ESTRO 2024 - Abstract Book

S1748

Clinical - Lung

ESTRO 2024

achieved a significant target reduction during chemoradiation and an adaptive planning performed. Patients’ characteristics are summarized in Table 1.

Table 1 Summarizes patients characteristics. Non Adaptive (n, %)

Adaptive (n, %)

Total (n, %)

p

Age <70 yrs >70 yrs

9 (25.7)

4 (11.4) 8 (22.9)

13 (37.1) 22 (62.9)

0.517

14 (40.0)

Histology Squamous

8 (22.9)

4 (11.4) 8 (22.9)

12 (34.3) 23 (65.7)

0.618

Adenocarcinoma

15 (42.9)

Stage IIIA

17 (48.6)

5 (14.7) 6 (17.6)

22 (62.8) 13 (38.2)

0.262

IIIB

7 (20.6)

Chemo before RT-CT Yes

9 (25.7)

5 (14.3) 7 (20.0)

14 (40.0) 21 (60.0)

0.583

Not

14 (40.0)

CTV mean reduction (%)

9.6

37.7

We investigate eight different multimodal late fusion rules and two patient-wise aggregation rules in comparison to unimodal approach. The experiments showed that the proposed fusion-based multimodal paradigm, achieving an AUC equal to 90.9%, outperforms each unimodal approach, suggesting that data integration can advance precision medicine. Figure 1 shows a radar chart plotting the performances in terms of AUC of the various unimodal and multimodal approaches. The different combination of pathomics, radiomics and semantic features obtained different AUC ranks. The combination of the three modality resulted in the highest performance of the model.