Images within tables (chemical compounds in patents) do not parse correctly

[geoHeil]

How can I handle tables of chemical compounds which include images?

See Page 2,3 as an example of this PDF:
WO2024040109-eval-small.pdf

I am currently exploring the use of docling with various OCR engines - also rapidocr and so far receive

In particular the reference to the images seems to be removed from the table. See:

- 146-

![Image](WO2024040109-eval_artifacts/000003_bac7203f58af638707de41cd7b0c02616048b5460f81dd4400286fb4ec612b68.png)

![Image](WO2024040109-eval_artifacts/000004_a0185af154173824b6eed213c2d7304c4cae5d1200c72d9235430b047fc4e92b.png)

Table 1

| 井   | Structure   | IUPAC                                                                                                                                                                  |   LCMS (M+H)+ | 1H NMR                                                                                                                                                                                                                                                                                                                                                                                                                          |
|------|-------------|------------------------------------------------------------------------------------------------------------------------------------------------------------------------|---------------|---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| 1-4  | HO          | 5-ethynyl-6-fluoro- 4-{8-fluoro-2-[(1- {[(3R)-3- fluoropyrrolidin-1- yl]methyl}cyclopro pyl)methoxy]-4- (morpholin-4- yl)pyrido[4,3- d]pyrimidin-7- yl}naphthalen-2-ol |         590.5 | DMSO-d6)  ppm = 10.23 - 10.13 (m, 1H), 9.07 (s, 1H), 7.98 (dd, J= 9.3, 6.0 Hz, 1H),7.50 - 7.44 (m, 1H), 7.40 (d, J= 2.5 Hz, 1H), 7.20 (d, J= 2.5 Hz, 1H), 5.28- 5.07 (m, 1H), 4.35 - 4.21 (m, 2H), 4.08 - 3.97 (m, 2H), 3.96 - 3.88 (m, 2H),3.86 - 3.79 (m, 5H), 2.91 - 2.76 (m,2H), 2.70 - 2.55 (m, 1H), 2.47 - 2.38 (m, 2H), 2.36 - 2.28 (m, 1H), 2.19 - 2.03 (m, 1H), 1.93 - 1.76 (m, 1H), 0.64 (m, 2H), 0.48 - 0.42 (m, 2H) |
| 1-5  | HO          | [(dimethylamino)m ethyl]cyclopropyl} methoxy)-8-fluoro- 4-(morpholin-4- yl)pyrido[4,3- d]pyrimidin-7-yl]- 5-ethynyl-6- fluoronaphthalen-2-                             |         546.2 | DMSO-d6)  ppm = 10.17 (dd, J= 5.0, 3.0 Hz, 1H), 9.11 - 9.06 (m, 1H), 8.02 - 7.95 (m, 1H), 7.51- 7.39 (m, 2H), 7.20 (d, J= 2.8 Hz, 1H), 4.28 - 4.24 (m, 2H), 4.04 - 3.90 (m, 4H), 3.83 - 3.81 (m, 3H), 2.53 (d)                                                                                                                                                                                                                  |

- 147 -

![Image](WO2024040109-eval_artifacts/000005_e6553cbad9c280959f1440ed881d5f4f42008773bf89543e24f59a17adb5f907.png)

![Image](WO2024040109-eval_artifacts/000006_8f5f2d11c1754182c3960441583c0e9d1b206e20331812db6172708c3b1053b5.png)

|     |      |                                                                                                                                                                                           | (m, 4H), 0.70 - 0.64 (m, 2H), 0.51 - 0.39 (m, 2H)                                                                                                                                                                                                                                                                                      |
|-----|------|-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
| 1-6 | HO、 | 4-[8-fluoro-4- (morpholin-4-yl)-2- ({1-[(piperidin-1- yl)methyllcyclopro pyl}methoxy)pyrid o[4,3-d]pyrimidin- 7-yl]naphthalen-2-                                                          | Hz, 1H), 7.52 - 7.43 (m, 1H), 7.40 (d, J= 2.8 Hz, 1H), 7.19 (d, J= 2.8 Hz, 586.1 1H), 4.30 (s, 2H), 4.07 - 3.99 (m, 2H),3.97- 3.90 (m, 3H), 3.85 - 3.79 (m, 5H), 3.18 (s, 2H)2.11 (s,                                                                                                                                                  |
| 1-7 | HO   | 4-(2-{[(2R,7aS)-2- fluoro-hexahydro- 1H-pyrrolizin-7a- yl]methoxy}-8- fluoro-4-{8-oxa-3- azabicyclo[3.2.1]oc tan-3- yl}pyrido[4,3- d]pyrimidin-7-yl)- 5-ethynyl-6- fluoronaphthalen-2- ol | 'H NMR (400 MHz, DMSO-d6) 8 ppm = 10.16 (s, 1H), 9.08 (s, 1H), 7.98 (dd, J=9.5, 6.0 Hz, 1H) 7.47 (t, J=9.1Hz, 1H), 7.40 (d, J= 2.5 Hz, 1H), 7.18 (d, J= 2.8 Hz, 1H), 602.3 5.46- 5.21 (m, 1H), 4.53 - 4.44 (m, 3H), 4.34 - 4.27 (m, 1H), 4.15 (s,1H),3.92 (d, J= 7.8 Hz, 1H),3.79 - 3.65 (m, 2H), 3.29 - 3.27 (m, 2H), 3.21 - 3.11 (m, |

the able is broken into alternating elements. How can I OCR the table with links/references to the images correctly so the connection of table row / compound ID and compound visualization is not broken?

[geoHeil]

#992 might be related

[geoHeil]

#1049 also might be related