"Radiology" in medicine ."Image processing "in Engineering

 ""Radiology" a Subject which is one of the most desired and most interesting field in Medicine and also where most of the money lies in" this is one of the doctor whom i am currently working with had to say for asking what do you feel about radiology , If some one would have asked me what to you mean "IMAGE PROCESSING" probably i would have given the same answer with reference to engineering

Next question i asked which modality do you prefer more  the question which came to mind spontaneously well the answer was more like expected  "MRI/CT" , 

From  a patient perspective ULTRASOUND is the safest but why MRI??? / CT???, "MRI and CT gives us more accurate results and so why take risk with ultrasound and when patients are   affording to pay  for higher modality scanning, and  with new machines the output are more faster ".

 

How accurately can you treat me if i have some problem was my next question "I don't treat i just diagnose ".

So the next question is the simplest one how accurately can i be diagnosed particularly using Ultrasound,Answer was look you cannot compare the modalities,some diagnosis particularly require Ultrasound only ,in that case we have to opt for ULTRASOUND ,In percentage wise say around 75% .So here starts the search for my project objective WHY??? only  75% not 90% or more ...

      With a Lot of Literature survey and meetings with radiologist ,the problem encountered using Ultrasound is it is a more of Subjective Type as all other modalities are ,and due to its image quality a bit like 1950s Black and white TV with lot of interference what we call "SPECKLE NOISE”. It is difficult to visualize, This article is not about removing speckle noise ,Lot of PhD have been done in this area ,I am referring to something else ,Why not make ULTRASOUND imaging a bit more objective ???.If it could bring accuracy above 5% at least then this work is worth doing. Instead of segmentation and other stuffs ,During my literature survey i have seen a lot of  segmentation and finally people concluding Cancerous, tumor, malignant is that very  necessary ???,The most craziest part is i saw most of them do there  MTECH project in a year and finally conclude look here is the TUMOUR ,i will segment it and tell you what kind of tumour, Are we gone Crazy ,lost our ability to think ,if i can detect a tumour in an image and conclude this is "that kind of Tumour" why do a doctor need to study 7 years ,BIOMEDICAL ENGINEERS do help in the diagnosis but do not conclude saying it what kind of and  Doctors are meant for that just give them a aid to visualize the area rather than giving them judgment

Something that drives me crazy is "compression algorithms " ,Biomedical engineers stop doing this ,this is not what you are meant to do compressing things

Here comes the most amazing part a Cardiologist by seeing an ECG can tell you what kind of arrhythmia it is in not more than 7-10 min ( value added after consulting). So do we really need to build an algorithm saying what kind of arrhythmia????, ok you want to build one ,surely proceed with it ,but don't conclude saying it is so & SO...GIVE SOME VALUES ,PUT THE BALL IN DOCTORS COURT ,THIS IS WHAT IT SEEMS TO POSSIBLY BE. 
If you are building something for self diagnosis(eg.BP monitoring system)  then you have to be pretty sure no "assume" ,"if" are possible

"In case you are building some algorithm for just a normal laymen not a doctor then that is a risk ,because you can never trust a persons body ,there is never a second chance"

MACHINE ALGORITHIMS should always be a supportive tool...Biomedical signals are always non deterministic you can never compare one persons signals with others ,so just build something to aid  the doctors judgement, not to help in concluding a doctors judgement . 
"DOING FOR THE SAKE OF DOING IS NOT WHAT A BIOMEDICAL ENGINEER IS MEANT TO BE , AN IMMENSE OPPURTUNITES AND FUN AWAITS IN BIOMEDICAL ENGINEERING ,COME JOIN THE FAMILY OF BIOMEDICAL ENGINEERING ",if you want to have fun with learning there is no other feild than this so for more info on biomedical engineering do visit BIOMEDICAL  ENGINEERING