@brainpilot/skills 0.0.6

package/skills/10_Clinical_Neuropsychology/lesion-symptom-mapping-guide/references/vlsm-pipeline.md

@@ -0,0 +1,182 @@
+# VLSM Pipeline: Step-by-Step Workflow
+
+This reference document supplements `SKILL.md` with a detailed step-by-step workflow for voxel-based lesion-symptom mapping.
+
+## Pre-Analysis Checklist
+
+Before starting VLSM, verify:
+
+- [ ] N >= 50 patients (N >= 100 preferred; Sperber, 2020)
+- [ ] All patients have T1 (and ideally FLAIR) images
+- [ ] Behavioral data available for all patients (continuous preferred)
+- [ ] Time post-onset is > 3 months for all (or most) patients
+- [ ] Etiology is consistent (e.g., all ischemic stroke) or will be modeled
+
+## Step 1: Lesion Segmentation
+
+### Manual Tracing Protocol
+
+1. **Imaging modality**: Use T1 for chronic lesions (hypointense); FLAIR for acute/subacute lesions (hyperintense) or white matter lesions
+2. **Software**: MRIcron, ITK-SNAP, or 3D Slicer
+3. **Tracing procedure**:
+ - Trace on each axial slice (or coronal if preferred)
+ - Include all clearly damaged tissue
+ - Exclude peri-lesional edema if visible
+ - Include both cortical and subcortical damage
+4. **Inter-rater reliability**: If multiple raters, compute Dice coefficient; target >= 0.85 (Pustina et al., 2016)
+5. **Save as binary mask** (0 = healthy, 1 = lesioned) in the same space as the T1
+
+### Semi-Automated Segmentation
+
+If using LINDA or lesion_gnb:
+1. Run the automated pipeline
+2. **CRITICAL**: Visually inspect EVERY patient's segmentation on all slices
+3. Manually correct errors (over-segmentation of ventricles, missed lesion regions)
+4. Document the correction process
+
+## Step 2: Spatial Normalization
+
+### Using ANTs (recommended)
+
+```bash
+# Create lesion cost function mask (inverted lesion)
+fslmaths lesion_mask.nii.gz -binv lesion_costmask.nii.gz
+
+# Register T1 to MNI with cost function masking
+antsRegistrationSyNQuick.sh \
+ -d 3 \
+ -f MNI152_T1_1mm.nii.gz \
+ -m patient_T1.nii.gz \
+ -x lesion_costmask.nii.gz \
+ -o patient_to_MNI_
+
+# Apply warp to lesion mask
+antsApplyTransforms \
+ -d 3 \
+ -i lesion_mask.nii.gz \
+ -r MNI152_T1_1mm.nii.gz \
+ -o lesion_mask_MNI.nii.gz \
+ -t patient_to_MNI_1Warp.nii.gz \
+ -t patient_to_MNI_0GenericAffine.mat \
+ -n NearestNeighbor
+```
+
+### Using FSL
+
+```bash
+# Register with lesion masking
+flirt -in patient_T1 -ref MNI152_T1_2mm -omat affine.mat \
+ -inweight lesion_costmask -cost mutualinfo
+
+fnirt --in=patient_T1 --ref=MNI152_T1_2mm --aff=affine.mat \
+ --inmask=lesion_costmask --cout=warp_coeff
+
+applywarp --in=lesion_mask --ref=MNI152_T1_2mm \
+ --warp=warp_coeff --out=lesion_mask_MNI --interp=nn
+```
+
+### Quality Control
+
+For every patient:
+1. Overlay the warped T1 on the MNI template
+2. Check anatomical landmark alignment (ventricles, central sulcus, Sylvian fissure)
+3. Overlay the warped lesion mask on the template
+4. Verify the lesion location is anatomically plausible
+5. Flag and exclude patients with poor registration quality
+
+## Step 3: Create Lesion Overlap Map
+
+1. Sum all binarized lesion masks in MNI space
+2. Inspect the overlap map to understand lesion distribution
+3. Set minimum overlap threshold:
+ - Include voxels where >= **N_min patients** have lesions
+ - N_min = max(10, 10% of sample) (Kimberg et al., 2007)
+ - This creates the analysis mask
+
+## Step 4: Run VLSM Analysis
+
+### Using NiiStat (MATLAB)
+
+```matlab
+% Load lesion masks and behavioral data
+nii_stat(lesion_folder, behavior_file, ...
