10.6 Femoral ROI Analysis: Femoral Neck, Total Hip, and Ward's Area

10.6 Femoral ROI Analysis: Femoral Neck, Total Hip, and Ward's Area

Quick Answer: Central DXA generates five proximal femur regions of interest (ROIs): Femoral Neck, Total Hip, Trochanteric, Intertrochanteric, and Ward's area. Under ISCD positions, adult osteoporosis diagnosis must use either the Femoral Neck or Total Hip, selecting whichever T-score is lower. Ward's area is an algorithm-derived search box identifying minimum pixel density; it has poor precision and systematically overdiagnoses osteoporosis, making it strictly prohibited for clinical diagnosis. Trochanteric and intertrochanteric sub-regions must not be used alone for diagnosis. Quality criteria require placing the neck box across the narrowest dimension of the femoral neck perpendicular to the neck axis, completely clear of the acetabulum and greater trochanter.

Standard Proximal Femur Regions of Interest (ROIs)

Densitometry software automatically identifies anatomical landmarks on the proximal femur and places specific regions of interest (ROIs) to calculate bone mineral content (BMC), area, and areal bone mineral density (BMD):

• Femoral Neck ROI: Rectangular region placed across the narrowest constriction of the femoral neck, with borders perpendicular to the neck axis. It measures a critical weight-bearing segment containing ~25%–30% trabecular and ~70%–75% cortical bone.
• Total Hip (Total Proximal Femur) ROI: Composite region integrating three sub-regions: femoral neck, trochanteric region, and intertrochanteric region. It excludes the femoral head, acetabulum, and shaft below the lesser trochanter. Enclosing a large area (~30–45 cm²), it provides the lowest measurement variability and greatest precision in hip densitometry.
• Trochanteric Sub-Region: Encompasses the greater trochanter (~50% trabecular bone). It responds dynamically to therapies, but soft-tissue variability limits precision.
• Intertrochanteric Sub-Region: Diagonal band between the neck and the lower margin of the lesser trochanter, composed of mixed trabecular and cortical bone.
• Ward's Area (Ward's Triangle): Small square search ROI (~1.0 cm × 1.0 cm) locating the lowest-density pixel cluster. It is an algorithmic calculation rather than a fixed anatomical structure.

ISCD Official Positions on Diagnostic Selection

The ISCD establishes explicit rules governing adult diagnostic classification (postmenopausal women and men aged ≥50):

Diagnostic Selection Rules

1. Core Diagnostic Rule: Diagnosis of osteoporosis in adults must use either the Femoral Neck or the Total Hip, selecting whichever site displays the lower T-score.
   • If femoral neck T-score is -2.6 and total hip T-score is -2.1, diagnosis is osteoporosis based on the femoral neck.
   • If femoral neck T-score is -1.7 and total hip T-score is -2.7, diagnosis is osteoporosis based on the total hip.
   • Both sites are validated against the female NHANES III reference database.
2. Longitudinal Monitoring Preference: While either site establishes diagnosis, the Total Hip ROI is preferred for longitudinal monitoring because its large area yields superior precision (CV ~1.0%–1.5%) and the smallest Least Significant Change (LSC).

Explicit Diagnostic Exclusions: Ward's Area and Sub-Regions

ARRT specifications emphasize regions that must never be used alone for clinical diagnosis:

Why Ward's Area is Prohibited

Under ISCD positions, Ward's area must never be used for diagnosis or monitoring:

1. Algorithmic Artifact: Ward's area is an automated search calculation seeking minimum density pixels. Its T-scores run 1.0 to 1.5 standard deviations below the femoral neck in normal individuals.
2. Systematic Overdiagnosis: Relying on Ward's area creates severe false-positive overdiagnosis of osteoporosis in healthy adults, leading to inappropriate drug therapy.
3. Poor Precision: Small shifts in leg rotation cause the algorithm to jump across different pixels, producing unacceptably high precision error (>3.0%–4.5%).

Trochanteric and Intertrochanteric Sub-Regions

Individual trochanteric and intertrochanteric sub-regions must not be used alone to diagnose osteoporosis. They are valid only as components of the composite Total Hip ROI.

Quality Criteria and Verification of the Femoral Neck Box

Automated ROI placement requires visual inspection and manual adjustment:

1. Neck Axis Alignment: The software places a midline along the femoral neck; the neck box long borders must sit perpendicular to this axis.
2. Narrowest Constriction: The box must be centered over the narrowest waist of the femoral neck.
3. Clearance from Acetabular Rim: The medial border must never overlap the acetabular rim or ischium. Overlapping pelvic cortical bone falsely elevates BMC and BMD, masking osteoporosis.
4. Clearance from Greater Trochanter: The lateral border must not touch the greater trochanter cortex, which falsely elevates femoral neck BMD.
5. Software Midline Bisector: The software midline must accurately bisect the femoral shaft and femoral head centers.

Window Level During Femur Analysis

The femur is the site where windowing earns its keep, because its critical boundaries are bone against bone: the femoral neck against the ischium medially and against the greater trochanter laterally. At default display settings those interfaces can be invisible, particularly in low bone density. Adjusting window level and window width separates them visually so the neck box can be verified and repositioned. As everywhere else, windowing changes only the display — not BMC, area, or BMD. Window level is a named scan-analysis topic in the ARRT content specifications effective January 1, 2027, and is described in full in the scan analysis section of the forearm chapter.

Reported Values and Graphical Displays

A proximal femur report lists each region separately — femoral neck, trochanter, intertrochanteric, Ward's area, and the composite total hip — and for each gives BMC (g), projected area (cm²), BMD (g/cm²), T-score, Z-score, and percent of the young-adult and age-matched reference means.

Two plausibility checks belong to this site specifically. Ward's area should be the lowest value on the report; if it is not, the analysis has gone wrong somewhere. And the femoral neck is normally lower than the total hip, because the total hip dilutes the neck with the denser intertrochanteric region — a femoral neck reading higher than the total hip suggests the neck box has captured ischium or trochanteric cortex.

The graphical displays accompanying the numbers are a reference-curve plot positioning the patient against age-related normative curves with T-score bands, and, once a prior study exists, a serial trend plot of BMD over time. Both are interpretive aids: classification comes from the lower of femoral neck and total hip, and significance of serial change comes from absolute BMD difference against the facility least significant change — never from the visual slope of a trend line. A serial plot that silently joins a right hip to a left hip, or points from two different scanners, will draw a perfectly convincing line through incomparable data.