Excedoeffluo's thoughts on the edge of the abyss...

excedoeffluo @ 2009-09-23T23:10:00

2009-09-23T22:11:25Z

25 in 50 minutes, yikes...

Nicked from Pete

2009-07-25T14:03:49Z

Leave me a comment and:

1. I’ll respond with something random about you.

2. I’ll challenge you to try something.

3. I’ll pick a color that I associate with you.

4. I’ll tell you something I like about you.

5. I’ll tell you my first/clearest memory of you.

6. I’ll tell you what animal you remind me of.

7. I’ll ask you something I’ve always wanted to ask you.

8. If I do this for you, you must post this on yours.

Long day

2008-10-18T20:34:55Z

First off, thanks guys for your comments on my previous post, made me feel heaps better about everything :)

Now, to the business end of this post. It's been a long day.

Drove to Bristol and back, thats about a 5 hour round trip, the longest i've ever personally driven, and took the keys to my new flat. Typically the landlord hadn't done some stuff he was meant to and will be seeing him next week to get it sorted. Bought kitchen stuff, iron, vacuum cleaner, kettle, bookshelf and wardrobe and cabinet set for bedroom and ordered tumble drier to add to the future deliveries of bed and sofa. Set up TV, no arial signal, nevermind Virgin Media will be along next week to hook me up.

Going down next friday to set up everything and indeed to stay, job starts on the monday after!

I'm pretty tired, but very pleased that things seem to be going according to plan! Hope you're all well :)

A brief update...

2008-08-15T23:05:57Z

Had my first face to face interview today for a job, if succesfull will have an assemsment centre to come. They said it;d be about and hour, but took an hour and a half, whihc I'm taking as good as they would surely have kept it short if they didn;t like me.

Got two more interviews for different companies next week, both in bristol. Here's hoping all goes well and I have one or more job offers sooner rather than later!

Hope you're all well :)

Graduation

2008-07-05T01:50:41Z

So, graduation day, last saturday:

...

Never Enough.

2008-05-19T00:04:38Z

Life begins here.

I have no idea what to do.

At all.

A qualification is soon at hand, but I don't know what to do with it.

All suggestions welcome!

Something new.

2008-04-11T00:12:51Z

So. Here I am. It's been a while, a while since I've presnted anything of substance at any rate.

I've been busy, on a variety of scales. I've been doing a lot of project work, I've been doing a lot of nothing, I've been doing some revison, I've not been doing much vis-a-vis talking to people.

I really have no excuse for not having kept in better topuch with people, none whatsoever. But right now is the final puch, right now is the knocking it out of the park. Right now is the revising, working hard, getting the goddam exams done, gettign the presnetations sorted and then...

Well then, is begining my life proper. Finding a job, a career. Maybe finding a girl and making a future. Life in general.

Problem is, I don't know what I want to do. I have no idea. Well, I have one idea, but I'm keeping that quiet for the moment.

I'm sorry that I've not been there for you guys more, I'm sorry that I haven;t been a better friend, and I hope that in the future we have an ever stronger relationship.

I wish you all the very best, keep safe my friends :)

So, I've been away from here for quite some time...

2008-03-03T02:06:29Z

I hardly ever post here anymore, but I do keep up with my friends on here, one of the most useful ways to catch up with everyones lives as I'm on MSN so infrequently these days. And, among other reasons,
A VOLUMETRIC MODEL FOR TUMOUR SPHEROIDS
Jonathan Hudson
MEng Cybernetics, siu04jh2@rdg.ac.uk

ABSTRACT
This report covers the creation of a new model of early stage tumour spheroid growth and its visualisation in a virtual environment. The report discusses the original formulation of the model and then its reworking and improvement into an improved model.

1. INTRODUCTION

In the modern world, cancer is constantly in the top list of causes of death. Cancer is the name given to a set of diseases in which cells divide without bound regardless of the limitations normally imposed upon their cell type. Statistically 1 in 3 people will develop some form of cancer at some stage in their lives and for one in 4 it will be the cause of their death. As such a great deal of research has been done into its detection, treatment and prevention. Until relatively recently most of this research was based heavily on molecular biology, cell biology and drug delivery methods, however mathematical and computational methods have been coming to the fore in more recent times. Large volumes of data have been found through research in the other sciences which can be analysed with mathematical approaches and utilised to form models of this deadly disease.
It is believed that by modelling cancer we can use biological data from both in vitro (experimentation performed outside of an organism in a controlled environment) and in vivo (experiments performed within a living organism) experiments to refine these models to create a better understanding of cancer and better ways to model it. If models can be created that portray cancer very well, then such models could be used in predicting a specific patients tumour progression, and better understanding of both how cancer behaves and grows can be taken from them.