+ 'design', 'continuous', ... % or 'binomial'
+ 'stat', 'brunner-munzel', ... % or 'ttest'
+ 'perm', 5000, ... % permutation count
+ 'minOverlap', 10); % minimum lesion overlap
+```
+
+### Using LESYMAP (R)
+
+```r
+library(LESYMAP)
+lsm = lesymap(
+ lesions.list = lesion_files,
+ behavior = behavior_scores,
+ method = "BM", # Brunner-Munzel
+ multipleComparison = "FWE",
+ nperm = 5000,
+ minSubjectPerVoxel = 10
+)
+```
+
+### Controlling for Lesion Volume
+
+**Method 1: Direct covariate (recommended)**
+```r
+lsm = lesymap(
+ lesions.list = lesion_files,
+ behavior = behavior_scores,
+ method = "regress", # Regression-based
+ covariates = data.frame(lesion_vol = volumes),
+ multipleComparison = "FWE",
+ nperm = 5000
+)
+```
+
+**Method 2: Residualization**
+```r
+residuals = resid(lm(behavior ~ lesion_volume))
+# Use residuals as behavior in standard VLSM
+```
+
+## Step 5: Interpret Results
+
+### What the Map Shows
+
+- Each supra-threshold voxel indicates a statistical association between damage at that voxel and worse behavioral performance
+- The map does NOT directly prove causation; it shows association (confounded by vascular territory collinearity)
+- Report peak coordinates, cluster extent, and statistical values
+
+### What the Map Does NOT Show
+
+- It does not identify "the region responsible for X" (collinearity caveat)
+- It does not account for white matter disconnection effects
+- It does not reveal whether the deficit is due to damage at the voxel or at a distant connected region
+
+### Supplementary Analyses
+
+After VLSM, consider:
+1. **Disconnection analysis**: Run BCBToolkit to identify which tracts are disrupted
+2. **ROI validation**: Extract lesion load in a priori ROIs and test association with behavior
+3. **Lesion network mapping**: Identify the functional network disrupted by the lesion locations
+
+## Common Quality Control Issues
+
+| Issue | Detection | Solution |
+|---|---|---|
+| Poor registration | Visual inspection of overlay | Re-run with different parameters or enantiomorphic approach |
+| Lesion extends beyond brain boundary | Visual inspection | Mask lesion with brain mask |
+| Ventricle labeled as lesion | Compare to template ventricles | Manually correct segmentation |
+| Behavioral outliers | Box plots, Cook's distance | Investigate; consider robust statistics (Brunner-Munzel) |
+| Insufficient lesion overlap | Overlap map shows sparse coverage | Lower voxel threshold or use ROI approach |
+
+## References
+
+- Avants, B. B., et al. (2011). A reproducible evaluation of ANTs similarity metric performance in brain image registration. *NeuroImage*, 54(3), 2033--2044.
+- Bates, E., et al. (2003). Voxel-based lesion-symptom mapping. *Nature Neuroscience*, 6(5), 448--450.
+- Brett, M., et al. (2001). Spatial normalization of brain images with focal lesions using cost function masking. *NeuroImage*, 14(2), 486--500.
+- Kimberg, D. Y., et al. (2007). Power in voxel-based lesion-symptom mapping. *Journal of Cognitive Neuroscience*, 19(7), 1067--1080.
+- Pustina, D., et al. (2016). Automated segmentation of chronic stroke lesions using LINDA. *Human Brain Mapping*, 37(4), 1405--1421.
+- Pustina, D., et al. (2018). Improved accuracy of lesion to symptom mapping with multivariate sparse canonical correlations. *Neuropsychologia*, 115, 154--166.
+- Sperber, C. (2020). Rethinking causality and data complexity in brain lesion-behaviour inference. *Cortex*, 126, 49--62.

package/skills/10_Clinical_Neuropsychology/neuropsych-battery-selector/SKILL.md

@@ -0,0 +1,250 @@
+---
+name: "neuropsych-battery-selector"
+description: "Domain-validated decision logic for selecting neuropsychological test batteries matched to suspected cognitive deficit profiles"
+domain: "clinical-neuroscience"
+authors:
+ - "Claude Code Agent"
+version: "1.0.0"
+papers:
+ - "Lezak et al., 2012"
+ - "Strauss et al., 2006"
+ - "Mitrushina et al., 2005"
+ - "Heilbronner et al., 2010"
+dependencies:
+ required:
+ - research-literacy
+review_status: "ai-generated"
+---
+
+# Neuropsychological Battery Selector
+
+## Purpose
+
+Selecting a neuropsychological test battery is a clinical judgment task, not a checklist exercise. A competent programmer without clinical neuropsychology training will get this wrong because:
+
+- **Not all "memory tests" test the same construct.** The CVLT-II/III assesses list learning with encoding strategies; Logical Memory tests narrative recall; the BVMT-R tests visual-spatial memory. Each is sensitive to different lesion profiles (Lezak et al., 2012, Ch. 11).
