If you are keen to do the experiment it should be possible. The volume measurement markings will no longer be exactly calibrated but the error should be smaller than casual measuring would require.

You need to prepare 3 things

• An adjustable temperature hot air gun (or a blow torch if looks don't matter)
• A curved insulating mask (thick card stock or many thin cards) with a hole the size of one indentation (or all of them if you are a gambler)
• A hose and connection to a source of vacuum such as your mouth suction that can be fitted to bottle neck

Hook up suction, align mask, heat with 120degC hot air and use judgement. Hold vacuum until plastic cools again.

If you are keen to do the experiment it should be possible. The volume measurement markings will no longer be exactly calibrated but the error should be smaller than casual measuring would require.

You need to prepare 3 things

• An adjustable temperature hot air gun (or a blow torch if looks don't matter)
• A curved insulating mask (thick card stock or many thin cards) with a hole the size of one indentation (or all of them if you are a gambler)
• A hose and connection to a source of vacuum such as your mouth suction that can be fitted to bottle neck

Hook up suction, align mask, heat with 120degC hot air and use judgement. Hold vacuum until plastic cools again.

EDIT:
The best source for a mask would be a cardboard shipping tube or roll material core that the bottle would fit in with not too much space. Open the hole with a few drill holes and shape it with a fine rasp.

The aim is to limit the amount of hot air that gets past the mask.

The required 120 degC is not very hot and most any non-plastic fibrous material will work.

Sticking a layer of tin/aluminium foil to the outside (or using some form of metallized heat shielding) might be the best if you want to go into production. The reflective shielding would probably allow you to use radiant heating (slower but more controllable) and it would basically be localised vacuum forming.