2. THE VOLUME MODEL

Work was carried out to implement a model [] that was considered by it’s creators to be a good model for in vivo (within a living organism) tumour growth. Work was then undertaken to extend this model into 3D so that it could more accurately represent tumour growth.
The model was built as a set of rules to be applied to a Cellular Automata, a common method used in discrete models. Each location in the CA is used to represent one or a number of cells and then rules are written to be applied to the system as time progresses.
It was decided to use the rules of this model as the basis for a scaleable model that could allow any point to represent any number of cells, and still allow for each of those cells to be treated individually within the element rather than in a group as a whole.
It was believed that a volume based model using simple parameters could be developed that could be grown in a small (and so easily visualised) domain but represent, hopefully realistically, a vastly larger number of cells than individually visualised items. The idea behind a variable volume representation model being that it might be able to work faster than a cell for cell system, or even a multi-cell to cell system in that it doesn’t care about the specific location of every individual cell however the rules still act on each cell in a given area allowing them to act as if they were each in a defined place. Strict parameterisation would be required to allow truly realistic behaviour, however ‘ball park’ figures can be used to show the potential of the system to grow realistically and further developed.
Another advantage of such a volume based system is that it does not restrict the actual arrangement of cells. If a slice of an actual tumour is taken, or indeed even of normal cells It can easily be seen that they do not grow in perfect grids (as a normal CA representation would imply), but in a less uniform configuration. Some models have been developed that use vorroni tessellations etc to model this disparity of cell alignment [ref]. By not explicitly caring about the alignment of cells and their neighbours, the model actually becomes more realistic, although in an abstract way. Of important note is the fact that the resulting interactive visualisation of this model is representative only, although there are ways to make it more realistic as are explored below.
This model would then by simple extension allow itself to be placed into an inhomogeneous and/or physically limited environment and so could be used as the basis for a variety of research areas.

2.1. Development of the basic model

Previous work led to an understanding of how to code an algorithm that can be applied to all the cells. Beyond that the set of process’s that would be required for the volume model was determined and are as follows:

1. Contribution – Each CA point that contained enough cells to reach the threshold for reproduction would then have a certain chance to contribute cells to itself or its neighbours. For each point that fulfilled these criteria the model would first calculate how many cells were determined to have divided. The next part would then determine their placement into the 26 surrounding cells and to itself with an emphasis on adding cells to locations that already had cells in them, so as to not have the program spread itself right across the environment leaving huge gaps with no cells.
2. VolumeCalc – This function would run through every point in the environment. At each location it would check the surrounding areas (and itself) to see if any other location was providing it with cells. If cells had been determined to be transferred, then they were added to the volume of the current location and removed from the contribution location of the donor location. If the volume was already full then the model would be informed that ReContribute should be called. If it could take some but not all of the donated cells then it would take what it could and reflect that in the contributions from the donor locations.
3. ReContribute – After the model had finished the main run it would then, if determined necessary, run this function. For each point which still had cells to be contributed, their contribution location was re-determined and then VolumeCalc was called again
Once the code was written and debugged it led to the creation of what appeared to be the spheroid shape that was expected. However the system would repeatedly create a model that was misshapen in the same way. Going through the code again some problems were determined and fixed but the same general results were being created.
The process was slow to run and upon looking through the system at another time, various errors became obvious, all three main functions were rewritten to create a better model, but problems were repeatedly encountered with both Contribute and ReContribute. It was decided to completely rewrite the code based on a different dynamic but the same basic volume and contribution idea.

2.2. Reformulation of the basic model

The new formulation removed the need for each cell to calculate contributions to every viable location surrounding it and removed the ReContribute function entirely; thus leading to a more streamlined and faster running model. The original versions of the volume method can be found in the appendix.
The new code consisted of only two main functions:

1. VolumeCalc – This now simply calls Contribute once and performs some minor information-to-console tasks.
2. Contribute – At each cell with a volume over the determined reproductive threshold the number of newly created cells is determined, as each one is created it’s location in the current cell or the surround is determined.
This new code only needed to run the Contribution function once, and as each cell’s final destination was determined as it was created, the need for any ReContribution function disappears.