+- **Test selection must match the referral question.** A dementia screen requires different instruments than a TBI return-to-work evaluation or a pre-surgical epilepsy workup.
+- **Normative data are not interchangeable.** Age, education, cultural background, and premorbid ability all determine which norms to apply and whether a given score is actually impaired (Mitrushina et al., 2005).
+- **Redundant tests waste time and fatigue patients.** Over-testing degrades performance and inflates apparent impairment, particularly in older adults and those with attentional deficits (Strauss et al., 2006).
+
+## When to Use This Skill
+
+Use this skill when you need to:
+
+- Select neuropsychological tests matched to a suspected cognitive deficit profile
+- Assemble a battery for a specific referral question (dementia differential, TBI, pre-surgical, forensic)
+- Advise on which cognitive domains to assess given a neurological condition
+- Evaluate whether a proposed battery has adequate domain coverage or problematic redundancy
+- Choose between brief screening vs. comprehensive evaluation
+
+Do NOT use this skill for:
+
+- Interpreting test scores (that requires a different skill)
+- Diagnosing neurological conditions from test results alone
+- Administering tests (this requires licensed clinical training)
+
+## Research Planning Protocol
+
+Before executing the domain-specific steps below, you MUST:
+
+1. **State the research question** -- What cognitive domain(s) are being assessed and why?
+2. **Justify the method choice** -- Why neuropsychological testing (not neuroimaging, behavioral paradigm)? What alternatives were considered?
+3. **Declare expected outcomes** -- What deficit pattern would support the clinical/research hypothesis?
+4. **Note assumptions and limitations** -- What does this battery assume about the patient? Where could it mislead?
+5. **Present the plan to the user and WAIT for confirmation** before proceeding.
+
+For detailed methodology guidance, see the `research-literacy` skill.
+
+---
+
+
+## ⚠️ Verification Notice
+
+This skill was generated by AI from academic literature. All parameters, thresholds, and citations require independent verification before use in research. If you find errors, please [open an issue](https://github.com/HaoxuanLiTHUAI/awesome_cognitive_and_neuroscience_skills/issues).
+
+## Step 1: Clarify the Referral Question
+
+The referral question determines everything. Map it to one of these categories:
+
+| Referral Type | Primary Goal | Typical Battery Length |
+|---|---|---|
+| Dementia differential diagnosis | Distinguish AD vs. FTD vs. VaD vs. DLB | 3--4 hours |
+| Mild cognitive impairment screening | Detect early decline, track progression | 1.5--2 hours |
+| TBI evaluation (acute/subacute) | Document deficits, guide rehabilitation | 2--3 hours |
+| TBI evaluation (chronic/forensic) | Quantify residual deficits, effort testing | 4--6 hours |
+| Pre-surgical epilepsy workup | Lateralize/localize function, predict risk | 3--5 hours |
+| Psychiatric differential | Distinguish cognitive vs. psychiatric etiology | 2--3 hours |
+| Return-to-work/fitness-for-duty | Functional capacity in specific domains | 2--4 hours |
+
+(Lezak et al., 2012, Ch. 5; Sweet et al., 2011 -- 78% of neuropsychologists use a flexible battery approach)
+
+---
+
+## Step 2: Identify Target Cognitive Domains
+
+Based on the referral question and suspected condition, select domains to assess. Every battery MUST cover at least attention/processing speed, memory, and executive function. Add domains based on the clinical picture.
