I would value the opinion of a sub specialised paediatric sonographer on this topic.
I will pick the brains of a couple of the partners of the practice I wok at who also work at the kids hospital.
From my perspective I would think the clinical presentation may be more important than the ultrasound findings to differentiate between the entities. I would also include blood results.
A post traumatic patient is likely to be afebrile and have normal white cell count etc indicating there is no evidence of an infective process. Skin bruising etc is an obvious clue as is the presence of fat contusion / necrosis in the sub-cutaneous fat superficial to the bony abnormality.
Septic arthritis: Any joint with an effusion in a paediatric patient with pain in the joint or a limp etc must be suspected to be a septic joint regardless of the “cleanliness” of the fluid within the effusion. Again the clinical picture and blood profile will push you towards or away from the decision to aspirate the joint for culture.
Osteomyolitis I think of as a disease of the shaft of the bone between the joints if you like. In osteomyelitis I am looking for periosteal changes like a hypo echoic line adjacent to the bone cortex and vascularity etc.
I will add more to this post after consulting my clever colleagues.