2.3. Introduction of ‘mechanical movement’

If new cells were determined to be made but there was nowhere for them to go, it was simply assumed that they weren’t created. This however is not very realistic; in fact tumour cells don’t care if there is nowhere for their daughter cell to go and create it anyway, pushing the surrounding cells away. A lot of models disregard this behaviour in order to save on the calculations involved in repositioning the surrounding cells, although there are exceptions [ref] [ref]. It was decided however that this issue would be tackled within this new model.
The volume approach of the model was considered to lend itself well to the idea of cells being pushed around by new ones due to individual cell location not being a factor. A new function was created that would be called every time Contribute determined that a cell was to be contributed to a zone that was already volumetrically speaking, full. This would look at the surrounding area of the overfilled location and determine a free place to shift a cell into. If none can be found then a location would be randomly selected and MechMove recalled.
Although more computationally intense, it is believed that this is more realistic model. In a set of soandso length runs it was seen that with a region of X cells a region of X contributions to full areas was made and X actual moves of cells were made.
The volume approach of the model means that only the volume need be taken into account in moving cells in full areas as although the number of cells at a location is known, their exact configuration is ignored. This means that cell movement due t overcrowding can be represented, but depended on the represented volume of each location, dramatically fewer calculations are actually required to do so.

3. VISUALISATION

There are a variety of ways to visualise a model such as this, due to the wish to make the visualisation interactive, a virtual reality type environment was selected. Two possible programs that could be used for this desired result are OpenSG and VTK, both allow data to be easily realised in a virtual realm and have been used as the visualisation process for several medical programs [ref] [ref] [ref] [ref]. OpenSG was chosen due to previous work conducted with the program that would mean less time would be required to get to grips with it to be used with this model.
A simple viewport representation of the model was generated. Due to the fact that this is not a cell by cell model, the resultant visualisation is to be taken as a representation of shape only.
The resulting visualisation can be manipulated through the use of the mouse, the model can be zoomed in on and panned around. Through a basic manipulation of the attributes of the rendered points, the model can be set to render parts with high translucency so as to allow the model to be viewed in cross-section.
Each element of the CA that contains at least one cell will render a sphere proportional in size to its volume. This delivers a fairly easy to understand rendition of the model, however it is neither positionally nor numerically accurate and so must only be considered as a representation. As an extension to this, to create a visualisation that is easier for the viewer to comprehend, geometric calculations can be made to move the rendered spheres to more positionally accurate places. In order to do this an algorithm was written that applies itself to the smaller (and therefore not entirely filled) spheres moving them closer to their most massive neighbours. The calculation for changing the position works to move the unit a distance proportional to its volume. In the case that several neighbours have the same highest volumes, the algorithm randomly chooses which one to be closer to. This can be further extended to move an equal distance to each of these equally massive neighbours, but at this time it has not been implemented. This gives a more realistic idea of where they would be if they actually had individual spatial locations, though again it must be noted that these are volumes and not individual units.
A more detailed look at the outside of the spheroid is desirable so as to provide a greater level of realism for the model’s exterior, even though the model deals with abstractions a better formulation of the visualisation can be created. A function could be run that takes the outermost spheres and actually calculates specific positions of each cell it represents based on their surrounding environment. It is a highly likely idea that they will be most heavily concentrated next to their most massive neighbour or neighbours, due to the high possibility that they had large contributions from them. A new array could be added into the system that keeps track of where the highest contributions have actually been made to a zone from its neighbours, and this information could be used instead of the more simple ‘highest volume’ selection process.
This would give a far more realistic visual representation of the model, however with a larger number of cells being represented by each CA location, the number of calculations and actually represented points would skyrocket and this approach is only suggested on a powerful machine capable of handling such intensive calculations and real-time rendering.

