AD_specific": "Medial temporal atrophy occurs in Alzheimer phenotypes but also in vascular or mixed dementia, hippocampal sclerosis, LATE and other conditions; hippocampal-sparing Alzheimer phenotypes may have little MTA.",
    "biological_AD_guard": "The 2024 Alzheimer's Association criteria define Alzheimer disease biologically using validated core biomarkers. A Scheltens score does not establish amyloid or tau pathology and cannot substitute for a required biomarker.",
    "negative_guard": "A low score does not exclude early, young-onset or hippocampal-sparing Alzheimer disease and does not exclude another cause of cognitive impairment."
  },
  "performance_and_reliability": {
    "meta_analysis": "A 2021 meta-analysis for Alzheimer disease versus healthy controls reported pooled sensitivity about 74 percent and specificity about 88 percent, with heterogeneous methods; these are group-level diagnostic-performance estimates, not a probability attached to an individual grade.",
    "heterogeneous_clinic_validation": "In 752 MRI examinations, correlations with quantitative hippocampal measures ranged from -0.20 to -0.68 and agreement was moderate to substantial; the cohort included Alzheimer, vascular and mixed dementias.",
    "normative_reliability": "A normative MRI study reported 90 percent agreement and an absolute-agreement ICC of 0.86, but reader agreement does not make the marker disease-specific."
  },
  "management_boundary": {
    "no_grade_action": "No grade independently starts, withholds or selects medication, disease-modifying therapy, biomarker testing, driving advice or level of care.",
    "appropriate_next_step": "When clinically relevant, integrate the bilateral score with cognitive and functional assessment, the complete MRI pattern, vascular burden, laboratory evaluation and disease-specific biomarkers according to the actual diagnostic question.",
    "anti_automation_guard": "Never transform an MTA cutoff into a diagnosis, prognosis, treatment eligibility decision or autonomous referral without the named protocol and patient context."
  },
  "missing_input_behavior": [
    "If age is unknown, provide morphology and bilateral scores but do not label them normal or abnormal by an age cutoff.",
    "If only one side is scorable, report that side and explicitly mark the other side ungradable.",
    "If the bilateral aggregation method is unknown, do not apply a published mean-score threshold.",
    "If cognition, function or biomarker status is unavailable, do not infer Alzheimer disease or its clinical stage.",
    "If the three anatomic axes disagree near a boundary, state the competing grades and the feature causing uncertainty."
  ],
  "supporting_sources": [
    {
      "role": "primary_scale",
      "citation": "Scheltens et al. JNNP. 1992;55:967-972",
      "doi": "10.1136/jnnp.55.10.967",
      "pmid": "1431963"
    },
    {
      "role": "age_specific_cutoff_protocol",
      "citation": "Claus et al. European Radiology. 2017;27:3147-3155",
      "doi": "10.1007/s00330-016-4726-3",
      "pmid": "28083697"
    },
    {
      "role": "heterogeneous_clinic_validity",
      "citation": "Molinder et al. BMC Neurology. 2021;21:289",
      "doi": "10.1186/s12883-021-02325-2",
      "pmcid": "PMC8305846"
    },
    {
      "role": "diagnostic_meta_analysis",
      "citation": "Park et al. European Radiology. 2021;31:9060-9072",
      "doi": "10.1007/s00330-021-08227-8"
    },
    {
      "role": "biological_AD_boundary",
      "citation": "Jack et al. Alzheimer's & Dementia. 2024",
      "doi": "10.1002/alz.13859",
      "pmid": "38934362"
    }
  ],
  "source_locator": "Scheltens et al. 1992, PMID 1431963, original bilateral coronal MRI visual scale; Claus et al. 2017, PMC5491609, decade-specific CT mean-score thresholds and oldest-group limitation; Cotta Ramusino et al. 2019, PMC6690662, coronal technique and normative distributions; Molinder et al. 2021, PMC8305846, heterogeneous-clinic validity and reliability; Park et al. 2021, DOI 10.1007/s00330-021-08227-8, pooled diagnostic performance; Jack et al. 2024, DOI 10.1002/alz.13859, biological AD criteria."
}