+
+### Cognitive Domain Framework
+
+**Attention / Processing Speed**
+- WAIS-IV Processing Speed Index (Coding, Symbol Search): **~15 min** (Wechsler, 2008)
+- Trail Making Test Part A: **~3 min** (Reitan, 1958; deficient if >78 sec, ages 25--54)
+- Continuous Performance Test (CPT-3): **~14 min** (Conners, 2014)
+- WAIS-IV Digit Span (Forward): **~5 min** (Wechsler, 2008)
+
+**Executive Function**
+- Wisconsin Card Sorting Test (WCST-64): **~15 min** (Heaton et al., 1993)
+- Trail Making Test Part B: **~5 min** (Reitan, 1958; deficient if >273 sec, ages 25--54)
+- Stroop Color-Word Test: **~5 min** (Golden, 1978)
+- Tower of London/D-KEFS Tower: **~15 min** (Shallice, 1982; Delis et al., 2001)
+- Verbal Fluency -- FAS: **~5 min** (Benton et al., 1994; mean ~36--44 words total for ages 25--54, education 12+ years)
+- Verbal Fluency -- Animals: **~2 min** (Strauss et al., 2006; mean ~20--24 animals for ages 25--54)
+
+**Memory**
+- WMS-IV (Logical Memory I & II, Verbal Paired Associates I & II): **~30--45 min** including delay (Wechsler, 2009)
+- CVLT-II/CVLT-3 (California Verbal Learning Test): **~30 min** (Delis et al., 2000/2017)
+- RAVLT (Rey Auditory Verbal Learning Test): **~15 min** (Rey, 1964; Schmidt, 1996)
+- BVMT-R (Brief Visuospatial Memory Test--Revised): **~25 min** including delay (Benedict, 1997)
+- Logical Memory (WMS-IV): immediate and delayed recall; **sensitivity 90--95% for MCI when combined with CVLT** (Rabin et al., 2009)
+
+**Language**
+- Boston Naming Test (BNT-60): **~15--20 min** (Kaplan et al., 1983)
+- Token Test (short form): **~10 min** (De Renzi & Vignolo, 1962)
+- Controlled Oral Word Association (COWA/FAS): **~5 min** (listed above under executive; also indexes language)
+- Western Aphasia Battery--Revised (WAB-R): **~30--60 min** (Kertesz, 2007; use for suspected aphasia)
+
+**Visuospatial / Visuoconstructional**
+- Rey Complex Figure Test -- Copy: **~5--10 min** (Osterrieth, 1944; Meyers & Meyers, 1995)
+- WAIS-IV Block Design: **~10 min** (Wechsler, 2008)
+- Judgment of Line Orientation (JLO): **~15 min** (Benton et al., 1994)
+- Hooper Visual Organization Test (VOT): **~15 min** (Hooper, 1983)
+
+**Motor Function**
+- Grooved Pegboard: **~5 min per hand** (Klove, 1963; Ruff & Parker, 1993)
+- Finger Tapping Test: **~10 min** (Halstead, 1947; Reitan & Wolfson, 1993)
+
+---
+
+## Step 3: Assemble the Battery
+
+### Core Battery (~2--3 hours)
+
+Every evaluation should include these unless contraindicated:
+
+| Domain | Recommended Core Test(s) | Time |
+|---|---|---|
+| Premorbid estimate | TOPF or WTAR | ~10 min |
+| Attention / Processing Speed | TMT-A + WAIS-IV Coding + Digit Span | ~20 min |
+| Executive Function | TMT-B + Verbal Fluency (FAS + Animals) + Stroop | ~15 min |
+| Verbal Memory | CVLT-II/III or RAVLT | ~30 min |
+| Visual Memory | BVMT-R or RCFT recall | ~25 min |
+| Language | BNT (30- or 60-item) | ~15 min |
+| Visuospatial | RCFT Copy or Block Design | ~10 min |
+| Motor | Grooved Pegboard (bilateral) | ~10 min |
+| Effort/Validity | TOMM Trial 1 or embedded measures | ~10 min |
+| **Total** | | **~145 min** |
+
+### Extended Battery (~4--6 hours)
+
+Add these for complex referrals (forensic, dementia differential, pre-surgical):
+
+| Domain | Additional Tests | Time |
+|---|---|---|
+| Intelligence estimate | WAIS-IV (4 index scores) | ~70 min |
+| Memory (expanded) | WMS-IV (full battery) | ~75 min |
+| Executive (expanded) | WCST-64 + Tower | ~30 min |
+| Language (expanded) | Token Test + WAB-R | ~40 min |
+| Visuospatial (expanded) | JLO + Hooper VOT | ~30 min |
+| Effort/Validity (expanded) | TOMM (full) + WMT or MSVT | ~30 min |
+| **Added time** | | **~275 min** |
+
+### Assembly Rules
+
+1. **One verbal learning test**: Choose CVLT-II/III OR RAVLT, not both. They measure overlapping constructs (Strauss et al., 2006--778).
