Extending Cancer-patients' survival times by quantified feedback

Following on from my original intention of living a long time by self-quantifying, this became more focused when I was diagnosed with terminal cancer with a ‘dead-by-date’ a year later ('Oct. '08).

Whilst I can never be sure why I’ve outlived the first three experts’ initial prognosis (a later one gave me about 5 years), I strongly suspect that it is because I was fairly healthy to begin with; and have been pro-active in incorporating cancer markers into my quantification - and adjusting my lifestyle partly on the basis of these.

So I’d like to discuss with others who have a similar objective: of using self-quantification to enhance their survival. In particular, data mining of all this data with a view to seeing what correlates with changing cancerousness (as indicated by cancer markers as well as ancillary factors like immune system measures.

This is an important story and I’d like to learn from it. What cancer markers are you using? What lifestyle adjustments have you tried?

I use 7 with various regularity: NMP-22, PSA, CEA, CA 19-9, CA-125, hCG, and now CxBladder.

Lifestyle changes (briefly): PACE instead of aerobics; low-carb instead of so-so carb/alt-day fasting; little sugar or alcohol vs. regular use; hyperthermia; vegan/vegetarianism/meat; fish oil vs. low fat. Along with all these changes, different supplementation regimes (presently trying one supposed to lower homocysteine; I also tried one to limit damage to my kidneys when I found one was shrivelling).

Ian would you be willing to share one of your charts, in the form of an image, which I think would inspire others?

I met a very interesting gentleman about a month ago, at the Foresight Reunion conference. His name is Ted Howard, he’s a software developer, politician, philosopher, pilot and cyclist, and he has survived melanoma essentially by adopting a vegan diet and taking high doses of Vitamin C. What makes Mr. Howard special is that he’s a very scientifically-oriented person, and Quantified Self material. He posted about his fight with melanoma here, and I’ve e-mailed him to join this discussion.

Just a quick into to say that I have accepted Dan’s invitation and joined this group.

Interested in software - particularly for either PC or Android phone, to make quantification and charting easier.

There is another page http://tedhowardnz.wordpress.com/cancer-treatment/ that contains information which may be useful to others with similar diagnosis.

Gary, I’m more than happy to share.

I’ve copied 4 of my graphs onto a Word doc. Alas, I don’t know how to post it on the forum.

So, any help would be appreciated; and will be rewarded by my posting 4 of my graphs.

I do also have a slew of biochemistry ones - such as neutrophils, albumin, etc, but I didn’t think you wanted those.


Hi Ian, instead of Word DOCs, I highly recommend PNG screen captures (smaller, higher quality, no risk of macro viruses).

To capture the screen in Windows 7, ignore the Alt+Print Screen combination. Instead, tap the Windows Key. The Start Menu will open. There, start typing “Snipping…” You’ll see the Snipping Tool right away. Launch it, then select the area to take a screenshot of. After you see the screenshot in the Sniping Tool window, go to Edit -> Copy.

Now in the new forum you can paste the screenshot directly into a post! No need to save to a file.

If you have screenshots saved already, you can upload them by clicking the button from the post editor toolbar.

Great Dan. Thanks for that clear, concise and coherent set of instructions. If only everyone did it so (especially when giving directions).

Herewith the png

Thank you for posting these, Ian.

Hi Ian,
It looks like you are doing and keeping track of a lot of different things for your health everyday. I’m curious as to how you manage all of this: What tools do you use to remember all your planned activities, to keep track of your activities and results, and to analyze your data?

Hi Rajiv,

Mainly Excel SS. There are ancillary stuff, such as for exercise HRM on proprietary software, but I transfer the summary to the SS. Some things are initially recorded on good old pencil and paper before transfer to the SS.

The analysis, to date, has mostly been simple correlations. I did use XLMiner briefly, but didn’t get too far (tho’ it seemed to do most of what I wanted from data mining); I’m hoping to get in to RExel - awaiting a guide.

Hi Ian - the correlations… are you just looking at the graphs, or are you doing anything more elaborate?

Initially I was just using Excel’s graphical correlations between two variables - body compositions and exercise. Latterly I did some XLMiner ANOVA, but got bogged down on its restrictions and so put it to one side whilst I backfilled my data (ongoing).

Hey Ted & Ian,

I’m working on an app to track user defined symptoms. I’m grappling with a recent Autoimmune DX, need to report symptoms to refine DX and treatment options. BUT at this point, my $$ is going to iPhone dev only…

Curious to what kind of metrics you’d be interested in tracking?

You may want to try my app, Tonic (www.tonicselfcare.com). People are using it to track, and remember to do, a very wide variety of health activities & symptoms. Many use it primarily to just make it easier to get through the day. But others have made use of the tracked information to do refine their health practices, as you want to do.

Hi Tasha,

Not sure anything I do can be of help here. I track lots, and then enter it into an SS:
exercise, by watch/chest HRM, aggregated over the day;
nutrition - by noting it in several dimensions (alcohol, carbs, sugar, specific items)
biochemistry: using urine dipsticks for 10 dimensions; 8 cancer markers; 57 others serum ones
lung capacity: Peak flow; and pulse rate after 2km steady walk.
Supplements: over 100 (not all at the same time)
CAM Treatments: hyperthermia, etc
prescription drugs: 29

Phew, that’s quite a few metrics. How can you test what works/what doesn’t when you have so many variables?

Well, that’s where data mining techniques come in.

I can (and do) simple inspection, two-variable correlations. I check each independent variable against the dependent ones - mainly the cancer markers.

But if anyone here is knowledgeable in data mining, I’d welcome some help.

I’ve tried XLMiner - and might go back to that, but it is expensive; I’m now trying to get to grips with RExcel.