Skip to content

Image processing in R for BMI calculation

5 messages · Richard O'Keefe, Jim Lemon, Sorkin, John +2 more

Original
Richard O'Keefe
# 
"Body Mass Index" is a rather bizarre thing:
body.mass.in.kg / height.in.m^2
I have never been able to find any biological
or physical meaning for this.  Yet clinicians
are solemnly advised to measure the weight to
the nearest 0.1kg and the height to the
nearest 0.1cm.

How do you propose to determine the weight from
a single image?  Even an R package cannot perform magic.
On Mon, 1 Mar 2021 at 05:39, Paul Bernal <paulbernal07 at gmail.com> wrote:

            

      
      
      
    

        

      
      
    

        

  
  


  


    

    

      
      

        


  

        

      Jim Lemon
    

    

      
      #
    

  


  

      
I must agree with the criticism of BMI as a diagnostic index. It is
easy to tell if a person is - ahem - wide and not very high with a
single glance. These elementary parameters can easily be deduced from
an image of said person. However, it does not convey that essential
ratio of muscle to - ahem - adipose tissue that is the stated reason
for its prominence. I suggest an older, but more valid, index that was
used in the identification of witches. Simply tie the person's hands
behind their back and throw him or her into the deep end of the pool.
Time to drowning is the response variable and I am certain that those
now chastised for their adipose tissue will vastly prefer it.

Jim

        
On Mon, Mar 1, 2021 at 10:24 PM Richard O'Keefe <raoknz at gmail.com> wrote:

            

      
      
      
    

        

      
      
    

            

      
      
      
    

        

      
      
    

  
  


  


      

    

    

      
      

        


  

        

      Sorkin, John
    

    

      
      #
    

  


  

      
Colleagues,

BMI has is failures, but it has demonstrated utility. BMI predicts multiple outcome measures including cardiovascular disease and mortality. Don't through out a useful metric because it is not the perfect metric. 

As to why BMI is computed as weight/height^2, it can be shown that dividing height by the square of weight decreases the correlation between weight and height. The optimum exponent (i.e. the power that most effectively minimizes the correlation between weight and height) is not the same in men and women and it differs by race. In general an exponent of 2.0 is best for men; for women an exponent of 2.2 is a bit better than 2.0.

In any event, don't let the perfect get in the way of the good.

John


John David Sorkin M.D., Ph.D.
Professor of Medicine
Chief, Biostatistics and Informatics
University of Maryland School of Medicine Division of Gerontology and Geriatric Medicine
Baltimore VA Medical Center
10 North Greene Street
GRECC (BT/18/GR)
Baltimore, MD 21201-1524
(Phone) 410-605-7119
(Fax) 410-605-7913 (Please call phone number above prior to faxing)

        

      
      
    

  
  


  


      

    

    

      
      

        


  

        

      Abby Spurdle
    

    

      
      #
    

  


  

      
I can't help but feel that a discussion on the merit of BMI is a
digression, from the OP's question.
In addition, to being of no relevance to "R Programming".

In relation to Richard's technical comments:
As per my previous post, it is possible to get *relative" measures.
(Assuming the images are not on a standardized scale, which they could be).

In relation to Jim's anatomical comments:
An improved algorithm could evaluate a subject's body type.
In addition to factoring in gender, ethnic characteristics and age.

In any case, I don't see why image-based classification is any less
worthwhile than conclusions based on timeseries plots and
scatterplots...

  
  


  


      

    

    

      
      

        


  

        

      Bert Gunter
    

    

      
      #
    

  


  

      
This discussion is completely offr topic here. Please take it elsewhere.

Bert Gunter

"The trouble with having an open mind is that people keep coming along and
sticking things into it."
-- Opus (aka Berkeley Breathed in his "Bloom County" comic strip )


On Mon, Mar 1, 2021 at 1:42 PM Sorkin, John <jsorkin at som.umaryland.edu>
wrote: