LINACwatch

UsING LINACWATCH for all treatment sessions delivered with IMRT techniques and for every patients Eliminates the need TO PERFORM machine QA?

NO - in order for pertinent use of LOG files, one must ensure that the data contained in the LOG files is in line with reality. For example, for leaf position and speed, by performing "Picket Fence" tests analysed in parallel; by LINACWATCH for the LOG files and by another software, like QUALIMAGIQ and its MLC-dyn module*, for portal images. However this second software must be capable of detecting deviations of less than 0.1 mm in slit position or width.
In sum, LINACWATCH is a complementary solution to machine QA and pre-treatment QA. It enables effortless checking of the stability of the main irradiation parameters during and between each irradiation session.
(*): MLC-dyn module performs a rapid, precise and comprehensive quality control of the MLC used in dynamic mode.

USING LINACWATCH FOR ALL TREATMENT SESSIONS DELIVERED WITH IMRT TECHNIQUES AND FOR EVERY PATIENTS ELIMINATES the need TO PERFORM pre-treatment QA?

NO - BUT using LINACWATCH for all sessions for every patient allows you to reduce the frequency of pre-treatment QA and to limit it to special cases not statistically studied by previously tested treatments.
By using LINACWATCH you will be more selective about your pre-treatment QA.
Furthermore, by combining the pre-treatment controls with LINACWATCH controls, you can be sure that the performance observed during the pre-treatment QA is stable for the duration of the treatment and not only just before the treatment...

Some dynamic log file analysers introduce a TPS in the control loop. LINACWATCH does not. Can you explain why?

The most important reasons are:
1- To provide analysis results less than 4 seconds after beam end. The therapist* can then easily and rapidly decide to stop the Linac in case of repeated bad results.
2- To be sure that deviations between the prescription and the treatment are attributable only to the Linac and not to the calculation algorithm and other beam data used by the second TPS.
3- To offer an honest solution, i.e. an ON LINE Beam QA integrated into patient workflow. LOG files contain no data concerning the patient. Don't ask LOG files to report more than they contain! A solution which includes a TPS may deceive you into thinking that it is performing a control of the treatment when in fact it performs a beam control diluted into a patient model that is not that of the patient being treated and also diluted into a dose calculation that may be less accurate than the one you do on your TPS. LINACWATCH is very efficient because it doesn't dilute original data contained in the LOG files. LINACWATCH follows the Linac in real time and helps you to efficiently identify the cause of beam defects.
4- To have a low priced solution which allows you to be easily equipped with a solution which strongly contributes to providing secure treatments for all of your patients treated with IMRT techniques.
(*): Another possibility with LINACWATCH, if you refuse to install an additional screen in the Linac control room, is to centralise results from all of your Linacs on one screen, for example in the dosimetry room, and to use audio alerts possibly usable with LINACWATCH, to warn people in the room in the event of non-compliant results.

Dynamic Log files are not measurements and therefore analysing these log files will never be as good as measuring FLUENCE at the exit of the collimator!

NO, this is not true. The dynamic log data contains results from a lot of measurements: chamber signals, encoder signals of each leaf motor rotation confirmed by the potentiometer signal produced by a pin attached to each leaf edge, encoder signals connected to the motor rotation for jaw translation, gantry and collimator rotations. And these measurements are taken every 20, 50 or 250 ms with an accuracy of 0.01 MU, 0.01 mm and 0.01°! 
When you measure the "dose" at the collimator exit, what are the spatial and time resolutions of your detector? 2.5 mm or 8.0 mm for spatial resolution. Worse, some detectors integrate everything to have the Dose.Area Product. But by integrating everything, you see nothing...

Could you compare AN EPID based solution to a LOG FILE BASED SOLUTION?

YES we can!
The main argument of EPID based solutions is to take the patient into account during the treatment.
Two types of EPID based solutions are on the market. The first one is only a gamma index comparison of the integrated EPID image acquired during fraction N with the one acquired during the first fraction. In this case the 1st fraction is the target for the rest of the treatment. As you know, more than 70% of non-conformities occur during the 1st fraction. With this type of solution, you lose the possibility of detecting the most frequent non-conformities definitively! Also please note that some studies show that when patient thickness increases or decreases 20 mm or when the patient is shifted 20 mm, nothing is really detected by the gamma index pass rate... 
The second type of EPID based solution consists in recalculating the dose distribution into the patient from the EPID signal. These solutions dilute beam defects into the patient and into a complex dose calculation algorithm. Moreover this complexity does not allow you to have real time detection of Linac defects.
In sum, all EPID based solutions dilute beam defects and reduce the capability of rapidly detecting a problem on the Linac. 
Another crucial problem is the premature wear of the EPID (20 to 50 times faster!), which will push the Linac manufacturer to significantly increase the cost of your maintenance contract...

Then finally the ideal solution is a dose calculaton based on the beam of the day defined by the DYNAMIC LOG file, applied to the CBCT of the Day?

YES for sure BUT this solution is not yet on the market because it still faces many technological blocks. The first is the ability of your TPS to accurately distort the reference structures based on the comparison of the CBCT of the day with the reference CT scan, especially when tissues disappear or appear which is a very common situation.
Please note that when your TPS will be ready for this adaptive workflow, LINACWATCH can already generate and export RT-Plans containing LOG file data, i.e. defining the beam of the day.

With DYNAMIC LOG FILES analysis, it's impossible to detect a calibration error!

YES, this is true. If you make a mistake, for example you didn't respect the field size requested by the Linac during the calibration, of course LINACWATCH will not detect this error. But this is not the goal of LINACWATCH. Its goal is to control the stability of the field size over time. Normally, following a jaw or a leaf calibration, you need to control the field size with the water tank, film or EPID images to detect a possible calibration error.
In sum, LOG file analysis will never replace a good machine QA, for example performed by QUALIMAGIQ, but a good machine QA and pre-treatment QA will never replace LINACWATCH, which controls the machine between your machine quality controls!

can LINACWATCH document THAT I trEateD a patient correctly during all of the irradiation sessions?

YES of course and in a most pertinent manner.
Indeed LINACWATCH can generate an RT-Plan file including all of the information contained in the dynamic LOG files of all of the sessions. You can then load this RT-Plan representing the entire treatment onto your TPS and, using the many tools available in it, compare in the model patient described by the reference CT scan, the dose distribution delivered and cumulated throughout the treatment relative to the treatment planned by your TPS.
You do not find this documentation very complementary, or more pertinent, than the documentation proposed by the pre-treatment QA? Because in the end the pre-treatment QA only proves that before treating the patient, your Linac and your TPS presented consistent dose distributions in a homogeneous water cylinder, once the beams planned for the patient had irradiated the cylinder.