4. DISCUSSION

This model is based around a volumetric representation that allows flexibility of cell position in an abstract manner. Although the model does allow cells to exist in any configuration they desire, this is only an abstract concept as there is no positional data related to a cell other than the general area it exists within, it is however still a more accurate representation than aligning cells into a spatial grid.
The volumetric nature of the program allows it to run far quicker than if every actual cell had a defined position. By reducing the amount of data that is involved with position, calculations are greatly reduced. Human cell’s are on average about 20 micrometers in diameter and tumour spheroids are often found to be between 0.1 and 0.5 millimetres in width []. A volume of a 0.5mm spheroid can be calculated as about 4 million tumour cells and this limit was applied to the model. A population of a few cells can be grown to this size and visualised in the OpenSG virtual environment in less than a minute. Extensions to the model would definitely impact the length of time the model takes to finish calculating, but it is still believed that a realistic model can be generated in a relatively short period of time.
The ability of the model to simulate cells continuing to divide and create new daughter cells, even in an environment where there is no space, by shifting it’s surrounding cells out of them way is a more realistic rule, however in it’s current state it has drawbacks. Although the algorithm has never broken down in any trial runs, it could theoretically get stuck in a loop, constantly moving between sites that are already full. This looping concept becomes obvious when the number of times a cell needs to be moved into a neighbouring site is compared with the number of cells that actually get shifted, typically they differ by a factor of 10.
One way to fix this would be to alter the algorithm so that it constantly tries to put shifted cells into the outer layers of the spheroid, this would dramatically cut down the number of moves required to successfully finish the shifting of cells.
The nature of coding the model as functions with global array definitions means that the code is very user friendly and that new or re-written functions can be added with a minimum of additional code required for the new code to work within the program. This means that the model is easily extendable.
Current additions that are being worked on for the model are the introduction of the model to an inhomogeneous environment that would then allow necrosis to be introduced to the model thus adding to the realistic properties of the model. Further to this the introduction of mutations as cells continue to divide, that change the daughter cells on some level will allow the model to represent different clonal populations within the tumour and allow cell competition in the spheroid to be modelled.

5. CONCLUSION

This Report has detailed the formulation of an original model of tumour spheroid growth and it’s visualisation within an interactive virtual environment.
The model is able to grow from just a few cells to a spheroid made up of a population of millions. The model contains the realistic properties of cell movement induced by a cell replicating in a mechanically enclosed environment and does not tie cells into a particular configuration or spatial pattern, although this is of lesser importance as no cell has a specific location.
It has been shown that the model is easily extendable and is the subject of ongoing adaptations to improve its lifelike properties.
It is believed that this model can be considered a new formulation of the widely used CA model’s of tumour spheroid growth.
Acknowledgement: The author would like to thank the project supervisor Dr Simant Prakoonwit.

So yeah, thats the first draft of my conference paper based on my MEng project. Kudos to anyone who manages their way through the whole thing, either you're bored or insane, either way if you struggled through I love you for the effort! I'd like people to know that I haven';t just been ignoring everyone, I've actually been working most of the time.

excedoeffluo @ 2007-09-24T00:47:00

2007-09-23T23:48:05Z

Wookiefest was amazing! Yay!

I'm 23 today. Meh...

Here again indeed

2007-09-08T00:35:41Z

Hey hey.

So yes, that was a blast. Spent 14 days in Malaysia and 16 in Japan with my bro. Highlights below, pictures to follow when I can figure a good way of getting them online in such a way that they don;t loose all their quality and look shit (*Paging the Icelandic Goddess of Photos*).

Malaysia:

Kualar Lumpar - Awesome city, Petronas Towers are great looking as is the view from the KL tower.

Perentian Islands - Pretty much the most beautiful place I've been in my life. Basically paradise isalnds, after initially rocking up with no reservations to find that the places we'd been aiming to stay were all full and that there was no other boat home, my brother missioned it around and found us a place a whole lot better, cheap and the sea was all of 20 meters from our door. Watching lightning storms and the stars from beach chairs with the sea lapping at our feet and sipping a few beers was pretty incredible. Some bloody good snorkling was to be had around the isalnds, stacks of coral and varieties of fish. Kyaking around the islands was certainly a highlight too.

Japan:

Tokyo - A fantastic bustling city of neon. Stayed in the Shinjuku area, wolrds busiest station there, with 3 million people going through it a day, it was certainly bloody busy. Got a great view of the city at night from the top of the Park Hyatt Hotel, where part of Lost in Translation was filmed.

Kyoto - Incredibly beautiful city with amazing Shrines and Temples. Saw the largest wooden building in the world, or atleast what you can of it, as it's currently within a new stucture whilst being repaired.

Nara - Nothing says 'safe place to go' like a huge woodedn building sat inside of which is a 19meter tall bronze buddah in the middle of a wood, filled with skittish deer, during a storm. At least thats what I was thinking as we headed in on the train and the clouds gathered. Turned out there were fewer trees than I thought, a lot more deer, all very friendly (especially if you had food), but the building and statue were truely awe inspiring, and thankfully the storm stopped shortly after we got there, still got soaked though.

Kummamoto - Stunning rebuilt castle and grounds.

Himeji - Most beautiful of Japan's suviving all wooden castle apparently. We certainly agreed.

Osaka - Great aquarium, worlds laregst IMAX (supposedly, 20 by 28 m) saw a 3d underwater documentary, was beautiful.

