HASCI - -How well does it work? Now we'll be able to find out!

[AlmostUndone]

I counted 1243 lost hairs in my donor.

In the "before" images, I can confirm a total of 4354 hairs which existed in the extraction sites. In the "after" images, using all the multiple viewpoints available, I could verify no more than 3248 hairs left. This simple calculation reveals 1106 lost hairs.

However, I observed a bunch of extraction sites, where we could see more hairs in the "after"-image than in the "before"-image. That totals up to 137 hairs (named in the excel sheet as "reduntant hairs"). This comparison shows that there were 4354+137-3248 = 1243 lost hairs in total.

I should be able to justify every hair count in the images by looking at the extraction sites from all the viewpoints we have. My analysis has the advantage over Arashi's in being relatively easy for anyone to verify. Instead of individual extraction points, the basic unit of my analysis is a group of 4 extraction points. Now hopefully I remembered to include everything:

EXCEL COUNT SHEET: https://dl.dropboxusercontent.com/u/...ount_march.xls



Left side, before: https://dl.dropboxusercontent.com/u/..._before_98.jpg
Right side, before: https://dl.dropboxusercontent.com/u/..._before_98.jpg

Left side, after: https://dl.dropboxusercontent.com/u/...t_after_98.jpg
Right side, after: https://dl.dropboxusercontent.com/u/...t_after_98.jpg

Left side, extractions: https://dl.dropboxusercontent.com/u/...ractions97.jpg
Left side, extractions (AN ADDITIONAL VIEWPOINT): https://dl.dropboxusercontent.com/u/...%20PICTURE.jpg
Right side, extractions: https://dl.dropboxusercontent.com/u/...ractions97.jpg
Right side, extractions (AN ADDITIONAL VIEWPOINT): https://dl.dropboxusercontent.com/u/...%20PICTURE.jpg

[Arashi]

Arashi eh, I'm working my ass off even when I'm not engaged to this project. My own analysis killed off a week at least, but I really wanted to make an easy way for anyone to see how many hairs were lost.

Dont get me wrong mate, I know you do. It's a shitload of work. It's just the suspense man, we're so close to the final conclusion, can't wait But again, I'm not complaining at all, in fact, I'm really grateful for all you're doing ! And I'm sure others are too. Without you we would never have had this analysis !

I counted exactly the same result as you previously did in the first area

Awesome !

[AlmostUndone]

Arashi eh, I'm working my ass off even when I'm not engaged to this project. My own analysis killed off a week at least, but I really wanted to make an easy way for anyone to see how many hairs were lost.

Verifying the analysis you made is slow work, as it's difficult to look up numbers, and with the need to compare before/after images side by side. In my analysis you don't always need to compare the images side by side. I composed a rectangular grid where all the circled extraction sites fall into groups of four. You can just look up any arbitrary group from the excel sheet to see if everything adds up, and circling errors are also easy to detect in this way.

I counted exactly the same result as you previously did in the first area. You counted 91 hairs lost in 104 extraction points. I counted 89 hairs lost in 105 extraction points plus 2 sites, where we saw more hair in the "after" photo than in the "before photos.
Meaning this: 216 out of 305 + 2 hairs grew back. The rest, 91 hairs, were lost.

https://dl.dropboxusercontent.com/u/81057175/CTEST_march.jpg

[Arashi]

I felt it necessary to begin my own analysis of the donor area. It's quite almost ready, but cost me those few days which I could have spent linking recipient photos. But I wouldn't mind linking them, so I won't need to upload 60 large-sized photographs for you?

You can't circle any grafts before I've taken the final photographs. No I can't promise it'll be done this month. It depends a bit on the sun's angle into my apartment; I rented huge electric lights for the "before" pictures, but it was more convenient to rely on daylight as I did in the rest of the photos.

Hmmm not even this month, that's quite a disappointment. But then again, it's your project man and I'm grateful you're doing all this.

Again, if there's any way I can help out to speed things up, take some work of your table, just let me know.

Good luck !

[AlmostUndone]

@Almostundone, is there any way I can help you out with recipient already ? Basically I'm just waiting for you, I have some free time and I'd love to dedicate it to this project. If you can think of any way I can help you at this stage, maybe linking photo's, maybe circling grafts, just let me know mate !

I felt it necessary to begin my own analysis of the donor area. It's quite almost ready, but cost me those few days which I could have spent linking recipient photos. But I wouldn't mind linking them, so I won't need to upload 60 large-sized photographs for you?

You can't circle any grafts before I've taken the final photographs. No I can't promise it'll be done this month. It depends a bit on the sun's angle into my apartment; I rented huge electric lights for the "before" pictures, but it was more convenient to rely on daylight as I did in the rest of the photos.

[Arashi]

@Almostundone, is there any way I can help you out with recipient already ? Basically I'm just waiting for you, I have some free time and I'd love to dedicate it to this project. If you can think of any way I can help you at this stage, maybe linking photo's, maybe circling grafts, just let me know mate !

[Arashi]

This is impressive. Arashi, you should do this for money, through a website or something! I would probably pay a few hundred dollars for you to map my hair counts before, and during different regiments, especially in the MPB zones. Does anyone offer this kind of thing? Certainly you could design some kind of software that is able to make it easier?