+2. **One copy figure**: RCFT copy OR Block Design for visuoconstruction screening. Use both only if visuospatial function is the primary question.
+3. **Delay intervals**: Schedule verbal memory delay recall (~20--30 min after learning) during non-memory tasks. Same for visual memory delay.
+4. **Fatigue management**: Place demanding tests (WCST, CVLT) early. Place motor tests as breaks. Offer rest periods every 60--90 min (Lezak et al., 2012, Ch. 6).
+5. **At least one validity measure**: Mandatory. Use TOMM Trial 1 (sensitivity 83%, specificity 93% at cutoff <=40; Denning, 2012) as a minimum. For forensic cases, use two or more PVTs from different modalities (Sweet et al., 2011).
+
+---
+
+## Step 4: Select Appropriate Norms
+
+### Normative Data Decision Tree
+
+1. **Age**: Always match. Most tests provide age-stratified norms.
+2. **Education**: Use education-corrected norms when available (e.g., Heaton et al., 2004 norms for TMT, WCST, verbal fluency).
+3. **Premorbid IQ**: For patients with estimated IQ far from average, IQ-adjusted norms improve accuracy over education alone. MOANS norms found BNT, Token Test, and JLO correlate more strongly with IQ (r = .47--.61) than with education (r = .24--.31) (Steinberg et al., 2005).
+4. **Cultural/linguistic background**: US-normed tests may overestimate impairment in non-English speakers or culturally diverse populations (Lucas et al., 2005; Pena-Casanova et al., 2009). Use population-specific norms when available (e.g., NP-NUMBRS for Spanish speakers).
+5. **Sex**: Match when norms are available. Grooved Pegboard shows significant sex differences: women faster than men (Ruff & Parker, 1993). Finger Tapping: men faster, especially in older groups.
+
+### Premorbid Estimation
+
+- **TOPF** (Test of Premorbid Functioning): 70 irregular words, co-normed with WAIS-IV/WMS-IV, IQ range 53--141 (Pearson, 2009). Preferred for current use.
+- **WTAR** (Wechsler Test of Adult Reading): predecessor to TOPF, co-normed with WAIS-III/WMS-III (Wechsler, 2001). Acceptable if TOPF unavailable.
+- **Caution**: Both underestimate premorbid IQ in high-functioning individuals and overestimate in low-functioning individuals (Bright & van der Linde, 2020). Supplement with demographic-based estimates.
+
+---
+
+## Step 5: Address Common Pitfalls
+
+### Practice Effects in Serial Assessment
+
+- Practice effects average **d = 0.24--0.28** on composite scores at 6--12 month retest intervals (Calamia et al., 2012).
+- No consensus on minimum retest interval; effects persist for **2+ years** on some measures (Heilbronner et al., 2010).
+- Tests most susceptible: PASAT, Stroop interference, verbal fluency, TMT-B (Beglinger et al., 2005).
+- Tests least susceptible: Digit Span, Letter-Number Sequencing (Beglinger et al., 2005).
+- **Mitigation**: Use alternate forms (CVLT-II has alternate form; RAVLT has multiple lists). Apply reliable change indices (RCIs) or standardized regression-based norms to interpret change (Chelune et al., 1993).
+
+### Ceiling and Floor Effects
+
+- **Ceiling effects**: TMT-A and simple attention tests may miss mild deficits in high-functioning individuals. Add more demanding measures (e.g., PASAT, D-KEFS verbal fluency switching) (Strauss et al., 2006).
+- **Floor effects**: WCST and complex tests may be too difficult for moderate-to-severe dementia. Substitute with simpler tasks (e.g., clock drawing, category fluency) (Lezak et al., 2012, Ch. 18).
+
+### Ecological Validity
+
+- Neuropsychological tests have modest correlations (**r = .3--.5**) with real-world functioning (Chaytor & Schmitter-Edgecombe, 2003).
+- Supplement with functional measures (e.g., Independent Living Scales, IADL checklists) when the referral question concerns everyday competence.
+- Executive function tests have particularly limited ecological validity; consider adding the Behavioral Assessment of the Dysexecutive Syndrome (BADS) or real-world task simulations (Wilson et al., 1996).