Apologies if any of this is garbled or confusing, got in at 6:30 this morning after a 12 hour flight, predceeded by a one hour wait, a 2 hour flight, a one hour wait and a 5 hours and 45 minutes flight. I have no idea how much sleep I got during that, but if it was much over 3 hours, i'll be surprised. So I'm in some kind of odd state right now and think that driving the car a short while ago was prehaps a bad idea, especially as the car felt like it was floating. Although I don't feel tired I can envision waking up with my face in the keyboard at some point in the near future.

Hope ya'll have been well Postcards will have British stamp as are being sent from here, never quite manged to send them when over there!

Photos will hopefully be sorted by sometime tonight or tomororow hopefully, took almost 700, so should hopefully be able to find at least a few that are good...

RIP Shaun Christopher

2007-06-15T14:49:51Z

A great loss. Goodbye my friend...

Shaun C was one of the very first people I met at OTFCC 2003. It was the night I got to the hotel and I got a call, it was Shaun asking if I wanted to meet up, I was tired but thoguht I'd be polite and go say Hi. Ended up meeting him in the lobby and chilling in his room for a bit just chatting about tf's and other 80's series. He was a great guy.

I saw him on and off during the convention and am sad that I didn't spend more time with him, now that I can never do that again. I remember him always being reallysmiley and happy, asking me how everything was going and if i was enjoying myself. And at the end he collaredme into having a photo taken with him. I need to find that photo and fucking frame it.

After that I chatted tohim on /msn and on the boards, but never as much as I could have. In all honesty I just never made time for him as I found him a little overbaring at times, which was me just being a dick. But I do remember having some great chats with him a few years back, to this day he is the only other person I've met to share my avid love of Sailor Moon.

And now he's gone.

That incredibly nice man, who always seemed to find something to like about people. That great guy that took the jokes in full stride. That proud man who always held his own. That fantastic guy who has been remembered so beautifully in this thread. Is gone. And I am sad. The tears have been flowing and I will be sad for much time to come.

I wish I'd made more time for him and I wish I'd been a better friend to him.

Thank you for everything Shaun. You, your family and friends are in my prayers.

Although I will not be at Botcon, let me know exactly when the pool jumping is planned for, I have a pool here and will jump into it fully clothed too if thats to be part of his tribute.

Goodbye my friend.

I only wish this was a drunk post.

2007-04-13T00:51:30Z

I am so screwed that i have moved beyond the mortal realm of screwed.

I am now in a place of sheer bollockedness that any description i could give would be pale in comparision to the real thing.

I have 12 days to basically write up my project, my dissertation, bachelors equivilent project.
I have 12 days to finish two (thankfully small) SLEASE assignments.
I have about 4 weeks to learn and revise and get to know my entire years worth of course.
I have no job lined up for the summer.
I have near zero prospects.
I know absolutely nothing.
Almost everyone is completely ignoring me.
I am being compeltly blanked by people that seem to view me as a source of help, but once it's given, I'm not even worthy of speaking to them more than once in several years.

Needless to say that this is all pissing me off. An awful lot of it is my own fucking fault, but the things that i can't control, that still act against me, no matter what I do, realy get to me.

And you know what pisses me off the most? I have a great fucking life. I truely do. But all I've ever manged to do is just fuck it up beyond belief.

But what really gets to me, what cuts right down to the bone, is the people that ignore me. People who only dain to speak to me once a year, when its in their own interest or if it suits their purposes.

If Bimma and Pete didn;t talk to me, I might start to believe that no-one on the face of this earth views me as worth while bothering with.

Ever screwed yourself over?...

2007-04-03T00:02:22Z

I ask simply because i've done it and was wondering if anyone else has managed to monumentally fuck themselves in someway or other through their own good intentions?

I'd never really come to terms with the fact that being the nice guy, solicited or no, doesn;t just let you help people, it leads to far more.

I don't regret a single time i've done anything whatsoever thats benefitted someone else, but I've come to realise that in my endeavours of good will, i've reached a little further than was neccesary, and compromised myself.

I hate seeing people in pain and suffering, be it physical, mental, spiratual or whatever, and feel a deep need to help. And through performing this help I find myself feeling a happier man, and i find myself being a man reaching beyond his sensible boundaries.

I'm no saint, not by a long freaking shot. I'm no good samaritan. I don't wnat anyone reading this to think that. I'm just posting this to see if anyone else has been in a similar situation, where you realise that your a 'help junkie', someone that can't help but help others, no matter the consequences.

