0:01

My next tip, don't miss the big one.

0:03

So I gave you an example of someone who

0:06

identified the lung cancer

0:07

and the mediastinal lymphadenopathy,

0:09

but missed the healed rib fracture.

0:14

Great. You know, if you have to make mistakes,

0:17

make mistakes on old things

0:18

that have no clinical importance,

0:21

but you have to maintain your vigilance

0:23

and be even more highly vigilant on things

0:26

that are the big ones, the rule out ones.

0:30

So for me, you know, strokes, aneurysms on, on, uh,

0:35

CT scans of the head and CTAs.

0:37

Those are the biggies.

0:39

But you know, you're gonna focus more carefully in the

0:44

screening of cancers for blood vessel detection,

0:47

detect dissections and emboli.

0:50

Um, you know, when we do RAD p

0:53

or when we do peer review missing

0:57

minor mucosal thickening in the trachea

1:03

or minor mucosal thickening in the bladder

1:07

or minor mucosal thickening in the sinuses,

1:10

I can handle those type of mistakes.

1:12

I feel horrible if I've missed a cancer, missed an aneurysm,

1:17

missed a bleed, for example.

1:19

So that's okay.

1:21

You know, be humble about missing cases and,

1:25

but when there are minor problems,

1:29

you can sort of dismiss them.

1:31

I think that this is where

1:33

artificial intelligence detection software really will

1:36

help radiologists.

1:37

It's already very useful for breast imaging

1:42

and mamma mammographic screening.

1:44

Um, I see it as sort of a solution for some

1:49

of the misses that we have, particularly at the end

1:52

of shift when we're exhausted.

1:54

So, um, I would say to the extent

1:58

that your practice can afford it

1:59

and has made the judgment that with the risk benefit

2:04

or time lost, time gained benefit

2:07

that you employ artificial intelligence for these big

2:12

deals of cancers, aneurysms, strokes, bleeds, et cetera.

2:18

In the breast imaging world,

2:20

obviously double reads are often useful

2:23

for not missing a breast cancer.

2:27

Um, hopefully AI will serve as

2:31

that sort of double read for increasing

2:36

applications in radiology.

2:39

So make sure that these diagnoses are paramount in your

2:42

head.