+
+### Symptom Validity Testing
+
+- **TOMM** standard cutoff (<45 Trial 2): specificity .96--1.00 but sensitivity only .15--.50 (Tombaugh, 1996). Use Trial 1 cutoff <=40 for better sensitivity (.83) at .93 specificity (Denning, 2012).
+- **WMT** (Word Memory Test): more sensitive than TOMM but higher false-positive rate in genuine MCI/dementia -- 67% of MCI patients classified as "poor effort" at standard cutoffs (Green, 2003). Use hard-easy comparison scores instead (sensitivity/specificity ~95%).
+- **Embedded PVTs**: Reliable Digit Span (RDS >= 7 as cutoff; Greiffenstein et al., 1994), CVLT-II Forced Choice <=15 (Delis et al., 2000). Use multiple embedded measures to supplement standalone PVTs.
+- **Rule**: In forensic and disability evaluations, include at least **two standalone PVTs and two embedded PVTs** (Larrabee, 2012).
+
+---
+
+## Step 6: Condition-Specific Battery Recommendations
+
+For expected deficit profiles by condition, see `references/deficit-profiles.md`. Below are summary battery modifications:
+
+| Condition | Add to Core | Remove/De-emphasize | Key Rationale |
+|---|---|---|---|
+| Alzheimer's (suspected) | WMS-IV full, CVLT-3 intrusion analysis, BNT-60 | May shorten executive battery | Memory encoding/storage is primary deficit (Weintraub et al., 2012) |
+| FTD (behavioral variant) | WCST, D-KEFS, social cognition measures, personality inventory | De-emphasize visuospatial | Executive/behavioral profile dominates (Rascovsky et al., 2011) |
+| Vascular dementia | Processing speed emphasis (Coding, Symbol Search), TMT-A/B, verbal fluency | May abbreviate language | Processing speed and executive function most affected (Sachdev et al., 2014) |
+| TBI (moderate-severe) | CPT-3, PASAT, verbal fluency, motor tests bilateral | None -- broad battery needed | Diffuse deficits: attention, speed, memory, executive (Rabinowitz & Levin, 2014) |
+| Temporal lobe epilepsy | Verbal/visual memory (laterality-specific), BNT, verbal fluency | May abbreviate motor | Memory lateralization critical for surgical planning (Jones-Gotman et al., 2010) |
+| Parkinson's disease | Verbal fluency (semantic + phonemic), JLO, clock drawing, Grooved Pegboard | None | Dual-syndrome: frontostriatal vs. posterior cortical (Kehagia et al., 2013) |
+| Multiple sclerosis | SDMT, PASAT, CVLT-II/III, BVMT-R | May abbreviate language | BICAMS recommended minimum battery (Langdon et al., 2012): SDMT + CVLT-II + BVMT-R |
+
+---
+
+## Quick Reference: Test-to-Domain Mapping
+
+For a comprehensive catalog of tests with administration times, normative samples, and sensitivity data, see `references/test-catalog.md`.
+
+---
+
+## Key References
+
+- Benton, A. L., Sivan, A. B., Hamsher, K., Varney, N. R., & Spreen, O. (1994). *Contributions to Neuropsychological Assessment* (2nd ed.). Oxford University Press.
+- Delis, D. C., Kaplan, E., & Kramer, J. H. (2001). *Delis-Kaplan Executive Function System*. Pearson.
+- Heaton, R. K., Miller, S. W., Taylor, M. J., & Grant, I. (2004). *Revised Comprehensive Norms for an Expanded Halstead-Reitan Battery*. PAR.
+- Heilbronner, R. L., Sweet, J. J., Attix, D. K., Krull, K. R., Henry, G. K., & Hart, R. P. (2010). Official position of the AACN on serial neuropsychological assessments. *The Clinical Neuropsychologist*, 24, 1267--1278.
+- Lezak, M. D., Howieson, D. B., Bigler, E. D., & Tranel, D. (2012). *Neuropsychological Assessment* (5th ed.). Oxford University Press.
+- Mitrushina, M., Boone, K. B., Razani, J., & D'Elia, L. F. (2005). *Handbook of Normative Data for Neuropsychological Assessment* (2nd ed.). Oxford University Press.
+- Strauss, E., Sherman, E. M. S., & Spreen, O. (2006). *A Compendium of Neuropsychological Tests* (3rd ed.). Oxford University Press.
+- Sweet, J. J., Nelson, N. W., & Moberg, P. J. (2011). The TCN/AACN 2010 "salary survey." *The Clinical Neuropsychologist*, 25, 218--245.