---

This project provides programs for predicting trunk muscle size and position values given sex, age, height, and weight. The predictions apply regressions developed based on CT measurements in a multi-ethnic sample of the Framingham Heart Study. This implements the regressions in Python code, specifically enabling users to run online via Google Colab notebooks. This may be of interest for researchers creating musculoskeletal models or other studies needing estimates of muscle morphometry.

The code (available in downloads) provides estimations of trunk muscle cross-sectional areas and positions at the vertebral levels between T4 and L4. Copy the Google Colab notebook to your Google Colab account to run. https://colab.research.google.com/

The form requires submission of a subject's sex, weight ( kg / lb ), height ( cm / in ), Age and ID (optional). After clicking the play button on the form an excel workbook will be downloaded. It contains four sheets: Cross-Sectional Area in mm^2 , Distance in mm^2 , Angle in degrees and an additional sheet describing your subject's inputs. Note that this calculator was developed with a dataset containing healthy adults between ages 40 and 90. Application outside this range may not be accurate, and this should not be used for children and adolescents.

FAQ.
1: What sample pool was used to generate the regression used in this calculator?

Table 1: Mean (SD) [Range] characteristics of participants included in the sample.

	Men (N=247)	Women (N=260 )
Age (years)	60.8 (14.1) [40-88]	61.8 (12.6) [40-90]
Height (cm)	173.8 (7.2) [155.5-193.7]	159.9 (6.6) [139.7-175.9]
Weight (kg)	86.0 (14.4) [47.2-122.9]	70.7 (15.3) [40.4-127.0]

2: Can this calculator be used for anyone?
The calculator can be used for anyone who falls within the data ranges noted above (i.e age 40 – 90, 140cm-195cm (4ft-6.4ft) and 70kg -130kg (154lbs-286lbs). Outside these ranges, the calculator may still be used, but will generate a warning that predictions are being extrapolated. Prediction intervals will also increase as inputs move outside the range of the sample. If an age < 40 is entered, the calculations will be performed for age = 40, as aging-related effects are likely not found in the same way for adults under 40.

3: How were the muscle distance and angle measurements calculated?
Measurements were performed in transverse plane CT scans at the mid-level of the vertebral body. After segmenting a muscle, the CSA is defined as its area in this plane. The distance and angle refer to the transverse plane polar coordinates representing position of a muscle’s centroid in relation to the centroid of the vertebral body, where the posterior direction is 0°.

4. What are the prediction intervals?
The prediction intervals are calculated at each vertebral level along with the predicted value. The prediction intervals for an outcome (lower 95% and upper 95%) provide a likely range of values for an individual with the given input sex, age, height and weight. A hard lower limit of 0 is applied for CSA predictions and distance predictions, and lower and upper limits of 0° and 180°, respectively, for angle predictions.

5. How do I run multiple individuals at once?
Use the MuscleCalculator_ForBulkUse. pynb code and fill in the arrays with your individuals' information, using commas for delineation and quotation marks for Sex, WeightUnits, HeightUnits and Names. Click run and your files will download. Your browser may ask you to allow multiple files to be downloaded. If you do not see a notification for the files being downloaded, check your google drive folder as some browsers may automatically send it there.