RegEMR: a natural language processing system to automatically identify premature ovarian decline from Chinese electronic medical records

Table 2 The scoring model of DOR, POI and POF

	Criteria	Score
The scoring model of DOR
AFC ≤ 5	Absent data	Null
	On no occasion	0
	Only on one occasion	1
	On at least two occasions < 4 weeks apart	2
	On at least two occasions > 4 weeks apart	3
AMH	Absent data	Null
	> 1 ng/mL	0
	≤ 1 ng/mL	3
FSH > 10 IU/L	Absent data	Null
	On no occasion	0
	Only on one occasion	1
	On at least two occasions < 4 weeks apart	2
	On at least two occasions > 4 weeks apart	3
Sum		DOR Score
The scoring model of POI
Menstrual cycle	Absent data	Null
	24–35 days	0
	< 24 or > 35 days	1
FSH > 25 IU/L	Absent data	Null
	On no occasion	0
	Only on one occasion	1
	On at least two occasions < 4 weeks apart	2
	On at least two occasions > 4 weeks apart	3
Sum		POI Score
The scoring model of POF
Menstrual cycle	Absent data	Null
	24–35 days	0
	< 24 or > 35 days	1
FSH > 40 IU/L	Absent data	Null
	On no occasion	0
	Only on one occasion	1
	On at least two occasions < 4 weeks apart	2
	On at least two occasions > 4 weeks apart	3
Sum		POF Score

The total score is the sum of the scores of its subsections. The POI and POF scores are calculated only for women under the age of 40

ISSN: 1472-6947