Thanks but one really needs the kind of high quality photo's that AlmostUndone shot, most people are not able to shoot them like that. And doing it manually, it's tons of work. And I just built upon what AlmostUndone did first, the linking of the photo's and then the marking of the grafts with circles, he must have spent at least the amount of time that I've spent, so doing it manually, that's a LOT of work.

Writing a computer program for it, thats an interesting thought though

[jamesst11]

This is impressive. Arashi, you should do this for money, through a website or something! I would probably pay a few hundred dollars for you to map my hair counts before, and during different regiments, especially in the MPB zones. Does anyone offer this kind of thing? Certainly you could design some kind of software that is able to make it easier?

[Arashi]

BTW, something we haven't discussed: sometimes you'll see that in the after photo a graft has *more* hair than in the before photo. In those cases I counted the result as 0. This could happen either due to photo's not really telling the whole story or, more likely, to hair coming out of resting state. So, since I didnt count those hairs that come out of resting state but did count the hairs that go into resting state (as hair loss), I think it makes a lot of sense to use that 6% correction I demonstrated before (so lower the final result with 6 %).

Actually I would bet that if you count the extra hairs in grafts in the after photo, you'd end up with that 6% too.

[Arashi]

I thought there were gonna be loads of mistakes in my work, but when I re-checked, I think I only found 1 badly placed circle in the afterimages. Pretty good, huh? That said, some circles had to be added. I'll take a closer look at Arashi's work sooner or later.

It's of course totally up to you but I'd love for you to first do the recipient ! I double checked my own work now and I'm pretty sure it's good. An option would be though, if you want to take get a quick idea of the correctness of my analysis, to analyse just one area, take a random area and analyse that area and you'll see that it's quite good

Also for anyone in general interested in verifying my work: Like AlmostUndone said before, make sure you are NOT using windows photo viewer: it anti-aliases and makes it harder to count. I'm using 'irfanview' myself, that doesnt anti aliase. And make sure you zoom in enough, sometimes you really need to zoom in to see if there's 2 or 3 hairs. Look good for really thin hairs, sometimes a hair gets a lot thinner in the after photo. And it's a good method to look for the hair ends, sometimes you see like a small balloon there at the end of the hair and you can look for these in difficult situations.

There will always be situations where it's really hard to tell how many hairs there are though in a graft. In those cases I tried to vary the estimates: first graft I really doubt I take the high estimate, 2nd the low etc. That's probably the best way to negate the uncertainty as much as possible.

[Arashi]

And oh yeah, 80% regrowth would now mean that we'll need to see 6.85 hair/graft on average in your recipient. Wow that would be something eh But since I didnt encounter any more than 5 hairs/graft in your donor (mostly 2-3 hair/graft), that only seems to be possible in HASCI fairytale land

[Arashi]

I thought there were gonna be loads of mistakes in my work, but when I re-checked, I think I only found 1 badly placed circle in the afterimages. Pretty good, huh? That said, some circles had to be added. I'll take a closer look at Arashi's work sooner or later.

Your works was really near perfect !! I think I only removed a circle around a graft 2 times. You did forget a couple of circles here and there, but it weren't many neither.

Anyway I've finished recounting everything ! I've uploaded the changed excel sheets. What I still haven't done is analysing alternative views. Nor have I checked the extraction points to be correct (I only did that for Area 1, you didnt make any errors there and you work very neatly in general so I figured the rest was ok too, but I might check a few area's for errors there).

Anyway, so my final result based on the above:

You have lost 1166 hairs in your donor and I've counted 848 extraction points with less hair than before of the total of 1599 extraction points that you had.

So 848 grafts that's just 6% more than you had done (800 grafts), I think it's pretty close ! Again, the reason here might be a combination of counting errors, hairs going into sleeping state and/or that HASCI damaged some grafts during failed extractions. Anyway if we go with the scenario that's best for HASCI and we're assuming that no grafts were damaged during failed extractions, then that would mean we'd need to lower the result with 6%, so that then would be 1096 hairs really lost in donor.

So, if your average recipient density would be higher than 1.37 hair/graft, then that would point at regrowth (well actually we'd need to see that 6% extra to be sure, so above 1.45 hair/graft would point at regrowth). Lower density than 1.37 hair/graft would equal net hair loss ! Keep in mind that in all 3 previous hasci recipient cases we've analysed, I didnt see a result above 1.3 hair/graft in donor.

Can't wait to look at your recipient though

[AlmostUndone]

I thought there were gonna be loads of mistakes in my work, but when I re-checked, I think I only found 1 badly placed circle in the afterimages. Pretty good, huh? That said, some circles had to be added. I'll take a closer look at Arashi's work sooner or later.

[Arashi]

Today I've recounted the left side, area 1 to Area 7B. Only made a few changes and the final result is again almost the same. Uploaded the edited files. Will try to finish recounting tomorrow.

After that I'll probably take a look at the alternative views too (still haven't done that)

[Arashi]

I've recounted the right side, the result is almost the same (uploaded the new file). I'll recount the left side this week too and while waiting for the donor pics I'll also compare with alternative viewpoints !