A Dose of Precaution

"and build safeguards against overexposure into their machines"

So subby thinks the engineers who designed the CAT scanners simply overlooked this?

Overexposure has always been due to operator error, overriding existing safeguards built into the machines themselves.

Who told the engineers how much too much was?

Did the bits of information get to the engineers via the managerial sieve?

Radiology, especially with regard to x-rays is a mature discipline with peer reviewed studies spanning over 50 years. Overexposure is extremely rare and to my knowledge always due to operator error.

Bridge building is even more mature, but bridges fall now and again. The difference is that you can see when it isn't happening with the bridge, not so with radiology and radiology is known to be a risk, while crossing bridges is not thought to be risky.

Over exposure issues and patient related incidents are mostly user related errors. This is the very reason why equipment designers should or must incorporate built-in safety features in their equipment designs. Fool proof built-in safety features that cannot be overridden by the operators nor by any third party service providers. Safety features that are interlocked for each and every modalities the machine is capable of safely conducting. Interlocking operations that are tied together with a normally functional equipment as stated in the manufacturer's specifications. This can be a software driven self-diagnostic routine that must be passed before the delivery of any form of energy. Ionized or non-ionized.

Overexposure is generally due to excessive xrays and scans being ordered to 1) satisfy the patient or 2) protect the doctor.

What I read is that the US population is the group most overexposed to radiation from medical sources in the world.

People from nuclear plants say that if they have a scan done and go to work all the alarms, bells and whistles go off due to the radiation they have received.

I also read that many of the scans, xrays are more precautionary than needed - lawsuit protection for the doctor/facility.

I don't believe the articles I have been reading are of the loony type - not knowing anything about medicine though I can not say that for sure.

As was mentioned, in overheming majority over exposure is the consequence of human errors or lack of knowledge. Many times the devices are neither controlled nor according to schedule calibrated.

I suppose that there are today possibilities to introduce more intelligence in the controllers and block the machine as far as some checks have not been done or if according to some data about the patient the machine requires the exposure time seems to be excessive.

Since human life is in danger some check lists should be used to oblige the personnel to verify the limits.

Of course the preparation level of the operators must be better and at regular time intervals checked.

"Since human life is in danger some check lists should be used to oblige the personnel to verify the limits."

I would say check lists that operator's has to go through prior to each and every case or procedure will be helpful. But this will not solve user-induced problems! Based on observed experiences, check lists are only good to refer to at the beginning when the equipment was just installed as well as an aide in user's training.

What will be helpful aside from having or following the things-to-do-list and a built-in self diagnostic routines will be another built-in sequentially arranged, preprogrammed steps that should be implemented prior to the delivery of any amount of energy. Internal routines that even an authorized operators, including field service engineers cannot bypassed even if it password protected.

Another is an internal automated, paralleled control systems that continuously monitors every stages as well as the overall equipment functionalities. This closed loop parallel control system will have the ability to shut down the machine if it senses any deviations on input settings versus proper output based on original equipment limits and specifications. Basically an input/ output correlations at every stage, as displayed and compared to the operator's manually set parameters .

Thank you far supporting the second § in my comment.

In fact what should be done is make intelligent machines able to control the operator skills.

When I mentioned check lists I did not think (was badly explained) only :step #5 do this +done, I thought about a check list with plausibility analysis by the controller of the machine. This was the reason I mentioned the necessity to supply data about the patient.

What is also to be done is the control of the emission, if it corresponds to the "command" the closed loop you mention is a very good aspect provided that it is also checked for consistency.

Actually it should be a series of smaller loops progressing to bigger loops until it reaches the final stage just before the delivery of any predefined and selected output. The continuing closely looped system that is built in into the parallel control system of the device. A separate or independent monitoring & controlled system that may or may not be similar in nature as compared with those of the user controls, but has the ability to shut down the whole system in case any user set parameters, or any circuit malfunction or output mis-calibration falls out of range and safe limits.

You touch a sensitive aspect the "calibration" on the spot. Do not forget that all references on a given place can get a derive of value. The always valid question who controls the controller?

With respect to the loop or loops the principle is important it is clear that a control cascade has to be implemented.

I know about a very unpleasant case, an irradiation machine for cancer treatment was not calibrated for a long time and ≈ 150 persons were over exposed more or less some of them even very much overexposed.

It was human failure because of lack of preparation. A simple timer set automatically at a calibration would have stopped the machine if the following calibration was done in due time.

It is not always necessary to make things complex some (or many) times simplicity is enough.

There are times when one can't help but to seek and look forward for a more complex non-corruptible aspect or part of technology for assistance and solutions! More often than not, patients now a days cannot just rely and count on anyone's conscientiousness, (operator's /field service techs or even MD's) to deliver proper care! Everyone seems to be in a rush like in a production environment, looking for shortcuts and trying to maximize production output, turn around times and profits!

The problem with complexity is that over proportional to it grows the risk of malfunctions.

An aspect I did not mention but has to do it now it the necessity for redundant systems as in military or space technology. So the reliability will be enhanced

That's exactly what a closely looped - paralleled control systems is all about... redundancy! Redundancy Not only in duplicating system functionalities but with a built-in intelligence. Intelligent enough to track and monitor overall equipment operational as well as the ability to override either correct simple deficiencies in operator's controlled settings or the ability to shut the whole system if deviations are sensed in any of the set and predetermined safe limits.