Anyone else out there like me?...

My Valentine Postbox

2007-02-15T00:52:28Z

First update in a long while, would post more, but need sleep, so i thought i'd post something completely ridiculous, like this:

Well its a few hours late, and i doubt anything will come of it, but screw it.

Valentine Postbox

leave excedoeffluo a valentine's message

get your own valentine postbox

excedoeffluo @ 2006-09-24T10:49:00

2006-09-24T09:50:29Z

22 today. Yikes!...

New Avatar

2006-04-03T00:59:05Z

Just testing out the new avatar, done in that yahoo avatar thing, expertly put into a LJ avatar format by Miss Tricky Disco, thanks Bimma!

I guess it looks a bit like me, certainly more like me than my south park version does!...

excedoeffluo @ 2006-02-05T00:44:00

2006-02-05T00:45:17Z

Get your own spectral analysis from Area 23^®

Hmm, purple, nice!...

excedoeffluo @ 2006-01-18T15:51:00

2006-01-18T16:03:41Z

So i haven't updated in ages, thanks for the nudge tricky :), so whats new with me?

Not much, same old same old. Been drunk alot, been VERY drunk alot, been to th casino a lot, actually won quite a bit. Uni works not really happening, really need to work on this neural networks projct.

Well thats it really, i'll post more if i think of it! Hope ya'll are good :)

excedoeffluo @ 2005-11-15T00:17:00

2005-11-15T00:20:58Z

So my uncle died today, or rather yesterday as i guess it is now, heart attack at 44. Damn fucking shame, i loved the guy, moustache like mario and a great guy to talk too. Guess the funeral is sometime soon.

Hope you guys are good. finally got my phone this week, my numbers the same as i was, would appreciate your guys numbers again.

Keep safe.

Gone for a bit...

2005-10-17T10:28:07Z

Sometime on saturday, whilst we were all downstairs, some twats swung into my room from the firescape next to it, and knicked my laptop and phone, so i'll be ogne for a bit from the net till i can get things sorted...

excedoeffluo @ 2005-10-09T16:14:00

2005-10-09T15:18:37Z

9 pubs, a strip joint and trying to join a casino.

And we were only meant to be watching the football...

Good day yesterday, went to a bunch of places i've not been to before, got very hammered. Strip club prices are fucking high, but the girls were hot. The casino told me they couldn;t accept my application form as it had to be atleast somewhat legible for their records, but what do you expect, i was very very drunk at the time.

Good times, good times!...

excedoeffluo @ 2005-09-24T00:46:00

2005-09-23T23:47:05Z

21 long years...

excedoeffluo @ 2005-09-13T12:27:00

2005-09-13T11:30:43Z

And the moral of the story is, i should drink less and think more. Apologies for last night to anyone that has any idea what i'm talking about. For those that don't, don;t worry it was just me being my usual stupid self, have a few drinks then getting upset and annoyed about something that doens;t actually matter, and something i can sort out fairly simply. I really need to get a grip of myself...

For everyone i hope you have a good day :).

excedoeffluo @ 2005-08-27T01:10:00

2005-08-27T00:10:36Z

The hunt for TT screamer has BEGUN!...

excedoeffluo @ 2005-06-30T02:13:00

2005-06-30T01:15:26Z

MIKE OLDFIELD LYRICS

"Moonlight Shadow"

The last that ever she saw him,
Carried away by a moonlight shadow.
He passed on worried and warning,
Carried away by a moonlight shadow.
Lost in a river that Saturday night,
Far away on the other side.
He was caught in the middle of a desperate fight
And she couldn't find how to push through.

The trees that whisper in the evening,
Carried away by a moonlight shadow.
Sing a song of sorrow and grieving,
Carried away by a moonlight shadow.
All she saw was a silhouette of a gun,
Far away on the other side.
He was shot six times by a man on the run
And she couldn't find how to push through.

[ Chorus ]

I stay, I pray
See you in heaven far away.
I stay, I pray
See you in heaven one day.

Four a.m. in the morning,
Carried away by a moonlight shadow.
I watched your vision forming,
Carried away by a moonlight shadow.
Stars move slowly in a silvery night,
Far away on the other side.
Will you come to talk to me this night,
But she couldn't find how to push through.

[ Repeat Chorus ]

Far away on the other side.

Caught in the middle of a hundred and five.
The night was heavy and the air was alive,
But she couldn't find how to push through.

Carried away by a moonlight shadow.
Carried away by a moonlight shadow.
Far away on the other side.
But she couldn't find how to push through.